# WELL-BEING OF SCHOOL TEACHERS IN THEIR WORK ENVIRONMENT

EDITED BY : Caterina Fiorilli, Wong Yau Ho Paul, Simona De Stasio, Florencio Vicente Castro and Paula Benevene PUBLISHED IN : Frontiers in Psychology

#### Frontiers eBook Copyright Statement

The copyright in the text of individual articles in this eBook is the property of their respective authors or their respective institutions or funders. The copyright in graphics and images within each article may be subject to copyright of other parties. In both cases this is subject to a license granted to Frontiers. The compilation of articles constituting this eBook is the property of Frontiers.

Each article within this eBook, and the eBook itself, are published under the most recent version of the Creative Commons CC-BY licence. The version current at the date of publication of this eBook is CC-BY 4.0. If the CC-BY licence is updated, the licence granted by Frontiers is automatically updated to the new version.

When exercising any right under the CC-BY licence, Frontiers must be attributed as the original publisher of the article or eBook, as applicable.

Authors have the responsibility of ensuring that any graphics or other materials which are the property of others may be included in the CC-BY licence, but this should be checked before relying on the CC-BY licence to reproduce those materials. Any copyright notices relating to those materials must be complied with.

Copyright and source acknowledgement notices may not be removed and must be displayed in any copy, derivative work or partial copy which includes the elements in question.

All copyright, and all rights therein, are protected by national and international copyright laws. The above represents a summary only. For further information please read Frontiers' Conditions for Website Use and Copyright Statement, and the applicable CC-BY licence.

ISSN 1664-8714 ISBN 978-2-88966-064-3 DOI 10.3389/978-2-88966-064-3

#### About Frontiers

Frontiers is more than just an open-access publisher of scholarly articles: it is a pioneering approach to the world of academia, radically improving the way scholarly research is managed. The grand vision of Frontiers is a world where all people have an equal opportunity to seek, share and generate knowledge. Frontiers provides immediate and permanent online open access to all its publications, but this alone is not enough to realize our grand goals.

#### Frontiers Journal Series

The Frontiers Journal Series is a multi-tier and interdisciplinary set of open-access, online journals, promising a paradigm shift from the current review, selection and dissemination processes in academic publishing. All Frontiers journals are driven by researchers for researchers; therefore, they constitute a service to the scholarly community. At the same time, the Frontiers Journal Series operates on a revolutionary invention, the tiered publishing system, initially addressing specific communities of scholars, and gradually climbing up to broader public understanding, thus serving the interests of the lay society, too.

#### Dedication to Quality

Each Frontiers article is a landmark of the highest quality, thanks to genuinely collaborative interactions between authors and review editors, who include some of the world's best academicians. Research must be certified by peers before entering a stream of knowledge that may eventually reach the public - and shape society; therefore, Frontiers only applies the most rigorous and unbiased reviews.

Frontiers revolutionizes research publishing by freely delivering the most outstanding research, evaluated with no bias from both the academic and social point of view. By applying the most advanced information technologies, Frontiers is catapulting scholarly publishing into a new generation.

#### What are Frontiers Research Topics?

Frontiers Research Topics are very popular trademarks of the Frontiers Journals Series: they are collections of at least ten articles, all centered on a particular subject. With their unique mix of varied contributions from Original Research to Review Articles, Frontiers Research Topics unify the most influential researchers, the latest key findings and historical advances in a hot research area! Find out more on how to host your own Frontiers Research Topic or contribute to one as an author by contacting the Frontiers Editorial Office: researchtopics@frontiersin.org

# WELL-BEING OF SCHOOL TEACHERS IN THEIR WORK ENVIRONMENT

Topic Editors:

Caterina Fiorilli, Libera Università Maria SS. Assunta University, Italy Wong Yau Ho Paul, Tung Wah College, Hong Kong Simona De Stasio, Libera Università Maria SS. Assunta, Italy Florencio Vicente Castro, University of Extremadura, Spain Paula Benevene, Libera Università Maria SS. Assunta, Italy

Citation: Fiorilli, C., Paul, W. Y. H., De Stasio, S., Castro, F. V., Benevene, P., eds. (2020). Well-Being of School Teachers in Their Work Environment. Lausanne: Frontiers Media SA. doi: 10.3389/978-2-88966-064-3

# Table of Contents


Nasr Chalghaf, Noomen Guelmami, Maamer Slimani, Giovanni Del Puente, Tania Simona Re, Riccardo Zerbetto, Juan José Maldonado Briegas, Ottavia Guglielmi, Sergio Garbarino, Fairouz Azaiez and Nicola Luigi Bragazzi

*20 When Does Work Interfere With Teachers' Private Life? An Application of the Job Demands-Resources Model*

Alessandro De Carlo, Damiano Girardi, Alessandra Falco, Laura Dal Corso and Annamaria Di Sipio

*33 Teachers' Self-Efficacy: The Role of Personal Values and Motivations for Teaching*

Daniela Barni, Francesca Danioni and Paula Benevene

*40 The Impact of Emotions and Hedonic Balance on Teachers' Self-Efficacy: Testing the Bouncing Back Effect of Positive Emotions*

Ilaria Buonomo, Caterina Fiorilli and Paula Benevene


Federica De Cordova, Sabrina Berlanda, Monica Pedrazza and Marta Fraizzoli

*75 Italian Teachers' Well-Being Within the High School Context: Evidence From a Large Scale Survey*

Barbara Barbieri, Isabella Sulis, Mariano Porcu and Michael D. Toland


Loredana Addimando


*125* Trans*-Cultural Validation of the "Academic Flow Scale" (Flow 4D 16) in Arabic Language: Insights for Occupational and Educational Psychology From an Exploratory Study*

Nasr Chalghaf, Chiraz Azaiez, Hela Krakdiya, Noomen Guelmami, Tania Simona Re, Juan José Maldonado Briegas, Riccardo Zerbetto, Giovanni Del Puente, Sergio Garbarino, Nicola Luigi Bragazzi and Fairouz Azaiez

*132 Subjective Happiness and Compassion are Enough to Increase Teachers' Work Engagement?*

Simona De Stasio, Caterina Fiorilli, Paula Benevene, Francesca Boldrini, Benedetta Ragni, Alessandro Pepe and Juan José Maldonado Briegas

*142 Teaching in the Suburbs: Participatory Action Research Against Educational Wastage*

Santa Parrello, Ilaria Iorio, Filomena Carillo and Cesare Moreno

*158 Trans-cultural Adaptation and Validation of the "Teacher Job Satisfaction Scale" in Arabic Language Among Sports and Physical Education Teachers ("Teacher of Physical Education Job Satisfaction Inventory"—TPEJSI): Insights for Sports, Educational, and Occupational Psychology*

Nasr Chalghaf, Noomen Guelmami, Tania Simona Re, Juan José Maldonado Briegas, Sergio Garbarino, Fairouz Azaiez and Nicola L. Bragazzi

*166 Teachers Between Job Satisfaction and Burnout Syndrome: What Makes Difference in Czech Elementary Schools*

Irena Smetackova, Ida Viktorova, Veronika Pavlas Martanova, Anna Pachova, Veronika Francova and Stanislav Stech

*174 Effect of Teachers' Happiness on Teachers' Health. The Mediating Role of Happiness at Work*

Paula Benevene, Simona De Stasio, Caterina Fiorilli, Ilaria Buonomo, Benedetta Ragni, Juan José Maldonado Briegas and Daniela Barni


Caterina Fiorilli, Paula Benevene, Simona De Stasio, Ilaria Buonomo, Luciano Romano, Alessandro Pepe and Loredana Addimando

*238 Teaching Beliefs on Developmentally Appropriate Practice Among Chinese Preschool Teachers: The Role of Personality* Paul Yau-Ho Wong

*245 Association Between Chronic Health Conditions and Quality of Life in Rural Teachers*

Pablo A. Lizana, Gustavo Vega-Fernandez and Lydia Lera


Larissa Maria Troesch and Catherine Eve Bauer

*274 Teaching Careers: Exploring Links Between Well-Being, Burnout, Self-Efficacy and Praxis Shock*

Julie Ballantyne and James Retell

# Editorial: Well-Being of School Teachers in Their Work Environment

Paula Benevene\*, Simona De Stasio and Caterina Fiorilli

Department of Human Sciences, Libera Università Maria SS. Assunta University, Rome, Italy

Keywords: teachers' well-being, teachers' resilience, teachers' happiness, teachers' health, school's environment, workplace well-being, positive psychology

**Editorial on the Research Topic**

#### **Well-Being of School Teachers in Their Work Environment**

Teachers' well-being has received much attention over the past decades, in the light of the major increase in sick leave as well as job quitting among teachers across different cultures and countries. It is in fact well-known in the literature that teaching is a demanding, challenging profession, exposed to stress, burnout, and more in general, a high attrition rate. The majority of studies have in fact targeted negative indicators of teacher functioning, but more recently, following the mainstream of positive psychology, more attention has been devoted to teachers' well-being. Wellbeing is not just the mere absence of illness at work. Rather, it refers to healthy and successful functioning of teachers at work. In fact, while physical, psychological, and mental health refers more to the lack of impairment, well-being refers more to the ability of teachers to develop a positive though dynamic equilibrium between teachers' resources and their challenges/demands (environmental, social, individual, physical, mental, psychological).

Well-being has been found to be linked to a positive relationship with students, colleagues, and families, as well as to higher academic results of the pupils. Well-being is a multifaceted concept, comprising cognitive and affective as well as physical and mental components; it encompasses dispositional, personal, organizational, and environmental factors. Addressing protective factors may help to develop resilience and more effective ways of addressing the impact caused by negative factors of the teaching workplace. Therefore, the goal of this Research Topic was to encourage new understandings of teachers' well-being from interdisciplinary psychological perspectives.

The articles collected in this special issue cover a huge variety of approaches (both qualitative and quantitative), countries (Western such as Italy, Finland, Switzerland, and Czeck, as well as non-Western ones, such as Chile, China, Tunisia, Australia) and types of studies (empirical research, research-action, training, validation of instruments). Studies presented here vary from the training of teachers to the individual and organizational factors affecting or protecting teachers' well-being and health.

We are happy to offer such a wealth of innovative material, with the aim of not only deepening the knowledge on the subject but also offering practical ways to promote teachers' well-being.

The first group of articles addresses the relationship between burnout, the work environment, and individual characteristics. These results are important because they offer suggestions on how to foster resilience among teachers, also in the light of the management of schools and educational systems.

More specifically, a study carried out in Czech Elementary Schools (Smetackova et al.) confirmed the close relationship between burnout and low self-efficacy, poor use of positive coping strategies, and low social support from colleagues and principals. In addition, teachers with developing or developed burnout syndrome feel more often unsatisfied about their profession. These results are interesting given that the data was collected in a post-communist country like Czechia, since the knowledge about this country is still underdeveloped and, as the authors point out, every educational system, with its specific values

Edited and reviewed by: Massimiliano Barattucci, University of eCampus, Italy

> \*Correspondence: Paula Benevene benevene@lumsa.it

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 21 January 2020 Accepted: 12 May 2020 Published: 12 August 2020

#### Citation:

Benevene P, De Stasio S and Fiorilli C (2020) Editorial: Well-Being of School Teachers in Their Work Environment. Front. Psychol. 11:1239. doi: 10.3389/fpsyg.2020.01239

**6**

and organizational structure, introduces different challenges, addressing different ways of tackling them.

These findings are in line with another article, presenting a comparison between two groups of primary school teachers, one Italian (Naples) and the other Swiss (Cantone Ticino) (Parrello et al.). In fact, from the authors' findings it is possible to hypothesize that different professional cultures impact differently on teachers' burnout, while macro factors of the work context have a limited impact. More in general, their study aimed at understanding if and to what extent the sociocultural and economic differences of the two contexts, as well as dispositional, relational, and organizational variables, impact teacher burnout differently.

The research carried out among a group of Finnish school teachers adopted a person-oriented approach, which offers the possibility of identifying how teachers may endorse simultaneously contradictory measures of work well-being, such as burnout and engagement (Salmela-Aro et al.). In fact, the authors examined different kinds of teachers' profiles, based on work burnout and engagement. This study revealed for the first time a profile of teachers being engaged but at the same time at risk of exhaustion and of feeling of inadequacy. Identifying risk factors for teacher burnout is important but also understanding protective factors is necessary.

The study carried out by Fiorilli et al. investigates the relation between teachers' levels of burnout on the one hand and their emotional intelligence on the other, as well as the internal and external social support received. Internal support comes from within the work setting itself (i.e., support from colleagues, supervisors, school principals), while external support refers to the teachers' private life (i.e., support from friends, family members, partners). Findings suggest that personal competence in the emotional domain may significantly contribute in teachers' well-being. More than that, emotional intelligence has proved to be susceptible of improvement through training programs, thus preventing teacher burnout.

Again, in terms of action that can be taken in order to reduce and prevent burnout, the study from Chirico et al. offers an interesting contribution. This study deals with an experience conducted in a group of teachers of a religious institute, in which prayer was used as a technique to prevent burnout. The study begins from a fact that techniques involving higher mental functions, such as transcendental meditation, have been used in stress and burnout prevention programs. Thus, praying could be effective, no less than meditation and other spiritual or mind–body techniques, in contrasting the negative effects of occupational stress and preventing burnout among teachers and, possibly, other human service professionals (Ballantyne and Retell).

The article written by Ballantyne and Retell presents a paper on praxis shock. More precisely, the authors investigate how teachers' self-reported levels of burnout and self-efficacy, wellbeing, and praxis shock vary as a function of time in the profession and how the constructs of well-being, burnout, and self-efficacy relate to teachers' experiences of praxis shock. It has to be noted that no previous work has approached the experiences of teachers' praxis shock as a way to investigate the interrelationships that might exist between self-efficacy, wellbeing, and burnout. Their paper reveals that praxis shock may occur at multiple points in a music teachers' career and that it is linked with reported burnout, well-being, and self-efficacy.

A second group of articles addresses the issue of teachers' subjective happiness, and its relation with factors of health and well-being among teachers, such as self-esteem and positive emotions.

The study of Benevene et al. deepens the understanding of the effects of happiness and self-esteem, as dispositional traits, on the health of teachers, as well as of the role played by the working environment in generating positive affection, thus mediating between the dispositional traits and teachers' health. The study is carried out in a group of full-time in-service teachers from Italy, and its findings show that self-esteem endorses healthrelated behaviors. Moreover, when teachers acknowledge their workplace as a context in which they feel happy, the impact of dispositional happiness and self-esteem on health conditions is higher.

The article by De Stasio et al. deals with kindergarten teachers, which are still poorly addressed in the literature on teachers' well-being. This study highlights that subjective happiness and compassion are linked. In addition, subjective happiness and compassion are related with work engagement, through the mediation of proactive strategies, that is, self-regulation and coregulation, meaning, respectively, the ability to identify and use resources for coping with stressors and the ability to seek and receive social support from colleagues.

So far, the relevance of offering training opportunities to teachers, following the mainstream of positive psychology (Rahm and Heise), has presented the results of a successful training aimed at fostering subjective well-being among a group of German teachers. Some of the main issues addressed in the course of the training intervention include the consequences of both positive and negative emotions, emotion regulation, time management, and gratitude. The analysis of the results of the experimental group, compared with the control group, confirmed the efficacy of the intervention.

A third group of articles collected in this special issue deals with the impact of the school context on factors affecting teachers' well-being. This is by far the largest group of articles, showing the increasing interest in the school dynamics as well as in the interaction of teachers with their workplace environment.

The article by De Carlo et al. examines the relationship between contextual work-related factors on the one hand, in terms of job demands and job resources, and work–family conflicts on the other. Building on the Job Demands-Resources (JD-R) model, authors hypothesize that job demands (that is, qualitative and quantitative workload) are positively associated with work–family conflicts among teachers. Moreover, in line with the buffer hypothesis of the JD-R, the authors explore how job resources, in terms of support from supervisors, job autonomy, and participation in decision making, have an impact on work–family conflicts.

The article by Addimando presents a study carried out in a group of in-service primary and junior schools in Switzerland, and it shows how positive working conditions leading to teachers'

well-being are associated with more effective teaching practices. The author explores how teachers who perceive a supportive and satisfying working environment (both in terms of internal and external resources) are more likely to be engaged in their activities and how this in turn leads to a more heterogeneous array of teaching practices with their students.

The paper from Parrello et al. presents a research action carried out in Italy, in 12 suburban secondary schools of six Italian cities. Teachers were involved in a Teacher Participatory Action Research (T-PAR) called the "Crossing Educational Boundaries" project. The article shows how T-PAR can be a powerful form of professional development for teachers: it actively involves them, improving their awareness of themselves, of their emotions and actions, and of their empowerment. Actively engaging the teachers and working with them to plan new methods to reduce the rates of educational wastage, in turn, is a way to reinforce their resilience and help them to give purpose to their jobs, thus improving their well-being at school.

De Cordova et al. present a paper on teachers who are exposed to violent behaviors on the part of students and/or their parents. Their study addresses an increasing phenomenon faced by teachers and represents one of the most serious workrelated stress factors affecting the teaching profession. This article deals with the capacity of teachers to deal with violence and to develop a more resilient mindset. In fact, teachers can experience occupational well-being even if they are subjected to aggressive behaviors: supportive leadership and good relationships with colleagues are valuable resources for fostering positive emotions and teachers' flourishing.

The paper by Barbieri et al. investigates the relationship between Italian teachers' well-being, socio-demographic characteristics and professional background. Using data from the 2015 wave of the Program for International Student Assessment (PISA), the study deepens the knowledge on how teachers' positive perception of the working environment, in terms of availability of adequate human and physical resources, and professional development opportunities, may develop a substantial level of well-being at work, and how these factors are related to teachers' job satisfaction. Moreover, the article addresses the key role of transformational leadership in promoting teacher's well-being.

Dal Corso et al. investigate the effects of perceived performance appraisal justice on teachers' well-being. Well-being factors that were considered in the study are job performance, job satisfaction, and life satisfaction. Performance management is a key factor to enhance professional development and improve teaching quality. This process is successful only if teachers perceive it as fair, clear, and effective, namely, if it is satisfying. The study contributes to better understanding the performance management process in educational settings. Furthermore, it highlights the importance of efficacy of the performance management, which is essential not only to improve individual well-being but also to enhance teaching quality.

The article presented by Cigala et al. presents the results of a training intervention on early childhood teachers. The issue addressed in the training is the management of a multiage classroom, perceived by the teachers themselves as making teaching and learning very difficult and ineffective. The study shows that it is possible to change the teachers' perception of their own work through training intervention and to reach positive results in terms of three specific dimensions of well-being: a greater sense of belonging to the group of colleagues, a greater sense of self-efficacy, and an idea of themselves as active and meaningful participants.

The paper by Wang and Li deals with technostress among a group of teacher of high education institutions (HEI) in China. More specifically, the authors investigate the relationships between multidimensional technostress and job performance. HEI requirements related to the use of Information and Communication Technologies (ICT) and the suitability of ICT for teachers' work are found to be critical factors affecting job performance. In addition, the study develops a comparison among teachers from different grade levels, analyzing if and to what extent HEIs' management related to ICT use tend to generate a different level of technostress among different teaching levels.

Also, the paper by Kuusimäki et al. deals with technology and teachers' well-being. Their article addresses the issue of digital communication (DC) between teachers and parents in a group of Finnish teachers. In fact, most parent–teacher communication nowadays takes place on digital platforms, but not much is known about the specific role of digital communication (DC) in building parent–teacher partnerships. This paper highlights how DC can be a strong, positive factor of promoting and supporting not only the parent–teacher partnership but also teachers' well-being.

The fourth group of articles addresses how individual factors may affect teachers' well-being.

Troesch and Bauer compared first-career teachers (FCT) with second-career teachers (SCT), that is, respectively, teachers who have no other previous professional experience but teaching and teachers who became teachers after having had a previous, different professional career. Their study explores whether FCTs and SCTs differ in how they feel challenged by professional demands associated with teaching. Their study suggests that SCTs' background as career switchers might be less important for coping with specific professional demands than the existent research literature suggest; indeed, they feel nearly as challenged when starting to teach as FCTs do.

The article by Barni et al. explores the relationship between teachers' values (conservation, openness to change, selftranscendence, and self-enhancement) and their self-efficacy, assuming as a starting point the Schwartz's theory of human values. In particular, their study aimed at analyzing the extent to which these relations are moderated by teachers' controlled and autonomous motivations for teaching. Interestingly, the study suggests that the relationships between openness to change and self-efficacy on the one hand, and self-transcendence and selfefficacy on the other, varied depending on teachers' motivations. These relations were stronger when teachers perceived less external pressure and felt to be self-determined toward teaching.

The article by Buonomo et al. deals with teachers' emotions. The authors investigate both the positive and negative emotions of teachers toward, respectively, their students and their professional role. More specifically, they explored how teachers' emotions toward students would explain their self-efficacy. This paper suggests that teachers' emotions toward professional roles may generate a broader emotional context that could influence the predictive effects of their emotions toward students.

Wong presents a study carried out in the Chinese preschool setting, observing teachers' personality and their teaching practice. More specifically, the author examines how preschool teachers' personalities are related to their beliefs of developmentally appropriate practices (DAP) and thus to their teaching practices. The paper suggests that there is a combined effect of teachers' personality and work experience in adopting DAP. Moreover, extraverted and intuitive teachers tend to show higher scores in DAP beliefs and practice than introverted teachers do.

Finally, the paper by Lizana et al. deals with teachers from a quite under-represented country in the literature on teachers' well-being: Chile. The authors carried out their study among a group of teachers working in rural schools of the Valparaíso Region, investigating the relationship between their perceived quality of life and their general health condition. Their paper suggests that resources must be made available to detect early mental and chronic health conditions of rural teachers.

A final group of articles deals with instruments useful to addressing and investigating teachers' well-being, through the validation of diagnostic instruments.

The study by Reina et al. provides evidence of the validity and reliability of the Spanish translation of presents of the Validation of a Physical Education Teachers' Self-Efficacy Instrument Toward Inclusion of Students With Disabilities (Block and Weatherford, 2013). The scale has also three subscales on intellectual, physical, and visual disabilities. The value of this work lies in the fact that teachers who receive prior training on working with disabled children would be more prepared to address diversity in the classroom, which is in turn related to their perception of self-efficacy.

Chalghaf et al. present another paper devoted to Teacher of Physical Education. More specifically, their study aims to develop a preliminary "Validation of the Teacher of Physical Education Burnout Inventory" (TPEBI) in the Arabic language, in Tunisia. Presently there is no reliable psychometric tool in the Arabic language that can be used to measure the

#### burnout level among sports and physical education teachers. The article by Chalghaf et al. thus deals with the translation of Maslach's Burnout Inventory instrument (according to its three-dimensional theoretical model: emotional exhaustion, depersonalization, and low personal accomplishment) in Arabic.

A second paper by Chalghaf et al. aims to validate the "Academic Flow Scale" (Flow 4D 16) in the Arabic language, again in Tunisia, and to test its factor structure, in terms of internal consistency/reliability, predictive validity, and sensitivity. The flow occurs when individuals are engaged in a specific activity with clear goals and high commitment, facing challenges in proportion to their skills, fully mobilizing their competencies, and dedicating their attention to the task.

A further paper by Chalghaf et al. deals with job satisfaction, as a relevant factor of well-being. The paper aims to trans-culturally adapt and validate "Teacher Job Satisfaction Scale" created by Pepe et al. (2017). This instrument has three subscales: satisfaction with colleagues, satisfaction with parents, and satisfaction with students. The authors thus developed the "Teacher of Physical Education Job Satisfaction Inventory" (TPEJSI) in the Arabic language. This scale was administered to a Tunisian population of sports and physical education teachers and analyzed according to Pepe's theoretical model.

The paper by Converso et al. develops a tool to offer support schools in making self-assessment evaluations of the internal organizational climate and teacher morale (TM). More specifically, the article deals with the Italian version of the School Organizational Health Questionnaire (SOHQ), developed by Hart et al. (2000). The original version of the SOHQ was translated from English into Italian, and it consisted of 57 Likert-type items grouped in 12 subdimensions: Morale, Appraisal & Recognition, Curriculum Coordination, Effective Discipline Policy, Excessive Work Demands, Goal Congruence, Participative Decision Making, Professional Growth, Professional Interaction, Role Clarity, Student Orientation, and Support Leadership.

# AUTHOR CONTRIBUTIONS

All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication.

# REFERENCES


**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2020 Benevene, De Stasio and Fiorilli. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Development and Preliminary Validation of the "Teacher of Physical Education Burnout Inventory" (TPEBI) in Arabic Language: Insights for Sports and Occupational Psychology

Nasr Chalghaf1,2,3,4, Noomen Guelmami1,5, Maamer Slimani1,2, Giovanni Del Puente<sup>2</sup> , Tania Simona Re6,7,8, Riccardo Zerbetto<sup>8</sup> , Juan José Maldonado Briegas<sup>7</sup> , Ottavia Guglielmi<sup>2</sup> , Sergio Garbarino<sup>2</sup> , Fairouz Azaiez1,2,3,4 and Nicola Luigi Bragazzi1,2,3,4,5,6,7,8 \*

<sup>1</sup> Postgraduate School of Public Health, Department of Health Sciences (DISSAL), University of Genoa, Genoa, Italy, <sup>2</sup> Department of Neuroscience, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health (DINOGMI), University of Genoa, Genoa, Italy, <sup>3</sup> Higher Institute of Sport and Physical Education, University of Sfax, Sfax, Tunisia, <sup>4</sup> Studies Group of Development and Social Environment, Faculty of Letters and Human Sciences of Sfax, University of Sfax, Sfax, Tunisia, <sup>5</sup> Higher Institute of Sport and Physical Education of Kef, Kef, Tunisia, <sup>6</sup> UNESCO Chair "Health Anthropology, Biosphere and Healing Systems", University of Genoa, Genoa, Italy, <sup>7</sup> Department of Psychology and Sociology of Education, University of Extremadura, Badajoz, Spain, <sup>8</sup> Centro Studi Terapia della Gestalt, Milan, Italy

Background: Burnout is an inappropriate response to chronic work stress, leading to emotional exhaustion (EE), depersonalization (D), and low personal accomplishment (PA). Burnout can affect workers in the helping professions. To quantitatively assess the burnout level among teachers, Maslach has adapted the "Maslach Burnout Inventory" (MBI) to the educational environment (the so-called MBI Educators Survey version or MBI-ES). Among teachers, sports and physical education teachers may suffer from burnout due to high workload.

Aims: No reliable psychometric tool in Arabic language exists that can be used to measure the burnout level among sports and physical education teachers. The objective of the present study was to develop a burnout measurement scale according to the Maslach's three-dimensional theoretical model for physical education teachers in Tunisia and to test its factor structure, in terms of internal consistency/reliability, predictive validity, and sensitivity.

Methods: A total of 525 Tunisian teachers teaching in secondary schools from different Tunisian governorates volunteered to participate in this study. The sample comprised of 285 males (54.3%) and of 240 females (45.7%). More in detail, 327 were teachers of primary school of physical education (62.3%) and 198 teachers of secondary

#### Edited by:

Paula Benevene, Libera Università Maria SS. Assunta, Italy

#### Reviewed by:

Alessandro De Carlo, Giustino Fortunato University, Italy Alessandra Falco, University of Padova, Italy

#### \*Correspondence:

Nicola Luigi Bragazzi robertobragazzi@gmail.com

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 22 December 2018 Accepted: 15 February 2019 Published: 09 April 2019

#### Citation:

Chalghaf N, Guelmami N, Slimani M, Del Puente G, Re TS, Zerbetto R, Maldonado Briegas JJ, Guglielmi O, Garbarino S, Azaiez F and Bragazzi NL (2019) Development and Preliminary Validation of the "Teacher of Physical Education Burnout Inventory" (TPEBI) in Arabic Language: Insights for Sports and Occupational Psychology. Front. Psychol. 10:456. doi: 10.3389/fpsyg.2019.00456

**10**

school (37.7%). Teachers were administered both the ad hoc developed "Teacher of Physical Education Burnout Inventory" (TPEBI) and the MBI-ES. Both exploratory [principal component analysis (PCA)] and confirmatory factor analyses (CFAs) were performed.

Results: The Cronbach's alpha coefficients were excellent (0.93, 0.94, and 0.91 for EE, D, and PA, respectively). The correlation matrix indicated significant correlations between the TPEBI and MBI-ED dimensions. However, CFA fit indices were not completely satisfactory.

Conclusion: Given the good PCA factor loadings, the correlation matrix, the sensitivity analysis, and the excellent internal consistency, it can be concluded that the TPEBI is a reliable psychometric tool that can be used to quantitatively assess the burnout level among teachers of physical education in the Arabic-speaking world. However, considering the CFA fit indices, further modifications to fully support the model are warranted.

Keywords: development and validation of a questionnaire, Arabic language, sports psychology, occupational psychology, work and well-being, burnout, teachers

# BACKGROUND

Initially observed inside a detoxification clinics and introduced by Freudenberger (1974), the concept of burnout has attracted considerable attention in the recent decades. Burnout can be described as an inappropriate response to chronic work stress, which leads to emotional exhaustion (EE), depersonalization (D), and low personal accomplishment (PA). According to Maslach, burnout is a syndrome of EE, D, and loss of self-efficacy, which may affect those individuals working with and caring for other humans, as a response to the emerging chronic emotional burden (Cherniss, 1980; Maslach, 1982; Maslach and Jackson, 1985, 1986; Golembiewski et al., 1986; Maslach and Schaufeli, 1993; Maslach et al., 2001).

Burnout, as such, can affect workers in the helping professions, like nurses, doctors, social workers, and teachers. These workers may experience high emotional stress due to frequent and intensive interactions with others. For instance, in the field of education, many studies have noted that the teaching profession is becoming more and more stressful, demanding and challenging, and that teacher burnout is an emerging, international psychosocial concern (Koustelios, 2001; Tsigilis et al., 2004; Koustelios and Tsigilis, 2005; Montgomery and Rupp, 2005; Hakanen et al., 2006; Stoeber and Rennert, 2008; Chang, 2009; Skaalvik and Skaalvik, 2011; Liu and Onwuegbuzie, 2012).

According to a meta-analysis, which has pooled together 2,527 correlational effect sizes from 65 studies, teachers are particularly vulnerable to stress and may suffer from ineffective mediating coping mechanisms over a long period of time (Montgomery and Rupp, 2005). Teacher burnout may have devastating consequences both for the mental health of teachers and students and may affect and compromise the quality of education (De Stasio et al., 2017; Benevene et al., 2018). Moreover, the higher the level of teacher's breakdown, the more likely the consequence of stress and burnout in the teaching profession (Koustelios, 2001; Tsigilis et al., 2004; Koustelios and Tsigilis, 2005; Montgomery and Rupp, 2005; Hakanen et al., 2006; Stoeber and Rennert, 2008; Chang, 2009; Skaalvik and Skaalvik, 2011; Liu and Onwuegbuzie, 2012).

With respect to the general population, teachers tend to have a higher level of burnout and, in particular, of EE, which can negatively impact on the learning process of students (Montgomery and Rupp, 2005).

Physical education teachers are particularly exposed to burnout. Indeed, physical education classes take place mainly outside the classrooms, which causes more problems to maintain discipline than in the classroom and requires constant vigilance for ensuring students' safety (Tsigilis et al., 2011). In addition, physical education teachers are often obliged to teach in particular situations and conditions, dealing with students who may be particularly turbulent, noisy, and intractable (Brouwers et al., 2011). At the end of the day, as such, physical education teachers may be physically and mentally exhausted (Brouwers et al., 2011).

In order to quantitatively assess the burnout level among teachers, Maslach has adapted the "Maslach Burnout Inventory" (MBI) to the educational environment (the so-called MBI Educators Survey version or MBI-ES). Many researchers have utilized this tool to investigate the phenomenon of burnout in various educational settings (primary, intermediate and secondary schools) in different cultural contexts such as the United States (Boles et al., 2000), Canada (Byrne, 1991), Holland (Schaufeli et al., 1994), Greece (Kantas and Vassilaki, 1996; Antoniou et al., 2006),Cyprus (Kokkinos, 2006), and Sweden (Arvidsson et al., 2016). The results of these investigations seem to suggest that teachers working in European countries suffer from lower levels of burnout than their colleagues in the United States or Canada, putatively due to cultural, societal, and organizational factors. Van Horn et al. (1997) have managed to replicate such finding, showing that Dutch teachers had a lower level of EE and D when compared to their Canadian counterparts. This result is of crucial importance, when validating

the psychometric tool in another language and adapting to another culture/context.

Among teachers, sports and physical education teachers may suffer from burnout due to high workload. For instance, a study carried out in Pelotas, in the southern state of Rio Grande do Sul (Brazil), found that 60.6% of teachers experienced high EE, 22.3% high D, and 34.0% low PA. According to the scores, 8.5% of teachers suffered from burnout syndrome (Sinott et al., 2014).

# AIMS

To the best of our knowledge, there is no psychometric tool in Arabic language that can be used to measure the burnout level among sports and physical education teachers. Therefore, the objective of the present study was to develop burnout measurement scale according to the Maslach's threedimensional theoretical model for physical education teachers in Tunisia and to test its factor structure, in terms of internal consistency/reliability, predictive validity, and sensitivity.

# MATERIALS AND METHODS

# Psychometrics Instruments

#### Development of the TPEBI

The "Teacher of Physical Education Burnout Inventory" (TPEBI) was devised based on the three dimensions of the Maslach's theoretical model (EE, D and PA).

The questionnaire was developed through different steps and phases: namely, an initial broad and comprehensive literature review on the topic (including, for instance, Chang, 2009; Skaalvik and Skaalvik, 2011; Liu and Onwuegbuzie, 2012). At this step, we could identify the main aspects that characterize the dimensions of burnout. Moreover, we made efforts to integrate the specific characteristics of the population under study into the questions. When drafting the items of the questionnaire, we selected a vocabulary that was clearly comprehensible and unambiguous. In the subsequent step, we performed a search of similar questionnaires in other languages and tested in other countries/settings (such as Montgomery and Rupp, 2005; Skaalvik and Skaalvik, 2011; Liu and Onwuegbuzie, 2012). In the next step, a focus group was arranged with experts in the Arab language, in sports and physical education, and in human and applied sciences. Then, several corrections made it possible to reformulate the questions that appeared to be unclear. This allowed the improvement of the tool. Finally, a pilot study was carried out, testing the preliminary properties and the readability of the questionnaire. This test confirmed the validity of the tool.

#### The MBI-ES

The MBI-ES is a psychometric instrument originally developed by Maslach et al. (1996), explicitly designed to quantitatively assess the burnout level among teachers. The tool is a 22 item questionnaire divided into three dimensions: seven items measuring EE, six items dealing with D, and nine items exploring PA. The scores are obtained on a six-point Likert scale and each dimension is calculated by computing the sum of the items. Several studies investigating both the original version and its cross-cultural adapted versions have confirmed the psychometric robustness of the tool, in terms of internal consistency and factor structure. In particular, two studies have confirmed the validity and reliability of the MBI-ES. Factor analysis studies by Iwanicki and Schwab (1981) and Gold (1984) supported, indeed, the MBI-ES 3-factor structure. Regarding reliability, Iwanicki and Schwab (1981) have reported Cronbach's alpha estimates of 0.90 for EE, 0.76 for D, and 0.76 for PA, while Gold (1984) reported estimates of 0.88, 0.74, and 0.72, respectively.

## Ethics Statement

The study protocol of the present investigation received ethical clearance from the UNESCO Chair "Health Anthropology Biosphere and Healing Systems," University of Genoa, Genoa (Italy), the Higher Institute of Sport and Physical Education of Sfax, Sfax (Tunisia), the Faculty of Letters and Human Sciences of Sfax, Sfax (Tunisia), and the Higher Institute of Sport and Physical Education of Kef, Kef (Tunisia). The project was approved by the Ethical Committee of the University of Sfax, Sfax, Tunisia.

All participants to the present study provided written, informed consent. Teachers were extensively informed about the purposes and procedure of the study, and were advised that the results would be made available to them upon completion of the study only in aggregate form, with no possibility to trace back to the single teacher's scores, thus ensuring anonymity and preserving the privacy of each participant.

The present investigation was carried out in accordance with the ethical principles of the 1964 Helsinki declaration and its subsequent amendments.

#### Procedure

Teachers who agreed to participate in the study were instructed how to proceed and complete the survey procedures required by the present study. Following the agreement of the secondary school principals, copies of the TPEBI and of the MBI-ES were simultaneously distributed to teachers at their work sites in off-peak hours. The entire procedure of questionnaires administration took over 2 months. A proper time (approximately 30 min) was ensured to each participant in order to answer the questionnaire thoroughly.

# Statistical Analysis

#### Descriptive Analysis

Before commencing any statistical analysis, data were visually inspected for potential outliers. Normality of data distribution was checked using the Pearson-D'Agostino omnibus test. Questionnaires scores were also checked for skewness and kurtosis, computing the Mardia's multivariate skewness and kurtosis statistics.

#### Internal Consistency/Reliability

The internal consistency of the instrument was examined computing the Cronbach's alpha coefficient for all the three dimensions of the inventory. More in detail, in order to properly

interpret the alpha coefficient, the following rule of thumb was used (Nunnally, 1978; George and Mallery, 2011): the coefficient was deemed excellent if the estimate was >0.90, whereas it was judged good in the range 0.80–0.90, acceptable in the range 0.70– 0.80, questionable or adequate in the range 0.60–0.70, poor in the range 0.50–0.60, and unacceptable if <0.50.

#### Inferential Statistics – Sensitivity Analysis

The sensitivity of the instrument was tested by performing a univariate analysis of variance (ANOVA), examining the impact of teachers' grade, gender, age, and their interaction effects on the TPEBI three dimensions scores.

#### Predictive Validity

Predictive validity was tested computing the Pearson's correlation between the dimensions of the ad hoc devised psychometric instrument (the TPEBI) and those of the MBI-ES. More in detail, the strength of correlation was measured using the rule of thumb described by Hinkle et al. (2003): the correlation was deemed negligible with the r coefficient in the range 0.00–0.30, low with r in the range 0.30–0.50, moderate with r in the range 0.50–0.70, high with r in the range 0.70–0.90, and, finally, very high with r in the range 0.90–1.00.

#### Principal Component Analysis (PCA)

The factor structure was initially investigated by conducting a principal component analysis (PCA) and a varimax rotation with Kaiser Normalization. More in detail, varimax rotation was chosen in that this approach enables to minimize factor complexity while, at the same time, maximes the variance of factor loadings (Tabachnick and Fidell, 2013).

Before proceeding with the PCA, the Kaiser–Meyer–Olkin (KMO) measure was computed in order to assess the sampling adequacy. Ideally, the KMO should be greater than 0.60. Once verified the sampling adequacy, a PCA iterative strategy was adopted in the present investigation. Different runs were carried out. First, an exploratory/preliminary run was performed on the 24 items of the questionnaire without conducting any rotation, in order: (i) to check if this approach could be deemed an appropriate technique for the matrix by examining whether the correlations among items were satisfactory (that is to say, yielding values > 0.30) and (ii) to control for the factorability of the correlation matrix utilizing the Bartlett's test of sphericity. In cases of statistical significance, this test enables scholars to reject the null hypothesis (that is to say the correlations in the correlation matrix are zero and the matrix is an identity matrix).

The likely number of factors was determined by: (i) computing the number of factors with eigenvalues greater than 1 (Field, 2009; Tabachnick and Fidell, 2013) and (ii) visually inspecting the Cattell's scree-plot. After checking the factor loadings, items were deleted in cases of unsatisfactory loading (that is to say, values less than 0.45). Moreover, items were suppressed if their factor loading conflicted with a sound theoretical explanation (Field, 2009; Tabachnick and Fidell, 2013). In the present case, no items were deleted, since all were retained.

Different runs with varimax rotation were, therefore, carried out in an iterative fashion, as previously explained, until a satisfactory, clearly interpretable solution was finally obtained. Cases of cross-loading were interpreted according to salience and explained variance, with theoretical considerations also being taken into account (Field, 2009; Tabachnick and Fidell, 2013).

#### Confirmatory Factor Analysis (CFA)

Then the model was tested by confirmatory factor analysis (CFA). As suggested and recommended by many scholars, a wide range of fit indices was calculated and reported, namely: (i) discrepancy indices [including the chi-squared and the Steiger– Lind's root-mean-square error of approximation (RMSEA)], (ii) tests comparing the target model with the null model [like the Bentler–Bonett's normed fit index (NFI), the Bentler–Bonett's not normed fit index (NNFI), known also as the Tucker– Lewis' index (TLI), the Bentler's comparative fit index (CFI), and the James–Mulaik–Brett's parsimony goodness-of-fit index (PGFI)], and (iii) information theory goodness-of-fit measures [the Joreskog's goodness-of-fit index (GFI), and the Joreskog's adjusted GFI (AGFI)].Concerning the cut-off and threshold values for discrepancy indices, the p-value associated with the chi-squared test should exceed 0.05 (that is to say, it should not be statistically significant). Further, the chi-squared divided by the degrees of freedom (df) value, should ideally be less than 2.0. As far as the RMSEA is concerned, values higher than 0.10 indicate poor fitting models (Steiger, 2000). Concerning the cutoff and threshold values for tests that compare the target model with the null model, NFI should exceed 0.90 according to Byrne (1994) or 0.95 according to Schumacker and Lomax (2004). NNFI/TLI should be above 0.95 according to Hu and Bentler (1995). PGFI is derived from NFI, correcting and compensating for model parsimony. CFI should exceed 0.95 (Bentler, 1990; Hu and Bentler, 1999) or 0.90 according to other scholars. Finally, regarding the cut-off and threshold values for information theory goodness-of-fit measures, GFI value should be higher than 0.90 (Byrne, 1994).

#### Statistical Software

All statistical analyses were carried out using the commercial software "Statistical Package for the Social Sciences" (IBM SPSS software for Windows, version 21.0, IBM Corp., Armonk, NY, United States; released 2012) whereas the CFA was performed by utilizing the commercial software "Analysis of a moment structures" (Amos software for Windows, version 21.0, IBM, SPSS, Chicago, United States; Arbuckle, 2012a,b).

For all statistical analyses, figures with p-value less than 0.05 were considered statistically significant.

# RESULTS

#### Univariate and Multivariate Normality

Concerning univariate normality, scores for all the items of the TPEBI items had univariate normal distributions with acceptable values of skewness and kurtosis.

As far as multivariate normality is concerned, the Mardia coefficients showed evidence of multivariate non-normality in

the data (multivariate kurtosis 782.35, z = 23.69, p < 0.001, and multivariate skewness 61.85, z = 29.95 p < 0.001).

# Development of the Psychometric Instrument: The TPEBI

Based on the different steps and phases mentioned in Section "Materials and Methods," the ad hoc devised psychometric instrument is made up of 24 items (eight items for each dimension, EE, D, and PA), and the scores of the dimensions are obtained by averaging the items scores. The answers are coded on a seven-point Likert scale. The 24 items of the TPEBI are reported in **Table 1**.

# Descriptive Analysis

At the collection and analysis of the questionnaires, there were no missing data. A total of 525 Tunisian teachers teaching in secondary schools from different Tunisian governorates volunteered to participate in this study. The sample comprised of 285 males (54.3%) and of 240 females (45.7%). More in detail, 327 were teachers of primary school of physical education (62.3%) and 198 teachers of secondary school (37.7%). Based on age distribution, the subjects were categorized into four age groups: namely, (i) age < 39 years, n = 113 (21.5%), (ii) age between 39 and 44 years, n = 220 (41.9%), (iii) age between 44 and 49 years, n = 105 (20.0%), and (iv) age > 49 years, n = 87 (16.6%).

# Principal Component Analysis (PCA)

Based on the eigenvalues and the inspection of the Cattell's scree-plot, PCA revealed a three-factor structure, accounting for 63.68% (**Tables 2**, **3**). Factor loadings ranged from 0.73 to 0.89 for EE, from 0.76 to 0.82 for D, and from 0.69 to 0.84 for PA.

# Internal Consistency/Reliability

The Cronbach's alpha coefficients were 0.93, 0.94, and 0.91 for EE, D, and PA, respectively. As such, the internal consistency/reliability of the ad hoc devised psychometric tool was found to be excellent.

# Inferential Statistics – Sensitivity Analysis

The results of the ANOVA are reported in **Table 4**. Grade significantly impacted on all the three dimensions of the TPEBI. Gender and age significantly impacted on the PA dimension of the TPEBI. Concerning the interaction effects, no significant influence could be found except for the interaction grade × age, significantly impacting on the D dimension of the TPEBI.

# Predictive Validity

The correlation matrix showed a number of statistically significant correlations between the TPEBI and the MBI-ES dimensions. Correlations are shown in **Table 5**. More in detail, concerning the TPEBI, correlation between PA and EE dimensions yielded a value of 0.19 (negligible, even though significant at 0.001 level). EE dimension of the TPEBI correlated with the EE (r = 0.71, high correlation, significant at 0.001 level), with the D (r = 0.18, negligible correlation, even though significant at 0.001 level) and with the PA (r = 0.17, negligible correlation, even though significant at 0.001 level) dimensions of the MBI-ES. The D dimension of the TPEBI correlated with the EE (r = 0.15, negligible correlation, even though significant at 0.001 level), and with the D (r = 0.65, moderate correlation, significant at 0.001 level) dimensions of the MBI-ES. Finally, the PA dimension of the TPEBI correlated with the EE (r = 0.20, negligible correlation, even though significant at 0.001 level), and with the PA (r = 0.71, high correlation, significant at 0.001 level) dimensions of the MBI-ES.

# Confirmatory Factor Analysis (CFA)

Concerning the CFA indices, the chi-squared yielded a value of 1592.69 (df = 249, chi-squared/df = 6.40). GFI and AGFI resulted 0.74 and 0.69, respectively. PGFI was 0.61, whereas TLI 0. 85 and CFI 0.86. RMSEA was 0.1. The findings of the CFA are pictorially shown in **Figure 1**.

# DISCUSSION

The objective of the present study was to construct and to test the factor structure, internal consistency/reliability, predictive validity, and sensitivity of an ad hoc burnout measurement scale for the Arabic-speaking world, devised according to the Maslach's three-dimensional theoretical model. Once developed, the 24 item tool was validated in a representative sample of physical education teachers in Tunisia, using both PCA and CFA. PCA factor loadings were good and the CFA fit indices satisfactory. The internal consistency/reliability of the three dimensions was found to be excellent. Furthermore, the instrument showed a number of statistically significant correlations with an already validated tool, the MBI-ES for teachers. Moreover, it was able to differentiate the burnout level according to the grade of the physical education teachers at the level of all the three dimensions, and, even though only partially, also according to age and gender of the teachers.

Generally, our findings are consistent with the existing scholarly literature on the topic. For instance, Kokkinos (2006) tested the three-factor burnout model in a population of physical education teachers in Cyprus using both PCA and CFA techniques. The results proved the psychometric robustness and soundness of the instrument in terms of factor loadings and, to a less extent, of CFA fit indices. The internal consistency/reliability was satisfactory, with Cronbach's alpha coefficients of 0.85 for EE, 0.63 for D, and 0.79 for PA. Female teachers seemed to be more emotionally exhausted than their male counterparts. Primary school teachers were more affected by EE, but D was higher among secondary school teachers. Overall, the results suggested that the Greek MBI-ES represented a valid and reliable adaptation of the original instrument, which can be used with confidence for in field investigations aimed to measure burnout.

In a recent cross-sectional study by Spittle et al. (2015), burnout was investigated in a sample of 49 high school physical education teachers, aged 25–63 years, stratifying the analysis by age and gender. Authors found that burnout scores did

TABLE 1 | Items of the ad hoc devised psychometric tool to quantitatively assess the burnout level among the teachers of physical education in the Arabic-speaking world, the "Teacher Physical Education Burnout Inventory" (TPEBI).

Teacher Physical Education Burnout Inventory (TPEBI)


0: Never; 1: Few times a year or less; 2: Once a month or less; 3: Many times a month; 4: Once a week; 5: Many times a week; 6: Every day.

not significantly differ by gender, whereas age impacted on the PA dimension, with younger teachers reporting lower levels of PA, therefore indicating greater levels of burnout. No significant interaction between gender and age was found for the aspects of EE or PA, but the effect was, instead, vital for the D dimension; younger male teachers reported, indeed, higher scores (moderate level) for D than older male teachers (low level), while younger and older female teachers reported comparable scores. These results indicated that the most inexperienced teachers in physical education tended to suffer more from burnout, which seems to be particularly true for young male teachers.

In another comparative study of Tsigilis et al. (2011) between physical education teachers at primary and secondary schools, it

TABLE 2 | Descriptive statistics reporting the scores of the "Teacher of Physical Education Burnout Inventory" (TPEBI) for each dimension found performing the principal component analysis (PCA).


F (female); M (male); SD (standard deviation); TPS (teacher of primary school): TSS (teacher of secondary school). Scores are broken down by teachers' grade, gender, and age group.


TABLE 4 | Impact on grade, gender, age, and their interaction effects on the "Teacher of Physical Education Burnout Inventory" (TPEBI) dimension and total scores.


Exploratory factor analysis was performed with varimax rotation and Kaiser normalization.

Emotional exhaustion Depersonalization Personal achievement Grade 30.03∗∗ 5.29<sup>∗</sup> 12.87∗∗ Gender 0.95 1.98 5.50<sup>∗</sup> Age 1.36 2.04 5.89∗∗ Grade × gender 2.98 0.22 0.01 Grade × age 1.67 2.75<sup>∗</sup> 1.64 Gender × age 0.71 1.26 0.93 Grade × gender × age 0.41 1.11 1.61

∗ (significant at 0.5 level); ∗∗(significant at 0.01 level).

Variable F

was shown that physical education teachers working in primary schools had significantly higher scores of burnout, namely, EE, compared to their secondary school colleagues. In addition, the strength of association between the three components of burnout was greater among primary school physical educators than in secondary school ones (Tsigilis et al., 2011).

In another investigation, Arvidsson et al. (2016) examined the prevalence rate of burnout among Swedish teachers, using a cross-culturally adapted version of the MBI-ES. Results reported low values in all the three dimensions of the inventory and there was no association between gender and increasing levels of burnout among teachers.

Studies addressing the relationship between sociodemographic data and teacher burnout have consistently



∗ (significant at 0.05 level); ∗∗(significant at 0.01 level); ∗∗∗(significant at 0.001 level); D (depersonalization); EE (emotional exhaustion); PA (personal achievement).

shown that some baseline factors can predict low but statistically significant variance in, at least some, subscales/dimensions of the burnout measurement questionnaire. Generally, age has been shown to be a significant predictor of EE, with younger teachers tending to score higher than older teachers. Considering gender, male teachers have been found to tend to have a higher score than female teachers on the D scale.

This finding is consistent with research conducted among other helping professions (Maslach and Jackson, 1985). Teachers who work with secondary school students tend to have a lower level of PA/personal success than their counterparts in primary school. Secondary school teachers also feel more depersonalized toward students than primary or junior teachers (Maslach and Jackson, 1985).

Grade represents an important variable that can be used to predict burnout levels. Differentiating between primary and secondary school teachers is essential in the field of burnout. Indeed, the survey by Kantas and Vassilaki (1996) on Greeks teachers showed significant differences between secondary and primary teachers. Furthermore, Kantas and Vassilaki (1996) showed for secondary school physical education teachers significantly lower scores for EE and D and significantly higher scores for PA.

Finally, teacher burnout syndrome seems to fluctuate depending on the cultural context and the educational system in which it is studied. For example, previous research has shown that North American teachers are more likely to be vulnerable to burnout than the Europeans (Koustelios, 2001).

# Strengths, Limitations, and Future Prospects

The major strength of the present investigation lays in its uniqueness and usefulness. The validated instrument is, indeed, expected to be concretely useful for improving the health and well-being of Arab speaking teachers of physical education. A further strength is given by the methodological rigor of the research: being the development and validation of an instrument, it follows the rules which are nowadays considered a standard in this kind of studies. An in-depth analysis of the psychometric properties of the questionnaire, using both PCA and CFA, has been carried out.

Among the limitations, even though in general from a psychometric standpoint the instrument is sound, based on the CFA findings, some modifications could be still required in order to further fully support the factor structure. Future studies should be conducted on other samples to replicate the current results.

Furthermore, given that the Arabic speaking world is quite vast and culturally various and that the current validation analysis was performed only in Tunisia, further studies in other Arabic speaking countries are warranted in order to strengthen the findings.

#### CONCLUSION

Our study aimed to develop and to test the factor structure, internal consistency/reliability, predictive validity, and sensitivity of a burnout measurement scale. Given the good PCA factor

#### REFERENCES


loadings, the correlation matrix, the sensitivity analysis, and the excellent internal consistency, it can be concluded that the TPEBI is a reliable psychometric tool that can be used to quantitatively assess the burnout level among teachers of physical education in the Arabic-speaking world. However, considering the CFA fit indices, some modifications are still required and further studies with higher samples are warranted in order to fully support the factor model.

#### AUTHOR CONTRIBUTIONS

NC, FA, and NB conceived the experiment and drafted the manuscript. NC performed the experiment. NC, NG, FA, and NB collected and analyzed the data. SM, GDP, TR, MS, RZ, JMB, OG, and SG critically reviewed the draft. All authors read and approved the final version.


Liu, S., and Onwuegbuzie, A. J. (2012). Chinese teachers' work stress and their turnover intention. Int. J. Educ. Res. 53, 160–170. doi: 10.1016/j.ijer.2012.03.006

Maslach, C. (1982). Burnout, the Cost of Caring. Englewood Cliffs, NJ: Prentice Hall. Maslach, C., and Jackson, S. (1986). Maslach Burnout Inventory Manual. Palo Alto,



**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Chalghaf, Guelmami, Slimani, Del Puente, Re, Zerbetto, Maldonado Briegas, Guglielmi, Garbarino, Azaiez and Bragazzi. This is an openaccess article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# When Does Work Interfere With Teachers' Private Life? An Application of the Job Demands-Resources Model

Alessandro De Carlo<sup>1</sup> \*, Damiano Girardi<sup>2</sup> , Alessandra Falco<sup>2</sup> , Laura Dal Corso<sup>2</sup> and Annamaria Di Sipio<sup>2</sup>

<sup>1</sup> Giustino Fortunato University, Benevento, Italy, <sup>2</sup> FISPPA Section of Applied Psychology, University of Padova, Padua, Italy

The purpose of this study is to examine the relationship between contextual work-related factors on the one hand, in terms of job demands (i.e., risk factors) and job resources (i.e., protective factors), and work-family conflict (WFC) in teachers on the other. Building on the Job Demands-Resources (JD-R) model, we hypothesized that job demands, namely qualitative, and quantitative workload, are positively associated with WFC in teachers. Moreover, in line with the buffer hypothesis of the JD-R, we expected job resources, in terms of support from supervisor (SS), job autonomy (JA), and participation in decision making (PDM), to affect this association, which is expected to be stronger when job resources are low. The study was conducted in an Italian secondary school. Overall, 122 teachers completed a self-report questionnaire aimed at determining WFC, as well as job demands and resources. The hypothesized relationships were tested using moderated multiple regression. The results of this study largely support our predictions. First, both aspects of workload were positively associated with WFC. Secondly, job resources, including SS and PDM, buffered this association, which was stronger when resources were low. On the contrary, JA did not buffer the association between workload and WFC. Overall, the results of this study are consistent with the JD-R model and contribute to the understanding of work–family conflict among teachers. More specifically, our study suggests that teachers with high levels of job resources, namely SS and PDM, can effectively cope with job demands, in terms of both qualitative and quantitative workload, thus preventing negative consequences such as conflict between work and family domains. Interventions aimed at preventing WFC among teachers should encourage organizations to optimize the balance between job demands and resources, as well as the identification and training of the workers at risk of WFC.

Keywords: work-family conflict, teachers, workload, support, participation, moderation

# INTRODUCTION

Over the past decades, the teaching profession has undergone several changes, often described in the literature as intensification (Hargreaves, 1992; Ballet and Kelchtermans, 2009; Van Droogenbroeck et al., 2014). According to this perspective, teachers are increasingly exposed to external expectations and pressures (e.g., from supervisors, parents, and policy makers), which

Edited by:

Caterina Fiorilli, LUMSA University, Italy

#### Reviewed by:

Tiziana Ramaci, Kore University of Enna, Italy Michela Cortini, Università degli Studi G. d'Annunzio Chieti e Pescara, Italy

\*Correspondence: Alessandro De Carlo alessandro.a.decarlo@gmail.com

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 06 February 2019 Accepted: 29 April 2019 Published: 21 May 2019

#### Citation:

De Carlo A, Girardi D, Falco A, Dal Corso L and Di Sipio A (2019) When Does Work Interfere With Teachers' Private Life? An Application of the Job Demands-Resources Model. Front. Psychol. 10:1121. doi: 10.3389/fpsyg.2019.01121

**20**

result in higher workload, both teaching- and non-teachingrelated (e.g., administrative work), and less time for interactions with colleagues as well as for one's private life. This may lead to a chronic feeling of work overload, both at school and at home, to the loss of specific professional skills, and to work-related stress.

Not surprisingly, work-family conflict (WFC) seems to play a central role in the stress process among teachers. Indeed, several job characteristics, such as workload, job insecurity, emotional labor, emotional investment in students' behavioral problems, and demanding interactions with parents, are associated with WFC among teachers (Cinamon et al., 2007; Noor and Zainuddin, 2011; Ilies et al., 2015; Richter et al., 2015; see also Michel et al., 2011, for a meta-analysis across different occupations). WFC may in turn have negative consequences for both teachers and their students, in terms of teachers' job burnout, poorer psychological and physical health, reduced organizational citizenship behaviors and job satisfaction, as well as lower students' perceived teacher autonomy support and autonomous motivation (Bragger et al., 2005; Cinamon and Rich, 2010; Bell et al., 2012; Guglielmi et al., 2012; Gao et al., 2013; Haslam et al., 2013; Santisi et al., 2014; Benevene and Fiorilli, 2015; Fiorilli et al., 2015; Shen et al., 2015; Bélanger et al., 2016; see also Amstad et al., 2011, for a meta-analysis across different occupations).

Therefore, in order to prevent illness and promote well-being and job performance among teachers, it seems important to investigate those precursors that may foster or prevent WFC. In this study, in light of the theoretical framework provided by the Job Demands-Resources model (JD-R; Demerouti et al., 2001; Bakker and Demerouti, 2007, 2017), we examined whether job demands (i.e., risk factors) are positively associated with WFC among teachers in an Italian secondary school, and whether job resources (i.e., protective factors) can moderate such association (Bakker et al., 2011). Previous research has shown that teachers in Italian secondary schools experience less work life-balance than those employed at other educational stages (Magnano et al., 2014b).

According to Greenhaus and Beutell (1985), WFC may be defined as a type of interrole conflict in which role pressures arising from work and family domains are mutually incompatible to some degree. This definition is consistent with the role stress theory (Kahn et al., 1964), according to which the resources available to individuals, such as time and energy, are limited, and the fulfillment of multiple roles unavoidably leads to the depletion of these resources. Therefore, individuals have to allocate their limited resources over work and family domains, in order to minimize the strain inexorably associated with the management of multiple roles (Geurts and Demerouti, 2003). Greenhaus and Beutell (1985) identified three dimensions of WFC, namely time-based, strain-based, and behavior-based conflict. According to this classification, the time devoted to one role, the strain produced by one role, or the behaviors required by one role, respectively, make it difficult to meet the demands of the other role. The interference between work and family domains may occur in both directions. In this regard, Netemeyer et al. (1996) proposed a distinction between WFC and family work conflict (FWC). WFC occurs when the demands of one's job, the time dedicated to one's job, and the strain generated by one's job interfere with responsibilities related to the family. Conversely, FWC refers to the interference between family demands, time dedicated to family activities, and strain created by the family on the one hand, and work-related responsibilities on the other. In this study, we focused exclusively on WFC, because we are interested in the effects of work context, in terms of job demands and resources, on teachers' functioning in the family domain.

The JD-R (Demerouti et al., 2001; Bakker and Demerouti, 2007, 2017) is a flexible model of job stress and well-being (Schaufeli and Taris, 2014) that has recently been applied to the interface between work and private life (Bakker et al., 2011). According to the JD-R, job characteristics of different occupations may be classified either as job demands or job resources. On the one hand, job demands are those aspects of a job (physical, psychological, social, and organizational) that require sustained physical and/or psychological (cognitive and emotional) effort from the employee and are therefore associated with certain psychological and/or physiological costs. On the other hand, job resources are those aspects of a job (physical, psychological, social, and organizational) that are functional in achieving work goals, reducing job demands and the associated costs (psychological and/or physiological), or promoting personal growth, learning, and development (Demerouti et al., 2001; Bakker and Demerouti, 2007, 2017). In line with the JD-R, high job demands may increase WFC due to the depletion of personal resources (e.g., time, physical and emotional energy; Bakker et al., 2011). If an individual has few resources available at the end of the day due to energy depletion at work, he or she will be less likely to be involved in his or her family role at home, which may give rise to work–family conflict (Ilies et al., 2015). Furthermore, the buffer hypothesis of the JD-R claims that high job resources may offset the harmful impact of job demands on work family conflict (Bakker and Demerouti, 2007, 2017; Xanthopoulou et al., 2007). Accordingly, specific job designs in which teachers are burdened with high job demands, but lack adequate levels of job resources, are particularly likely to increase WFC (Bakker et al., 2011). The conceptual model is depicted in **Figure 1**.

That being said, there are different reasons why distinct job demands and job resources may influence WFC in teachers. More specifically, in this study two different facets of workload were considered as job demands, namely time pressure and problem solving. The former reflects quantitative workload, the amount of work to be done in a given time, whereas the latter refers to qualitative workload, which pertains to the difficulty or complexity of the job, for which the worker is not trained or does not have enough resources to deal with (Xie et al., 2008; Bowling and Kirkendall, 2012). We focused on these specific job demands because previous research has shown an association between said factors and WFC (Byron, 2005; Michel et al., 2011). Moreover, in line with the definition proposed by Greenhaus and Beutell (1985), we expect quantitative and qualitative workload to influence two relevant dimensions of the construct, namely the time- and strain-based ones (Geurts and Demerouti, 2003; van Hooff et al., 2005). Finally, these two facets of workload reflect important job demands for teachers

that are acknowledged in the literature, such as time pressure, heavy workload (e.g., the necessity to prepare for teaching in the evenings or weekends), the diversity of tasks required (e.g., teaching- and non-teaching-related workload), and bureaucracy (Mearns and Cain, 2003; Skaalvik and Skaalvik, 2010; Antoniou et al., 2013; Van Droogenbroeck et al., 2014).

With respect to job resources, in this study we focused on support from supervisor (SS), job autonomy (JA), and participation in decision making (PDM) for several reasons. First, at a more general level and in line with the Self-Determination Theory (Deci and Ryan, 2000; Deci et al., 2017), these job resources may promote the satisfaction of basic human needs, such as, respectively, the need of relatedness (i.e., feeling part of a group, loved, and cared for), autonomy (i.e., to experience a sense of psychological freedom and volition), and competence (i.e., feeling effective in the interaction with the environment; Van den Broeck et al., 2010). The satisfaction of these needs, in turn, may lead to motivation and work-engagement (Van den Broeck et al., 2008), which help workers to cope effectively with job demands, thus preventing negative consequences such as WFC and job burnout (Xanthopoulou et al., 2007; Bakker et al., 2011).

Furthermore, there are other specific reasons why these job resources may buffer the relationship between job demands, in terms of qualitative and quantitative workload, and WFC in teachers. JA is associated with more opportunities to cope with job demands, and it might be of crucial importance for teachers, who can develop their own strategies to deal with educational (e.g., selecting student goals and their own teaching methods), and administrative demands (e.g., paperwork, meetings, and accountability demands; Van Droogenbroeck et al., 2014; De Neve et al., 2015). With respect to SS, school leaders such as principals and vice principals play a central role in supporting teachers, providing them with emotional and instrumental support (e.g., encouraging teachers to collaborate, providing support for policy changes and during conflicts with students or parents, providing extra job resources or opportunity for professional development; Kossek et al., 2011; Honingh and Hooge, 2014; Bélanger et al., 2015). Moreover, several facets of organizational support are associated with higher job satisfaction and reduced psychological strain, which may help to prevent negative consequences for the individual such as WFC (Riggle et al., 2009; Cortini et al., 2016). Finally, encouraging PDM may help teachers to affect the organization of their work, providing them with the opportunity to influence the allocation of resources, to deal with external or internal pressures, or to reduce hindrances to job performance, therefore increasing their effectiveness, both in the educational and the school administration domain (Bakker et al., 2007; Sarafidou and Chatziioannidis, 2013). This is particularly relevant in shifting and unstable work environments, in which the collaboration of different players is useful to create more effective work practices (Scaratti et al., 2017).

Therefore, based on these arguments and given the assumptions of the JD-R, we hypothesized that job demands will be positively associated with WFC, and that job resources will affect this association, which will be particularly strong when JR are low.

> Hypothesis 1: problem solving (H1a) and time pressure (H1b), two facets of workload that reflect job demands, will be positively associated with WFC;

> Hypothesis 2: the association between problem solving and WFC (H2a), as well as the association between time pressure and WFC (H2b), will be moderated by job resources, in terms of SS, JA and PDM, so that job resources will attenuate the positive association between job demands and WFC.

# MATERIALS AND METHODS

# Participants and Procedure

The present study was conducted in an Italian secondary school as part of a work-related stress risk assessment. Participants were teachers, who were informed beforehand about the aims of the investigation and took part in the study on a voluntary basis. All participants gave their written informed consent before the administration of the questionnaire, in accordance with the Declaration of Helsinki. The study was carried out in accordance to the rules of AIP (Associazione Italiana di Psicologia – Italian Association of Psychology), according to which there was no need for previous ethics approval, since it would not deal with animals or vulnerable groups, or would involve risk for the well-being of participants, or use biomedical devices or invasive investigation tools. Our study did not need ethics approval, according to our national regulations as well as to the Ethics Committee

of the University of Padova. A self-report questionnaire aimed at determining job demands, job resources, and WFC was administered. Overall, 150 teachers completed the questionnaire. However, 28 participants had missing values in at least one of the variables considered in the study and were therefore excluded from subsequent analyses. Accordingly, the final sample comprised of 122 participants. Fifty-four participants were women and 67 men (one missing value). Most respondents were aged between 46 and 55 years (47.6%), 31.9% were younger than 46 years, and 20.5% were older than 55 years. With respect to work experience, the majority of them (52.5%) had been teaching at the school for 10 years or less (46.7% for more than 10 years; one missing value). Most participants were married (70.5%) and had a permanent contract (83.6%). Finally, with respect to parental status, 52.4% of participants had two or more children, 23.8% had one child and 23.8% had no children.

#### Measures

To determine the dimensions under investigation, the following self-report measures were administered. All the scales were taken from the Q-Bo test, an instrument standardized for the Italian context (De Carlo et al., 2008). The six-point response scale ranged from 1 (strongly disagree) to 6 (strongly agree).

> Job demands, in terms of workload, were determined using ten items, designed to detect problem solving and time pressure (i.e., qualitative and quantitative workload). Examples of items were "My job requires me to constantly solve new problems" (problem solving, five items) and "I have to work very fast" (time pressure, five items). Cronbach's alpha was 0.85 for problem solving and 0.87 for time pressure.

> Job resources were assessed using fifteen items, aimed at determining SS (five items; e.g., "My supervisor values the work I do"), JA (seven items; e.g., "I can organize my work autonomously"), and PDM (three items; e.g., "Teachers are involved in making important decisions and the definition of work goals"). Cronbach's alpha was 0.86 for SS, 0.92 for JA, and 0.63 for PDM.

> Work-family conflict was determined using five items (e.g., "I devote too little time to my family because of my job"). Cronbach's alpha was 0.89.

#### Data Analysis

The hypothesized relationships were tested using moderated multiple regressions analyses following the procedure outlined by Aiken and West (1991); see also Cohen et al., 2003). WFC was the dependent variable, whereas job demands (i.e., problem solving and time pressure) and job resources (i.e., SS, JA, and PDM) were the independent and the moderating variables, respectively. The scores of each job demand and job resource were centered, and then the cross-products of centered variables were computed.

Overall, six different models were estimated. In Model 1 (M1), the centered scores of problem solving (JD) and SS (JR), as well as the interaction term, were entered in the regression model. In Model 2 (M2) and Model 3 (M3) JA (M2) or PDM (M3) were the job resources, respectively, whereas problem solving was the job demand. Model 4 (M4), Model 5 (M5), and Model 6 (M6) were similar, except that time pressure was the job demand. These models (M1–M6) were also estimated omitting the respective interaction term, to assess the additional variance explained by each of them.

To interpret the nature of the moderating effect, significant interactions were presented graphically, following the procedure outlined by Aiken and West (1991). Finally, if a significant interaction was found, then a simple slope analysis was conducted, to determine whether job demands are associated with WFC at high (+1SD) and low (-1SD) levels of job resources (Aiken and West, 1991). The analyses were performed using R version 3.5.1 (R Core Team, 2018).

# RESULTS

Descriptive statistics and correlations between study variables are reported in **Table 1**. Interestingly, both time pressure (r<sup>120</sup> = 0.51, p < 0.001) and problem solving (r<sup>120</sup> = 0.40, p < 0.001), the two facets of workload that reflect job demands, were positively associated with WFC. Conversely, there was a negative association between WFC and job resources, in terms of SS (r<sup>120</sup> = −0.36, p < 0.001), JA (r<sup>120</sup> = −0.34, p < 0.001), and PDM (r<sup>120</sup> = −0.25, p < 0.01).

The results of the regression analyses are presented in **Table 2** (M1–M3) and **Table 3** (M4–M6). Overall, problem solving (M1– M3) and time pressure (M4–M6) were positively associated with WFC, and therefore H1a and H1b are supported. The interaction between problem solving and SS (M1) accounted for an additional 5% of the variance in WFC, Fchange(1, 118) = 8.13, p < 0.01. The same also occurred for the interaction between problem solving and PDM (M3), which accounted for an additional 3.8% of the variance in WFC, Fchange(1, 118) = 5.64, p < 0.05, whereas the interaction between problem solving and JA (M2) was not significant.

To interpret the nature of the moderating effect, these interactions were presented graphically. The association between problem solving and WFC was stronger for individuals with low levels of SS (**Figure 2**). The same pattern also occurred for PDM (**Figure 3**). The simple slope analysis showed that the relationship between problem solving and WFC was positive and significant when SS was low (b = 0.64, p < 0.001) or when PDM was low (b = 0.61, p < 0.001), but non-significant when SS or PDM were high. These two job resources, therefore, buffered the positive association between problem solving and WFC, whereas JA did not. Overall, H2a was partially supported.

Similarly, the interaction between time pressure and SS (M4) accounted for an additional 2.8% of the variance in WFC, Fchange(1, 118) = 4.86, p < 0.05, and the same also occurred for the interaction between time pressure and PDM (M6), which accounted for an additional 2.4% of the variance in WFC, Fchange(1, 118) = 4.01, p < 0.05. Again, the interaction between time pressure and JA (M5) was not significant. To interpret the nature of the moderating effect, these interactions were presented graphically. The association between time pressure and WFC was stronger for individuals with low levels of SS (**Figure 4**).

#### TABLE 1 | Means, Standard Deviations, and Correlations Between Study Variables (N = 122).


Cronbach's alpha values are displayed along the diagonal of the correlation matrix.

∗∗p < 0.01. ∗∗∗p < 0.001.

TABLE 2 | Results from Moderated Multiple Regression Analyses: Model 1, Model 2, and Model 3 (N = 122).


All variables were centered at their means.

<sup>∗</sup>p < 0.05. ∗∗p < 0.01. ∗∗∗p < 0.001.

TABLE 3 | Results from Moderated Multiple Regression Analyses: Model 4, Model 5, and Model 6 (N = 122).


All variables were centered at their means.

<sup>∗</sup>p < 0.05. ∗∗p < 0.01. ∗∗∗p < 0.001.

The same pattern also occurred for PDM (**Figure 5**). The simple slope analysis showed that the relationship between time pressure and WFC was positive and significant either when SS was high (b = 0.32, p < 0.01) or low (b = 0.64, p < 0.001). The same pattern of results also occurred in M6, given that the relationship between time pressure and WFC was positive and significant either when PDM was high (b = 0.35, p < 0.01) or low (b = 0.65, p < 0.001). Overall, these two job resources buffered the positive association between time pressure and WFC, whereas JA did not. Therefore, H2b was partially supported.

#### DISCUSSION

Over the past decades, teachers have been increasingly exposed to external expectations and pressures (e.g., from supervisors, parents, and policy makers). As a consequence, teachers may experience higher workload (both teaching- and non-teachingrelated), less time for interactions with colleagues, more conflict between work and family domains, and job burnout (Hargreaves, 1992; Ballet and Kelchtermans, 2009; Van Droogenbroeck et al., 2014). With the aim of preventing WFC and its consequences, the

present study examined contextual work-related factors that may influence WFC among teachers in an Italian secondary school.

Our hypotheses were based on the JD-R model (Demerouti et al., 2001; Bakker and Demerouti, 2007, 2017), a relevant model of job stress and well-being that has recently been applied to the interface between work and private life (Bakker et al., 2011). Building on the JD-R, our first hypothesis was that job demands (i.e., risk factors) are positively associated with WFC, whereas our second hypothesis stated that job resources (i.e., protective factors) affect this association, which is expected to be particularly strong when JR are low. The results of this study largely support our predictions. First, job demands, in terms of qualitative and quantitative workload, were positively associated with WFC. Secondly, job resources, including SS and PDM, buffered this association, which was stronger when resources were low.

Overall, we believe that this study provides a valuable contribution to the literature on the interface between work and private life among teachers. First, our results suggest that workload, in terms of both the amount of work to be done in a given time (i.e., quantitative workload) and the difficulty or complexity of the job (i.e., qualitative workload), may play a central role in the onset of WFC in teachers. From a theoretical standpoint, this is consistent with the idea that demands at work deplete teachers' personal resources (e.g., time, energy, and mood), which may lead to poorer outcomes in the family domain (e.g., reduced quality of care for family members), and WFC (Bakker et al., 2011; ten Brummelhuis and Bakker, 2012). Moreover, empirical research has shown workload to be positively associated with WFC (Grzywacz and Marks, 2000; Byron, 2005; Michel et al., 2011). Notably, Ilies et al. (2015) recently found in a sample of teachers that daily workload was associated with experiences of strain-based work–family conflict, and this relationship was mostly mediated by emotional fatigue. The authors argued that the teaching profession may drain individuals of the emotional resources that are needed at home to effectively take part in family life, thus resulting in strainbased WFC.

Secondly, our study showed that job resources, namely SS and PDM, may protect teachers against the negative consequences of workload, in terms of conflict between work and family domains. Accordingly, high workload may not necessarily result in WFC if teachers have adequate levels of job resources, which may help individuals to preserve or replenish personal resources such as time, energy, and mood (Bakker et al., 2011; ten Brummelhuis and Bakker, 2012). On a theoretical level, this is consistent with the buffer hypothesis of the JD-R, which claims that high job resources may offset the harmful impact of job demands on WFC (Bakker et al., 2011), job burnout (Bakker and Demerouti, 2007, 2017; Xanthopoulou et al., 2007), and work-related stress (Falco et al., 2018). Moreover, from an empirical standpoint, our findings are also consistent with the ones described by Bakker et al. (2011), who found that job resources, such as PDM and supervisory coaching, attenuated the adverse effect of job demands (i.e., work overload, emotional, and cognitive demands) on WFC among medical residents.

Interestingly, in our study JA did not buffer the association between workload and WFC. This is an intriguing finding, albeit not completely unexpected. Indeed, although JA plays a central role in models of work-related stress and motivation such as the job demand-control model (Karasek, 1979) and the job characteristics model (Hackman and Oldham, 1976), previous research has shown that JA does not attenuate the relationship between job demands (including workload) and WFC (Bakker et al., 2011; Falco et al., 2013a). A possible explanation is that, at least in certain circumstances (e.g., when confronted with tasks that can be performed at home, such as preparing for lessons and grading assignments), JA may contribute to blur the boundaries between work and private life (Blair-Loy, 2009). Another possible explanation is that teachers in our sample particularly benefit from other resources, such as SS and PDM, whenever confronted with their specific job demands, whereas the independence in scheduling the work and in determining the procedures to be used is less relevant to them. It is also possible that, compared to other job resources, teachers attribute less value to JA when coping with high demands at work, because autonomy is perceived as an intrinsic feature of their job. However, given the relative scarcity of studies on the possible moderating role of JA in the relationship between job demands and WFC among teachers, more research is warranted to generalize our findings.

Taken together, the findings of this study allow us to extend to the teaching profession the results of the seminal work by Bakker and colleagues, who claimed that "the JD-R model provides a fruitful framework for explaining which particular job designs facilitate, or instead prevent, work–home interference" (Bakker et al., 2011, p. 178). According to the JD-R, the specific combination of job demands and job resources should be considered, to understand which particular job designs may result in work–family conflict. In this regard, our study suggests that teachers with high levels of job resources, such as SS and PDM, can effectively cope with job demands, in terms of both qualitative and quantitative workload, thus preventing negative consequences such as WFC.

At a more general level, the results from this study are also in line with the Conservation of Resources (COR; Hobfoll, 1989, 2001) theory, which provides a theoretical framework to understand stress (acute and chronic) in various settings of people's life. The COR theory claims that individuals try to acquire, retain and protect resources, which may include conditions (e.g., tenure, marriage, or occupational status), personal characteristics (e.g., self-esteem), or energies (e.g., time, money; Grandey and Cropanzano, 1999; Geurts and Demerouti, 2003). Psychological stress occurs when individuals' resources are threatened with loss or actually lost, or when individuals do not gain adequate resources following a relevant investment of resources (Hobfoll, 1989, 2001). According to the COR theory perspective, the balancing of work and family roles may lead to a loss of resources and work family conflict (Grandey and Cropanzano, 1999; Richardson and Thompson, 2012). Consistently, our study showed that job demands, in terms of quantitative and qualitative workload, are associated with WFC in teachers.

Moreover, the COR theory claims that individuals have to invest resources to protect against resource loss (Hobfoll, 2001). With respect to the interface between work and home, this means that contextual demands in a domain (e.g., high demands at work) may result in an initial loss of resources (e.g., energy or time) with more immediate consequences in the other sphere (e.g., spousal conflict). To effectively cope with this situation and prevent further resource loss, individuals may invest additional resources, which might lead to a loss spiral over time (ten Brummelhuis and Bakker, 2012; Neto et al., 2016). In this perspective, individuals with greater resources are less vulnerable to resource loss than those with fewer resources (Hobfoll, 2001). With greater job resources available, conflict between work and family domains becomes less likely (Grandey and Cropanzano, 1999). Consistently, our study showed that greater resources at work may protect against the negative consequences of high job demands, in terms of WFC.

Interestingly, aging could play a role in this process. For example, teachers at the beginning of their career may experience higher levels of WFC, since they invest considerable amount of energy and personal resources to meet work goals and to deal with stressful work-related demands while, at the same time, they are also involved in parental duties (Antoniou et al., 2006). Conversely, older teachers may have lower levels of WFC, given that age is associated with greater job resources (e.g., tenure, status), coping resources, occupational experience, and organizational identification (Grandey and Cropanzano, 1999; Avanzi et al., 2012; Mauno et al., 2013). Future studies should investigate the possible moderating role of age in the relationship between job demands and resources on the one hand, and WFC and enrichment on the other.

Our study has some limitations. First, in this study we focused only on WFC, although it is possible that work and family domains may positively influence each other. For example, when resources are adequate to deal with high demands in one domain (e.g., at work), individuals may feel stimulated to learn and grow (e.g., in their job), and energies will be mobilized, rather than depleted (Geurts and Demerouti, 2003). This may in turn improve the quality of life in the other domain, such as family life (i.e., work-to-family enrichment; Greenhaus and Powell, 2006). Moreover, it should be noted that the crosssectional design of the study precludes conclusions about the causal direction of the observed relationships. Future studies should investigate the longitudinal relationships between job demands, resources, and work family conflict, given that past research suggested that these constructs may influence each other over time (Demerouti et al., 2004). Moreover, studies using within person designs could shed light on the dayto-day processes through which work influences the family domain (Ilies et al., 2015). It should be also mentioned that the constructs of interests were determined using the same measurement method (i.e., self-report questionnaires), and the observed relationships could be affected by common method bias (Podsakoff et al., 2012). Accordingly, future studies could adopt, for example, observer-rating of WFC (e.g., spouse rating;

Grandey et al., 2005), as well as biomarkers of stress that could be related to insufficient opportunities for recovery (Girardi et al., 2015a). Finally, future research should consider the possible role of additional job demands (e.g., emotional demands) and job resources (e.g., job security, support from colleagues, opportunities for professional development) that may be relevant for teachers (Xanthopoulou et al., 2007; Falco et al., 2008; Barbieri et al., 2014).

We believe that this study has relevant practical implications. More specifically, our results suggest that interventions should be aimed at preventing WFC, in order to avoid negative consequences for teachers and their students, such as job burnout, sickness absenteeism, and presenteeism, as well as reduced teachers' effectiveness (Amstad et al., 2011; Falco et al., 2013b; Shen et al., 2015). These interventions should target the organization as well as the individuals (i.e., primary and secondary prevention).

Regarding primary prevention, organizations should be encouraged to optimize the balance between job demands and job resources (following a top-down approach; van Wingerden et al., 2017). On the one hand, this can be achieved by reducing job demands whenever possible, especially the ones perceived as hindrances (e.g., interpersonal conflict, role ambiguity, which may result in work overload), or minimize their negative consequences. For example, with respect to the time-based dimension of WFC, work activities that can interfere with teachers' private lives, such as working overtime or working at home during the week-end, could be discouraged, whereas workers who reach their objectives operating within regular working hours could be rewarded. At the same time, to reduce strain-based WFC, activities that favor recovery should be promoted (Geurts and Sonnentag, 2006). For instance, teachers could learn simple physical exercises (e.g., stretching) or relaxation techniques to practice at home or at work, in specifically designated areas during work breaks (Taylor, 2005).

On the other hand, organizations should provide teachers with adequate levels of job resources, such as SS and PDM, in order to offset the negative consequences of job demands on WFC. Clearly, school leaders such as principals and vice principals play a central role in this process. For example, supervisors training programs could be directed at increasing the adoption of family supportive supervisor behaviors (Hammer et al., 2011). These behaviors refer to emotional support (e.g., listening and care for teachers' work-family demands), instrumental support (e.g., providing effective support for day-by day teachers' workfamily demands), role-modeling behaviors (e.g., actively showing how to balance work and family domains), and creative work-family management (e.g., restructuring job to encourage teachers' effectiveness in work and family domains). Moreover, school leaders could involve teachers in decision making (e.g., when new protocols are adopted or when resources are allocated), so that they can optimize their work processes. Furthermore, interventions could include case studies that adopt a situational approach for improving managerial competence in school leaders and teachers

(Ripamonti and Scaratti, 2012; Scaratti and Ivaldi, 2015). In this respect, the active participation of different subjects (e.g., principals and vice principals, teachers, clerical workers), including practitioners such as work and organizational psychologists or therapists, is essential to gain a deeper insight into the specific issues reported and develop effective organizational policies and practices (Cunliffe and Scaratti, 2017).

With respect to secondary prevention, interventions should be directed at the identification and training of the workers at risk of WFC. Again, school leaders play a key role because they are able to identify teachers who may need support in balancing work and family domains, as well as benefit from specific interventions. In this regard, specific training programs could help teachers to develop personal resources, such as selfesteem, self-efficacy, and courage (De Stasio et al., 2017; Farnese et al., 2017; Magnano et al., 2017; Benevene et al., 2018b,c) as well as skills needed to effectively cope with high workload and time pressure (e.g., time management, goal-setting; Richardson and Rothstein, 2008). This in turn may also improve students' personal resources, motivation, satisfaction, and performance (Magnano et al., 2014a,b; Ripamonti et al., 2018; Santisi et al., 2018). The aforementioned interventions should be directed specifically at individuals with high levels of personal demands (e.g., perfectionism; Falco et al., 2014), which may influence the effort they invest in their work (Bakker and Demerouti, 2017). For example, individuals with high levels of perfectionism could be trained to gain awareness about the dysfunctional role played by work-related irrational beliefs (Girardi et al., 2015b; Falco et al., 2017). Finally, interventions should be aimed at fostering job crafting, in which teachers are encouraged to proactively optimize their own job designs (bottom-up approach; Tims et al., 2012). For example, teachers could try to increase social resources (Benevene et al., 2018a), such as seeking social support from the school leader when confronted with high demands at work, or more structural resources, such as a more active PDM.

# REFERENCES


Overall, building on the seminal study conducted by Bakker et al. (2011) on a sample of medical residents, the aim of this paper was to extend the Job Demands–Resources model to WFC in teachers. Our results showed that job demands, in terms of qualitative and quantitative workload (i.e., problem solving and time pressure, respectively), were positively associated with WFC in teachers. Furthermore, in line with the buffer hypothesis of the JD-R, job resources, in terms of SS and PDM, modulated the association between job demands and WFC, which was stronger when resources were low. Therefore, in order to prevent WFC among teachers, job designs that optimize the balance between job demands and resources should be encouraged.

#### ETHICS STATEMENT

All participants gave their written informed consent before the administration of the questionnaire, in accordance with the Declaration of Helsinki. The study was carried out in accordance the rules of AIP (Associazione Italiana di Psicologia – Italian Association of Psychology), according to which there was no need for previous ethics approval, since it would not deal with animals or vulnerable groups, or would involve risk for the well-being of participants, or use biomedical devices, or invasive investigation tools. Our study did not need ethics approval, according to our national regulations as well as to the Ethics Committee of the University of Padova.

# AUTHOR CONTRIBUTIONS

ADC developed the research project, with the contribution of DG, AF, and LDC. DG carried out the data analysis, with the contribution of ADS. ADC reviewed the literature. AF and LDC developed the questionnaire. ADS developed the administration procedure and prepared the data set.



three-wave study. J. Vocat. Behav. 64, 131–149. doi: 10.1016/S0001-8791(03) 00030-7


spillover between work and family. J. Occup. Health Psychol. 5, 111–126. doi: 10.1037/1076-8998.5.1.111



**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 De Carlo, Girardi, Falco, Dal Corso and Di Sipio. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fpsyg-10-01645 July 11, 2019 Time: 17:36 # 1

# Teachers' Self-Efficacy: The Role of Personal Values and Motivations for Teaching

#### Daniela Barni<sup>1</sup> \*, Francesca Danioni<sup>2</sup> and Paula Benevene<sup>1</sup>

<sup>1</sup> Department of Human Sciences, LUMSA University, Rome, Italy, <sup>2</sup> Family Studies and Research University Centre, Catholic University of Milan, Milan, Italy

Teachers' personal values drive their goals and behaviors at school. Moreover, values can support subjective well-being and an individual sense of self-efficacy. Teachers' self-efficacy, namely teachers' beliefs in their ability to effectively handle the tasks, obligations, and challenges related to their professional activity, plays a key role in influencing important academic outcomes (e.g., students' achievement and motivation) and well-being in the working environment. Based on Schwartz's well-known theory of human values, this study sought to examine the relations between teachers' values (i.e., conservation, openness to change, self-transcendence, and self-enhancement) and their self-efficacy. In particular, it aimed at analyzing the extent to which these relations are moderated by teachers' controlled and autonomous motivations for teaching. Two hundred and twenty-seven Italian high school teachers (73.6% females; M = 44.77 years, SD = 10.56) were involved in the study and asked to complete a self-report questionnaire. Results showed that teachers' conservation values were positively associated to sense of self-efficacy regardless of the type and level of motivation for teaching. More interestingly, the relationships between openness to change and self-efficacy on the one hand, and self-transcendence and self-efficacy on the other, varied depending on teachers' motivations. These relations were stronger when teachers perceived less external pressure and felt to be self-determined toward teaching. Implications of these results for teachers' practices and well-being in their work environment and further developments of the study are discussed.

Keywords: teachers, self-efficacy, values, motivations for teaching, well-being

# INTRODUCTION

Teachers' self-efficacy has progressively gained an important role in school psychology research as a result of its implications for teaching effectiveness, instructional practices, and for students' academic achievement (Klassen et al., 2009; Klassen and Tze, 2014). Considerable research has shown that teachers with high levels of self-efficacy experience higher levels of job satisfaction, lower levels of job-related stress and face less difficulties in dealing with students' misbehaviors (Caprara et al., 2003). Thus, understanding the main antecedents of self-efficacy may have important payoffs in working for teachers' well-being and school effectiveness and improvement.

#### Edited by:

Tiziana Ramaci, Kore University of Enna, Italy

Reviewed by: Marinella Coco, University of Catania, Italy Giuseppe Santisi, University of Catania, Italy

> \*Correspondence: Daniela Barni d.barni@lumsa.it

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 11 April 2019 Accepted: 28 June 2019 Published: 12 July 2019

#### Citation:

Barni D, Danioni F and Benevene P (2019) Teachers' Self-Efficacy: The Role of Personal Values and Motivations for Teaching. Front. Psychol. 10:1645. doi: 10.3389/fpsyg.2019.01645 fpsyg-10-01645 July 11, 2019 Time: 17:36 # 2

The concept of self-efficacy derives from Bandura's socialcognitive theory of behavioral change (Bandura, 1977). It refers to a teacher's belief in his/her ability to successfully cope with tasks, obligations and challenges related to his/her professional role (e.g., didactical tasks, managing discipline problems in the class, etc.) (Caprara et al., 2006). This belief is determined by several factors, among which personality characteristics – in particular personality traits – which have led to a rise in the academic interest on the topic. For example, based on the five-factor model of personality (Costa and McCrae, 1992), Djigic et al. (2014) ´ found that teachers with higher levels of openness to experience and conscientiousness reported a stronger sense of efficacy. Sousa et al. (2012), in their study involving frontline service employees, showed that also personal values are significant predictors of workers' self-efficacy. Specifically, openness to change values (i.e., selfdirection, stimulation, and hedonism) and self-enhancement values (i.e., power and achievement) are both positively related to self-efficacy. These values are part of the well-known Schwartz's (1992) model of human values, which also includes conservation values (i.e., tradition, conformity, and security) and self-transcendence values (i.e., benevolence and universalism). According to Schwartz, personal values can be defined as trans-situational goals that vary in importance and serve as guiding principles in people's lives. Schwartz (1992, 2012) identified ten basic value types: power (social status, dominance over people and resources), achievement (personal success according to social standards), hedonism (pleasure or sensuous gratification), stimulation (excitement, challenge, and novelty), self-direction (independence of thought and action), benevolence (preserving and enhancing the welfare of people to whom one is close), universalism (understanding, tolerance, and concern for the welfare of all people and nature), tradition (respect and commitment to cultural or religious customs and ideas), conformity (restraint of actions and impulses that may harm others or violate social expectations), and security (safety and stability of society, relationships, and self). These ten values can be organized into a motivational two-dimensional system. The first dimension contrasts openness to change, which emphasizes change and independence, and conservation, where the emphasis is instead on self-restraint, preserving traditional practices, and safeguarding stability. The second dimension contrasts selfenhancement, where people prioritize their personal interests even at the expense of others, and self-transcendence, where people transcend their selfish concerns to promote the welfare of others. More recently, Schwartz (2012) has highlighted that self-transcendence and conservation values are characterized by a social focus, that is concern with outcomes for others or for established institutions, whereas self-enhancement and openness to change values by a personal focus, namely concern with outcome for self.

Despite their great influence on many attitudes and behaviors (Schwartz, 2015), also in the workplace (e.g., Koivula, 2008) and at school (e.g., Barni et al., 2018), the relationship between personal values and self-efficacy has been completely under investigated among teachers. This study intends to fill this gap. In particular, we analyzed in a sample of 227 teachers how a comprehensive set of personal values (i.e., conservation, openness to change, self-transcendence, and self-enhancement) is related to self-efficacy, and whether this relationship is moderated by controlled versus autonomous motivations for teaching. According to the Self-Determination Theory (Deci and Ryan, 1985), control-motivated teachers perform teaching activities because they wish to receive external rewards (e.g., school principal's approval) or to avoid feelings of guilt, whereas autonomously motivated teachers perform teaching activity because of the intrinsic value they attribute to it (Roth et al., 2007). Given that values to exert influence must be activated and the target behavior needs to be personally relevant to the individual (Bardi and Schwartz, 2003), it is likely that the relation between values and self-efficacy is stronger when teaching behavior is self-determined (i.e., low controlled motivations and high autonomous motivations). Indeed, as typically shown by studies dealing with the person-situation controversy, the stronger the perceived situational pressure to hold a value or to act in a particular way, the weaker should be the importance of stable individual differences such as personal values (e.g., Bardi and Schwartz, 2003; Barni et al., 2016; Danioni and Barni, 2017).

# METHOD

#### Participants and Procedure

Participants were 227 teachers (73.6% females) working in Italian high schools whose age ranged between 25 and 68 years (M = 44.77, SD = 10.56). On average, the teachers involved had been working in the schools for 7.93 years (SD = 7.80, range = 1–34). Most of the participants were tenured teachers (95.1%); 43.3% of them taught scientific and technical subjects, 41.0% taught humanistic subjects whereas the remaining 15.7% taught foreign languages. Participants were recruited with the collaboration of the schools where they worked and were informed about the main objectives of the study. The researcher directly contacted the school principal to obtain approval to carry out the research, who then introduced the researcher to the teachers. Teachers participated in the study on a voluntary basis. Overall, sixteen schools were involved.

Teachers who gave written informed consent were informed about the main objectives of the study and filled in an anonymous self-report questionnaire (response rate: 72.7%). The study was approved by the Ethics Committee of the Catholic University of Milan and followed the APA ethical guidelines of research. Data treatment follows the Italian legal and privacy restrictions (Italian Data Protection Code – Legislative Decree No. 196/2003) and the General Data Protection Regulation 2016/679.

#### Measures Personal Values

The Portrait Values Questionnaire (PVQ) (Schwartz et al., 2001; Italian validation by Capanna et al., 2005) was used to measure personal values. The PVQ is composed of 40 verbal portraits of a person and his/her objectives or aspirations, which indirectly reflect the importance of a value. For example, "Thinking up new ideas and being creative is important to him/her. He/She likes to do things in his/her own original way" describes a person for whom openness to change is important. Respondents' values are inferred from their self-reported similarity (from 1 = not like me at all to 6 = very much like me) to people described. We computed four mean indexes assessing the importance personally given to conservation (α = 0.86), openness to change (α = 0.79), self-transcendence (α = 0.80), and self-enhancement (α = 0.77).

#### Motivations for Teaching

fpsyg-10-01645 July 11, 2019 Time: 17:36 # 3

The Autonomous Motivations for Teaching Scale (Roth et al., 2007), translated into Italian and back-translated into English, was used to assess motivations for teaching. This scale includes 16 items on a 5-point Likert answer scale (from 1 = totally disagree to 5 = totally agree). Eight items measure teachers' controlled motivations for teaching (e.g., "When I devote time to individual talks with students, I do so because I want the parents to appreciate my knowledge and familiarity with their children;" α = 0.74) and the remaining eight items measure teachers' autonomous motivations for teaching (e.g., "When I devote time to individual talks with students, I do so because I like being in touch with children and adolescents;" α = 0.65).

#### Self-Efficacy

The Teachers' Self-efficacy Scale (Caprara, 2001) was used. The scale is composed of 12 items on a 7-point Likert answer scale (from 1 = totally disagree to 7 = totally agree). An example of an item is: "As a teacher, I am capable of getting recognition and appreciation from my students". We computed one mean index measuring teachers' beliefs in their ability to effectively handle various tasks, obligations, and challenges related to their professional role (α = 0.83).

# Data Analysis

#### Preliminary Analysis

We described the study variables in terms of means, standard deviations, and range. Associations between variables were measured by bivariate Pearson correlations.

#### Moderation of Controlled and Autonomous Motivations for Teaching on the Relation Between Teachers' Personal Values and Self-Efficacy

We conducted two multiple hierarchical regressions to test the moderation hypothesis. Teachers' personal values (i.e., conservation, openness to change, self-transcendence, and selfenhancement) were used as independent variables, testing separately the interaction effects with the two moderators (i.e., controlled motivations and autonomous motivations for teaching) for the outcome of teachers' self-efficacy.

Each of the two regression models, the first one with teachers' controlled motivations as moderator and the second one with autonomous motivations as moderator, was tested in the following steps. In step 1 predictors (i.e., teachers' personal values) and moderator variable (i.e., teachers' controlled motivations or autonomous motivations) were entered in the model to test their direct relations with teachers' self-efficacy. In step 2 the interaction term between predictor and moderator variable was entered to the model to test the statistically significance of the interaction effect. The independent variables were centered on their means before computing the interaction terms to minimize multicollinearity and for easier interpretation of model coefficients (Aiken and West, 1991). Lastly, simple slope analysis was performed to probe any significant interaction effect. We used SPSS 24.0 to conduct all the analyses.

# RESULTS

Descriptive statistics of the study variables and their intercorrelations are reported in **Table 1**. Teachers gave the greatest importance to self-transcendence values, followed by conservation, openness to change, and self-enhancement. They scored higher on autonomous motivations for teaching than on controlled motivations, t(225) = 25.22, p < 0.001, and on average they reported a high level of self-efficacy.

Teachers' personal values were all significantly correlated to each other, with the only exception of self-transcendence and selfenhancement values. Openness to change, self-transcendence, and self-enhancement were positively associated with teachers' autonomous motivations for teaching. Self-enhancement


TABLE 1 | Pearson correlation coefficients, means, standard deviations (SD) and ranges of teachers' values, motivations for teaching, and self-efficacy (N = 227).

∗∗p < 0.01 and <sup>∗</sup>p < 0.05.

fpsyg-10-01645 July 11, 2019 Time: 17:36 # 4

and self-transcendence were also associated, respectively positively and negatively, with teachers' controlled motivations. Finally, teachers' self-efficacy was significantly correlated to conservation, openness to change and self-transcendence, but not to self-enhancement values or motivations for teaching (**Table 1**).

**Table 2** shows the results of moderated regression analyses. Both models, the first one (Model 1) with controlled motivations as moderator and the second (Model 2) with autonomous motivations as moderator, were statistically significant [Model 1: F(9,215) = 4.063, p < 0.001; Model 2: F(9,215) = 3.391, p < 0.01]. They both accounted for significant amount of variance in teachers' self-efficacy (Model 1: R <sup>2</sup> = 0.145, p < 0.001; Model 2: R <sup>2</sup> = 0.135, p < 0.001). Teachers' conservation values were positively associated to self-efficacy regardless of the type and level of motivations for teaching. The relations of openness to change and self-transcendence with the criterion variable were instead moderated by teachers' motivations. Examination of the interaction plot showed that at low levels of controlled motivations (1 standard deviation below the mean), teachers' openness to change supported their sense of efficacy (**Figure 1**). At higher levels of autonomous motivations (1 standard deviation above the mean), a positive association between self-transcendence and self-efficacy emerged and was stronger than that found at low levels of autonomous motivations (**Figure 2**). Finally, teachers' self-enhancement values were not related with self-efficacy.

#### DISCUSSION

Teachers' self-efficacy has been repeatedly demonstrated to be a relevant factor for the effectiveness of the teaching activity, as it is a powerful drive influencing the behavior of teachers in the classroom and the effort put in the endeavor (Klassen et al., 2009; Klassen and Tze, 2014). Therefore, improved teacher self-efficacy can result in improved teacher mental health and job satisfaction, and students' academic performance (Bandura, 1977).

Based on this, it becomes of extreme relevance to understand what influences teacher's belief in his/her ability to successfully cope with tasks, obligations and challenges related to his/her professional role (Caprara et al., 2006; Tschannen-Moran and Hoy, 2007). With this purpose, teachers' personal values may be something of great interest to explore. Indeed, values have been shown to effectively predict workers' self-efficacy (e.g., Sousa et al., 2012), but despite the relevant heuristic power that personal values have in many life contexts, they have been completely under investigated in the school context. Besides trying to fill this gap of knowledge, the present study also aims at investigating whether and the extent to which teachers' values–self-efficacy relation changes as a function of teachers' motivations for teaching (Roth et al., 2007). Specifically, the relations between personal values and self-efficacy was expected to be stronger for self-determined teachers, namely those teachers characterized by low controlled motivations (i.e., performing teaching activity because of external rewards) and high autonomous ones (i.e., attributing an intrinsic value to their professional activity). Self-determination may activate values and related behaviors, thus strengthening the values–selfefficacy relation.

From this study it emerged that teachers gave the greatest importance to self-transcendence and conservation values. This result is in line with previous studies on teachers (e.g., Verkasalo et al., 1996) but also with studies on the Italian adult population in general (e.g., Caprara et al., 2011). Thus, the teachers who participated in the study recognized the importance of values like the welfare of ingroup members, tolerance, social justice, world beauty (i.e., benevolence and universalism, conformity, security, and tradition). Interestingly, according to Schwartz's (2012) refined model of values, self-transcendence and conservation are both values focused on social outcomes, namely values primarily regulating how one socially relates to others and affects them,

TABLE 2 | Moderated regression analysis results (dependent variable: teachers' self-efficacy).


∗∗p < 0.01 and <sup>∗</sup>p < 0.05. CM, controlled motivations; AM, autonomous motivations.

fpsyg-10-01645 July 11, 2019 Time: 17:36 # 5

whereas openness to change and self-enhancement are mostly focused on self-interest. Thus, in their value priorities teachers appear to be socially oriented, both in terms of caring for others (i.e., benevolence and universalism) and in terms of maintenance of the current order (i.e., tradition, conformity, and security).

All in all, teachers' personal values were shown to be important predictors of teachers' self-efficacy. Specifically, conservation was positively associated to teachers' self-efficacy, both for those teachers driven by controlled motivations in their work and for those teachers driven by autonomous motivations. It is likely that respecting norms and safeguarding stability in the field of education (where teachers are supposed to feel the responsibility to transmit knowledge and to "take care" of students) makes teachers experience a sense of security in planning, organizing, and carrying out activities required to attain educational goals. The specific context where the study was carried out explains therefore the contrasting finding compared to the results by Sousa et al. (2012), where conservation had a negative impact on selfefficacy. The result is, however, more in line with Djigic et al. ´ (2014), who found teachers' conscientiousness to be positively related with self-efficacy.

The relations between openness to change and selftranscendence values and self-efficacy varied instead as a function of teachers' motivations for teaching. Indeed, the relation between openness to change and self-efficacy was positive at low levels of controlled motivations, but negative at high levels of controlled motivations. In other words, teachers for whom values such as novelty, freedom, and choosing one's own goals are important for their perception of efficiency actually feel more efficient in their teaching activities when they feel independent from external conditions. On the contrary, where the external pressures to behave in a certain way are strong, the more teachers give importance to openness to change, the more they perceive themselves as inefficient. According to the personenvironment fit theory (see Caplan, 1987), this result shows that teachers' self-efficacy is higher when their personal values and work environment attributes are perceived as compatible.

In this line, also a genuine interest for others – in terms of importance assigned to self-transcendence values – for teachers who recognized the intrinsic value of their teaching activity was found to be a significant factor in fostering teachers' selfefficacy. Teachers who show concern for their students and "value" themselves in terms of what they do, can profit in terms of self-efficacy.

In interpreting these results, some limitations need to be taken into consideration. The main shortcoming of the present study is the involvement of a single-country convenience sample, since participants were selected according to the high school collaboration, and its cross-sectional design, which limits casual inferences from the data as well as possible considerations with regards to bidirectionality of the links. Caution is needed in generalizing results since the sample was mainly composed of female teachers (73.6%); however, this reflects the Italian reality, where the majority of teachers are women.

Based on the relevance of self-efficacy for teaching effectiveness and students' academic achievement, future research may also address the role of other variables in influencing teachers' self-efficacy. For example, it would be of great interest to expand the study of teachers' self-efficacy by considering their emotional intelligence (Magnano et al., 2017b) and organizational mindfulness (Magnano et al., 2017a).

However, it is worth noting that, to the best of our knowledge, this is the first study focused on the relations between teachers' personal values and their self-efficacy. We believe the strongest point of this study is to have shown that values, under certain motivational conditions, are precious resources to improve the quality of teaching experience for teachers and indirectly for their students. Self-knowledge and conscious endorsement of personal values could therefore help teachers to healthily handle the challenges of teaching and to prevent work-related stress and burnout. Further research about teachers' values and their implications for teachers' wellbeing and teaching practices should be promoted, particularly in culturally specific contexts.

#### DATA AVAILABILITY

fpsyg-10-01645 July 11, 2019 Time: 17:36 # 6

The datasets for this manuscript are not publicly available because of local legal and privacy restrictions (Italian Data Protection Code – Legislative Decree No. 196/2003). However, the raw data supporting the conclusions of this manuscript can be made available by the first author to qualified researchers upon request.

#### ETHICS STATEMENT

The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation. All participants gave written informed consent in accordance with

#### REFERENCES


the Declaration of Helsinki. The study was approved by the Ethics Committee of the Catholic University of Milan, Italy.

#### AUTHOR CONTRIBUTIONS

DB designed and carried out the study, contributed to the analysis of the results and to the writing of the manuscript. FD contributed to the analysis of the results and to the writing of the manuscript. PB contributed to the writing of the manuscript.

#### FUNDING

Preparation of this article was supported by a grant from the LUMSA University, Rome, Italy.


fpsyg-10-01645 July 11, 2019 Time: 17:36 # 7

employees. Int. J. Sel. Assess. 20, 159–170. doi: 10.1111/j.1468-2389.2012.00 589.x


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Barni, Danioni and Benevene. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# The Impact of Emotions and Hedonic Balance on Teachers' Self-Efficacy: Testing the Bouncing Back Effect of Positive Emotions

#### Ilaria Buonomo\*, Caterina Fiorilli and Paula Benevene

Department of Human Science, Libera Università Maria SS. Assunta, Rome, Italy

Emotions toward students (e.g., Chan, 2004) and professional role (e.g., O'Connor, 2008) impact teachers' self-efficacy (TSE) beliefs. The effect of positive emotions (PEs) can be explained by the broaden and build theory, stating that the higher the PEs individuals attribute to themselves, the higher the chance to build positive aspects of the self (Fredrickson, 2001). At the same time, negative emotions (NEs) at school inversely influence TSE, reducing teachers' confidence (Chan, 2004). Furthermore, Fredrickson et al. (2000)'s studies inform about the bouncing back effect of PEs on the detrimental effects of NEs on self-efficacy. Starting from these considerations, this study (1) evaluated the bouncing back effect of PEs on negative ones, when predicting self-efficacy; (2) verified whether emotions toward professional role moderated the bouncing back effect. Self-efficacy and emotions in teaching (MESI, Moè et al., 2010) were measured. Two hundred and seventy-two Italian secondary school teachers (F = 73%) were involved. PEs toward students might act as buffering factors against the detrimental effect of NEs on self-efficacy [F(2,270) = 26.17, P < 0.001, R <sup>2</sup> = 0.199]. Finally, emotions toward students and emotions toward professional role do not interact when predicting self-efficacy. The relationships with students seem to have an highly protective effect on teachers' mental health. At the same time, the perception of one's own role as detached from the teaching community may have a role in justifying the non-significant effect of emotions toward professional role in the model and shed new light on intervention objectives.

#### Keywords: teachers, self-efficacy, trait emotions, undoing effect, professional role

# INTRODUCTION

Tschannen-Moran et al. (1998) defined teachers' self-efficacy (TSE) as the teacher's belief in his or her capability to organize and execute the actions required to successfully accomplish professional tasks in a specific context (Tschannen-Moran et al., 1998). Teacher beliefs about professional efficacy influence several aspects in the teaching–learning process, including classroom environment, student performance, and teacher practices (Goddard et al., 2007; Santisi et al., 2014; Zee and Koomen, 2016; Benevene et al., 2018; Granziera and Perera, 2019). Accordingly, teachers with low self-efficacy are at high risk for absenteeism, physical illness, and burnout (Zee and Koomen, 2016). Given the impact of TSE on school life, influencing factors should be

Edited by: Tiziana Ramaci, Kore University of Enna, Italy

#### Reviewed by:

Harsha Perera, University of Nevada, Las Vegas, United States Vincenzo Russo, Università IULM, Italy

> \*Correspondence: Ilaria Buonomo i.buonomo1@lumsa.it

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 04 February 2019 Accepted: 02 July 2019 Published: 17 July 2019

#### Citation:

Buonomo I, Fiorilli C and Benevene P (2019) The Impact of Emotions and Hedonic Balance on Teachers' Self-Efficacy: Testing the Bouncing Back Effect of Positive Emotions. Front. Psychol. 10:1670. doi: 10.3389/fpsyg.2019.01670

individuated, to shed a new light on well-being promotion at school. Consistently with Bandura's (1986) social-cognitive theory, Tschannen-Moran et al. (1998) individuated four main sources for TSE: vicarious experiences, sense of mastery, verbal persuasion, and physiological and emotional arousal. At the same time, the authors considered physiological states and associated emotions as less effective in predicting TSE when compared to other sources. This is due to the fact that physiological states and emotions must be carefully attended to have an effect on TSE (Tschannen-Moran et al., 1998). Several studies have shown that, when workers attend their positive emotional experiences, their self-efficacy raise (Wright and Staw, 1999; Lyubomirsky et al., 2005; Tsai et al., 2007; Staw et al., 2008; O'Malley and Gregory, 2011; Xanthopoulou et al., 2012). Furthermore, current research suggests that teachers are aware of their emotional involvement in their job. More specifically, several authors show that teachers recognize the effects of emotions and relations experienced at school on their teaching goals, experiences, and results (Sutton and Wheatley, 2003; Sutton and Harper, 2009; Borrelli et al., 2014; Fiorilli et al., 2017). Consistently, some studies report that teachers who generally experiment more positive than negative emotions (NEs) feel more confident too (Sutton and Harper, 2009).

The effect of positive emotions (PEs) on self-efficacy can be read in light of the broaden and build theory, stating that the higher the PEs individuals attribute to themselves, the higher the chance to build positive aspects of the self (Fredrickson, 2001; Fredrickson and Losada, 2005). At the same time, teaching is loaded with NEs, too, for example because of student misbehavior or low social recognition (Becker et al., 2015; Mevarech and Maskit, 2015; Frenzel et al., 2016; Fiorilli et al., 2017). The emotional correlates of such problems and conflicts at school inversely influence TSE, reducing teachers' confidence (Chan, 2004). Therefore, it seems useful to take into account the cooccurrence between positive and NEs (Diener and Emmons, 1984), and how this may influence teacher self-efficacy.

Positive and NEs, indeed, activate different biological and behavioral patterns and are independently regulated over time (Diener et al., 1985; Watson, 2000).

How positive and NEs influence one another, when influencing self-efficacy? A possible explanation is provided by the construct of undoing effect (Fredrickson et al., 2000), interpreted as the ability of PEs to restore psychological resources in face of NEs and experiences (Gloria et al., 2013). When individuals make the most of their PEs, they are better at reducing the detrimental effects of negative affect on personal interpretations of events (Tugade and Fredrickson, 2004, 2007). Consistently, previous studies reported that positive affect mediated the relationship between NEs at work and psychological resources in teachers (Gloria et al., 2013). Recognizing PEs at work, indeed, could act as a buffer, through which the potential detrimental effect of NEs on self-efficacy would be lowered. Despite this, to the best of our knowledge, no studies have explored such effects on self-efficacy.

When studying teacher emotions, a distinction should be made between emotions toward students and emotions toward the teaching professional role (Moè et al., 2010). While many studies addressed the firsts, few studies addresses the seconds, despite showing meaningful implications for the profession (Sutton and Wheatley, 2003). More specifically, teachers' emotions toward their role guide their practices and shape their professional identities, independently from the actual teaching experience in the classroom (O'Connor, 2008; Gonzalez-Calvo and Arias-Carballal, 2017). In other words, emotions seem a crucial point of teachers' professional life. According to Palmer (1998), this is due to the fact that the teaching profession is at the crossroad of personal and public life. With this regard, teaching professional role involves two interacting dimensions: intrapersonal factors, such as emotions and personal history (e.g., Zembylas, 2003; Van Veen and Sleegers, 2006) and social factors, such as personal and professional relations and social representations about the profession (Holstein and Gubrium, 2000; Roa-Tampe, 2017). Teachers, indeed, spontaneously refer to emotions when asked to enumerate the main dimensions of their profession (Buonomo and Fatigante, 2017; Gonzalez-Calvo and Arias-Carballal, 2017; González-Calvo and Fernández-Balboa, 2018). This broader "emotional context," connected with teachers' identities, influences emotions experienced with students during classroom interactions. O'Connor (2008), indeed, argued that teachers justify their caring attitude toward students as a main part of their professional identity. At the same time, current research suggests that teachers experience several NEs toward their professional role, above all related to low social recognition of their profession and feeling of social underestimation (OECD, 2014; Buonomo et al., 2017). Teachers able to balance the contrasting emotions felt toward professional role may be better at mitigating the detrimental effects of NEs on TSE, thus strengthening the protective impact of positive classroom experiences on their sense of confidence. The ability to balance positive and NEs (namely, hedonic balance; Diener et al., 1999) regarding the role, indeed, was related to efficacy beliefs in Italian teachers (Caprara and Steca, 2006).

Overall, emotions related to students and to the teaching role can be considered as distinct sources of information about teachers' emotional experiences. Generally speaking, the emotional acknowledgment at work predicts self-efficacy beliefs (e.g., Xanthopoulou et al., 2012). For this reason, we hypothesized that emotions toward students would explain teachers' selfefficacy. At the same time, as teachers' identities are embedded with emotions (O'Connor, 2008), we considered emotions toward professional role as generating a broader emotional context that could influence the predictive effects of emotions toward students. More specifically, this study aimed to verify two hypotheses:

Hypothesis 1 – positive emotions toward students mediate the detrimental effect of negative emotions toward students on self-efficacy;

Hypothesis 2 – hedonic balance related to professional role moderates the mediating effect of positive emotions, so that the higher the hedonic balance, the higher the mediating effect of positive emotions.

Participants were 272 Italian school teachers (Females = 73%), aged 26–66 (M = 51.50; SD = 7.96) years and with 1– 41 years of experience (M = 21.74; SD = 10.37). The most of them were class (91%) and permanent (89%) teachers. Data collection occurred during school meetings, with principals' consent. Teachers were instructed by written informed consent that they could leave the study and ask for further information at any time. Teachers' self-efficacy and emotions were assessed with the Metacognitive Questionnaires for Teachers, validated on the Italian teaching population (MESI; Moè et al., 2010). Self-efficacy was measured with 24-item, on a 10-point Likert scale (1 = not effective, 10 = totally effective) (sample item: "I am good at managing students' oppositional behaviors"). Teachers' PE and NE, toward students and professional role, were measured with 30 items ("how frequently do you feel these emotions when you are with your students/when you think about your profession?"), measured on a 5 point Likert scale (1 = almost never, 5 = almost always). The subscale measured 17 NEs (shame, anger, uneasiness, inadequacy, wrath, exasperation, indignation, sadness, sense of failure, guilt, resignation, irritation, antipathy, frustration, discouragement, nervousness, disappointment) and 13 PEs (cheerfulness, enthusiasm, caring, commotion, admiration, complacency, pleasure, happiness, satisfaction, fulfilment, sense of accomplishment, joy, enrichment). NE and PE are computed as total mean scores of the answers to the 17 and 13 items, respectively (Moè et al., 2010). Starting from the scales of PE and NE toward the professional role, hedonic balance related to professional role was computed as the difference between PE and NE (Diener et al., 1999). Consequently, when the hedonic balance is high, the respondent perceives more PE than NE. Reliability was examined by estimating Cronbach's alpha and McDonald's omega (McDonald, 1999), as the last has been advocated as more informative than Cronbach's alpha (Revelle and Zinbarg, 2009), especially for heterogeneous scales. Omega is an indicator of general factor saturation, interpreted as the precision with which scale scores estimate a latent variable common to all test items. Overall, Cronbach's alphas ranged 0.835 (NE toward professional role) to 0.975 (PE toward professional role), while McDonald's Omegas ranged from 0.92 (NE toward professional role and NEs toward students) to 0.95 (PE toward professional role).

The associations among self-efficacy, sociodemographic variables, student-related emotions, and role-related hedonic balance were calculated with Pearson's product-moment correlations. As sociodemographic variables did not correlate with other variables, these were not included in the analyses. Two multiple regression analyses were conducted: (1) Hypothesis 1 (PEs toward students mediate the detrimental effect of NEs toward students on self-efficacy) was verified with a mediation model, in which beta weights are used to measure direct and indirect effects and bootstrap method (5000 samples) is used to test the effect size of the indirect effect; (2) Hypothesis 2 (hedonic balance related to professional role moderates the mediating effect of PEs) was verified with the beta weight of the interaction between hedonic balance and PE toward students. Variables were mean centered prior to the formation of the interaction term (Aiken and West, 1991), to differentiate the moderating effect at low, medium, and high levels of hedonic balance. Predicted probability plots, residuals scatterplots, and variance inflation factor (VIF) < 5, and Tolerance > 0.80 criteria were used to test, respectively, the normality, homoscedasticity, and non-multicollinearity assumption for regression. All the variables for all the regression models fulfilled the assumptions. Finally, a p > 0.001 Mahalanobis' distance criterion was used to identify and skip multivariate outliers. Analyses were run with the PROCESS macro (Hayes, 2012) for IBM SPSS (vr. 23) and with the psych package in R (vr. 3.5.3) (Revelle, 2016).

# RESULTS

**Table 1** shows mean and standard deviation values and correlations among self-efficacy, student-related PE and NE, and role-related hedonic balance. All the associations were significant (p < 0.01). Overall, the higher the PEs perceived with students, the higher the hedonic balance related to professional role, the higher the perceived self-efficacy.

**Table 2** shows the simple mediation model. As expected (Hypothesis 1), student-related PEs mediated the relationship between student-related NEs and self-efficacy [F(2,270) = 26.17, p < 0.001, R <sup>2</sup> = 0.199]. The effect of student-related NEs on self-efficacy was significant [b = −0.4918, t(273) = −5.0161, p < 0.001], as well as the effect of student-related PEs on selfefficacy [b = 0.3681, t(271) = 3.73, p < 0.001]. The indirect effect was significant (95% CI = −0.292 to −0.093; bootstrapped unstandardized indirect effect = −0.161, bootstrapping run on 5000 samples). Overall, results show a partial mediation of PEs in the relationship between NEs and self-efficacy.

**Table 3** shows the moderated mediation model (or indirect conditional effect; Preacher et al., 2007). Contrarily to Hypothesis 2, the role-related hedonic balance did not moderate the mediation of student-related PEs on the relationship between student-related NEs and self-efficacy [interaction effect was b = −0.08, t(271) = −1.35, NS].

# DISCUSSION

Our findings showed that PEs toward students partially mediate the relationship between NEs toward students and self-efficacy in a sample of Italian teachers, confirming our first hypothesis. At the same time, our results showed that the hedonic balance toward professional role did not influence the effect of PEs on self-efficacy, contrarily to our second hypothesis.

Overall, this study confirmed the role of emotions in predicting teaching self-efficacy, strengthening socialcognitive theory statements about the role of emotional arousal in predicting efficacy beliefs (Bandura, 1986; Tschannen-Moran et al., 1998).

#### TABLE 1 | Mean, standard deviations, and correlations among variables.



Unstandardized regression coefficients are reported. Bootstrap sample size: 5000. LL, lower limit; CI, confidence interval; UL, upper limit.

TABLE 3 | Regression results for conditional indirect effect.


N = 272. Unstandardized regression coefficients are reported. Bootstrap sample size: 5000.

According to our first hypothesis, when teachers frequently experience NEs toward their students, they are at higher risk for believing to be not effective in the classroom. Emotional labor is a possible explanation for this finding, as it may have influenced teachers' answers. Emotional labor is the strain emerging from the contrast between felt emotions and organizational rules about emotion expression (Schutz and Lee, 2014). In other words, teachers may feel they are not allowed to express certain emotions because of the school's rules about emotional disclosure (Yin and Lee, 2012). According to the emotional labor framework, indeed, the higher the emotions felt but not expressed, the higher the stress and burnout risk (Yin and Lee, 2012). At the same time, and contrarily to our results, other studies (Sutton and Harper, 2009) showed that teachers tend to express or fake anger toward their students in order to maintain discipline and that, by means of this strategy, they feel more effective at managing their classroom. To better understand these apparently contrasting results, the type of studied emotions should be considered. Studies about teachers' emotions, indeed, usually address basic NEs (Sutton and Harper, 2009; Yin and Lee, 2012). In our study, teachers answered about their complex, self-conscious emotions (e.g., sense of failure, guilt, resignation, frustration, discouragement; Tracy and Robins, 2007), linked to experiences of loss or blame (Thamm, 2006). This may explain the reduction of self-efficacy. Individuals with high negative self-conscious emotions, like blame, guilt, and discouragement, are at higher risk to experience low selfesteem (Kuppens and Van Mechelen, 2007). Therefore, it is likely that these emotions may have a role in reducing selfefficacy, too.

According to our first hypothesis, findings showed that PEs toward students partially protect teachers from detrimental effect of negative ones on self-efficacy. As reported by Llorens et al. (2007), positive emotional states may help workers perceiving themselves more efficacious in performing daily tasks at work. Moreover, according to the broaden and build theory and its applications at work (Tugade and Fredrickson, 2004, 2007; Gloria et al., 2013), PEs may restore psychological resources (e.g., self-efficacy) despite the acknowledgment of NEs. How teachers' PEs may have such a role? van der Want et al. (2018) stated that each teacher has a professional interpersonal standard, depending on personal beliefs and values, and related to the relationship with students.

According to the authors, affective and cognitive appraisals about personal well-being and emotions and coping with stressful situations are represented accordingly to the interpersonal identity standards and guide teachers' experience and expression of emotions. Accordingly, we have found that teachers recursively make use of the garden metaphor, reporting to be emotionally refueled by their students' recognition (Buonomo and Fatigante, 2017). This point may clarify the bouncing back effect (Tugade and Fredrickson, 2004) found in this study: if teachers are willing to "accept" a contextual stressor, hoping to reach a more favorable outcome, then it is likely that PEs may help to build stronger efficacy beliefs regarding their teaching abilities.

Finally, contrarily to our second hypothesis, PEs toward professional role did not moderate the mediation effect shown in the study. This finding may be due to the characteristics of the Italian educational system itself. According to Bracci (2009), the web of accountability in Italian schools is getting confused because of several changes and reforms: in such a condition, it is likely that teachers take an individualistic point of view, giving more saliency and attention to the personal experience of teaching, than to the idea of being part of a specific group of professionals. Moreover, previous studies showed that being happy to be part of a school institution does not necessarily promote teachers' efficacy beliefs (Malinen and Savolainen, 2016; Buonomo et al., 2017).

Regarding the limitations of the study, the use of longitudinal data is needed, to better address the mediation effect, giving more robust support to the findings, as mediations on crosssectional data are known to be biased because of autoregressive effects (Maxwell and Cole, 2007). The impossibility to distinguish between "actual" predictors and results in cross-sectional designs (namely the impossibility to define a temporal lag between predictor and mediator, and mediator and outcome) weakens the inferences that can be made from our model. Moreover, the lack of a temporal lag didn't allow to verify whether, as reported from previous studies (Canrinus and Fokkens-Bruinsma, 2011; Rodríguez-Sánchez et al., 2011; Lohbeck et al., 2018), feeling effective as a teacher may have influenced the level of PE and NE experienced in the classroom. For example, indeed, previous studies acknowledged that teachers with high self-efficacy are more prone to experience PE than NE with students, and vice

#### REFERENCES


versa (Tschannen-Moran et al., 1998). Furthermore, a distinction between the effects of trait- and state-emotions on self-efficacy should be done, as they differently impact teachers' professional experiences (Goetz et al., 2015). Finally, a multimethod approach would give more information on how teachers perceive their emotions and how emotional experiences impact self-efficacy (Sutton and Wheatley, 2003).

Despite these limitations, our study shed new light on the effect of emotions on self-efficacy in teachers and have some practical implications concerning teachers' training, regarding positive psychology interventions on school climate (Jennings and Greenberg, 2009). Improving teachers' abilities to acknowledge positive exchanges and experiences within the school context may heighten their abilities to recognize personal attainments and successes as professionals. Moreover, our study suggests taking into account the organizational dimension of the teaching profession when intervening on teaching communities. Cultivating collaboration and sense of community within schools, indeed, may strengthening the effect of PEs: from one side, by increasing PEs during daily school life (Jennings and Greenberg, 2009), from the other, by improving positive perceptions of the teaching profession.

#### ETHICS STATEMENT

The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation. All participants gave written informed consent in accordance with the Declaration of Helsinki. Ethical implications were evaluated by the Scientific Board of Libera Università Maria SS. Assunta (LUMSA), the body responsible for providing ethics approvals for studies, that approved the study.

#### AUTHOR CONTRIBUTIONS

IB designed and carried out the study. IB and CF contributed to the analysis of the results and to the writing of the manuscript. PB supervised the study design and the manuscript draft.

Bracci, E. (2009). Autonomy, responsibility and accountability in the Italian school system. Crit. Perspect. Account. 20, 293–312. doi: 10.1016/j.cpa.2008.09.001




and personal resources. Eur. J. Work Organ. Psychol. 21, 489–517. doi: 10.1080/ 1359432X.2011.584386


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Buonomo, Fiorilli and Benevene. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# School Burnout, Relational, and Organizational Factors

#### *Santa Parrello1 \*, Alice Ambrosetti2 , Ilaria Iorio1 and Luciana Castelli2*

*1 Section of Psychology and Educational Sciences, Department of Humanities, University of Naples Federico II, Naples, Italy, 2 Centre for Innovation and Research on Education Systems (CIRSE), Department of Education and Learning, University of Applied Sciences and Arts of Southern Switzerland (SUPSI), Locarno, Switzerland*

Work stress and burnout affect teachers to a significant extent. The objective of this study is to evaluate and compare the impact of relational and organizational factors on teacher burnout in two samples of primary school teachers, one Italian (Naples) and the other Swiss (Cantone Ticino). The hypothesis is that, given the socio-cultural and economic differences of the two contexts, the variables under investigation impact teacher burnout differently. We collected data through a self-reported questionnaire containing the following scales: Copenhagen Burnout Inventory, Satisfaction with Life Scale, Life Orientation test, organizational identification, colleague support, and workload. The Swiss sample consists of 964 teachers (26% kindergarten and 73.7% primary school teachers); the Italian sample consists of 104 teachers (20% kindergarten and 80% primary schools teachers). Descriptive analyses, mean comparison (*t* test), correlational analyses, and multiple linear regression analyses were conducted. There are no significant differences between the two samples with respect to burnout, colleague support, and workload. Correlations between burnout and the variables under investigation are significant in both samples, except for optimism in the Italian sample. Regression analysis shows that optimism and colleague support have an impact on burnout only in the Swiss sample; organizational identification has a stronger impact on burnout in the Italian sample.

Keywords: factors of school burnout, organizational identification, optimism, teachers' well-being, comparative study

# INTRODUCTION

Teachers' well-being is a resource for schools that is constantly at risk and deserves to be protected (Hakanen et al., 2006). Recent studies on risk and protective factors of teacher burnout explore some relational and contextual aspects, such as the complexity of the educational role, the burden of responsibility, the support of colleagues, the relationship with the organization, and the social devaluation of the profession (Fiorilli et al., 2015; De Rosa et al., 2017).

In particular, the support received from colleagues (Brouwers et al., 2011) and organizational identification (Ashforth and Mael, 1989) is considered as protective factors from teacher burnout. In fact, it was found that a strong sense of organizational identification is negatively correlated to stress and burnout (Bizumic et al., 2009) and positively related to well-being (Van Dick and Wagner, 2002). Moreover, work burnout interrelates with social support (Avanzi et al., 2018) and collective efficacy (Avanzi et al., 2015).

#### *Edited by:*

*Simona De Stasio, Libera Università Maria SS. Assunta, Italy*

#### *Reviewed by:*

*Pina Marsico, University of Salerno, Italy Carlos Prado Romero, National Autonomous University of Mexico, Mexico*

> *\*Correspondence: Santa Parrello parrello@unina.it*

#### *Specialty section:*

*This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology*

*Received: 06 June 2019 Accepted: 05 July 2019 Published: 02 August 2019*

#### *Citation:*

*Parrello S, Ambrosetti A, Iorio I and Castelli L (2019) School Burnout, Relational, and Organizational Factors. Front. Psychol. 10:1695. doi: 10.3389/fpsyg.2019.01695*

**47**

This study analyzes and compares the impact of some relational and organizational factors on teacher burnout in two different countries, Switzerland (Ticino, the Swiss Italianspeaking Canton) and Italy. These countries differ across various socio-cultural and economic aspects, making a comparison legitimate. Indeed, as explained above, burnout does not only depend on individual factors but also on contextual factors. The following indicators partially show the contextual differences in which the teachers of the two countries carry out their daily professional activity.

First, in Ticino, kindergarten and primary schools are under the communal and the cantonal authority; in order to obtain the teaching degree, students have to obtain a bachelor's degree in a professional university. Teaching is a highly secure job, and working conditions are good. The weekly timetable is between 32 and 25 h (depending if school meals are included); moreover, meetings and interviews with the parents have to be taken into account. In Italy, schools are under the national authority, namely the Ministry of Education. In the past, high school degree and teaching degree were mandatory in order to become a teacher; now, a 4-year bachelor's degree is mandatory. Teaching is a highly secure job, but working conditions are quite unfavorable. The weekly timetable is of 22 h of classroom teaching, 2 h of school planning, 40 h per year for meetings and interviews with the parents, and 40 h per year for training.

With regard to working life, from the point of view of the economic situation and working conditions, Switzerland is above the OECD average with respect to the average level of disposable income per capita per year, while Italy is below the OECD average (OECD average: 30,563 USD; Switzerland: 36,378 USD; Italy: 26,063 USD). With respect to contractual wages, Italian salaries are lower than those of Switzerland and of the OECD average (OECD, 2017); the unemployment rate (IMF, 2018) in Italy is 11.26%, while in Ticino, it is 3.19% (DFE, 2019). The average life satisfaction in Switzerland is 7.9, compared to the OECD average of 7.3 and to the Italian average of 7.0 (OECD, 2017). In both countries, students (age 11–15) do not have a great deal of sympathy for the school, but Swiss students have a greater perception of the effectiveness of their schools and of the social support received in it compared to their Italian peers (HBSC, 2016). The rate of early school leaving in Italy is very high, with significant regional differences (Cederna, 2016; Parrello et al., 2019).

The aim of this study is to analyze the impact of relational and organizational factors on teacher work burnout, studentrelated burnout, and life satisfaction in Naples and Cantone Ticino.

The main hypothesis is that contextual factors have an impact on teacher burnout and on the relational and organizational factors related to it.

We expect that Italian teachers to have higher levels of work burnout (H1), higher levels of student-related burnout (H2), and lower life satisfaction (H3) compared to Swiss teachers. Furthermore, we assume that optimism, workload, colleague support, and organizational identification have an impact on work burnout, student-related burnout, and life satisfaction (H4).

# MATERIALS AND METHODS

We examined our hypothesis on a sample that comprised two sub-samples of kindergarten and primary school teachers; the first one was composed of teachers from Naples, and the second one of teachers from Ticino.

The Italian sample was a convenience sample selected from a school located in the urban area of Naples. In Ticino, the whole population of kindergarten and primary school teachers was invited to participate in the survey (51.8% response rate).

The Swiss sample comprises 964 teachers: 710 (73.7%) of primary school, 251 (26%) of pre-primary school, and three subjects did not provide this information. In the Italian sample, 20% of teachers (*N* = 21) work in kindergarten, and the rest in primary school. In Ticino, 20% (*N* = 195) of teachers are male, while in the Italian sample, just two teachers are male.

Mean age is 42.79 (SD = 11.35) in the Swiss sample and 53.89 (SD = 7.1) in the Italian sample.

Data have been collected through a self-reported questionnaire; in Switzerland, the questionnaire was administered online; in Italy, the questionnaire was paper-pencil administered.

The following measures have been administered to both subsamples.


successes are my successes"). Items are rated on a 5-point Likert scale, ranging from 1 "Totally disagree" to 5 "Totally agree."


On the data collected, we first performed a comparison of means (Student's *t* test) between the two sub-samples for burnout, life satisfaction, workload, optimism, organizational identification, and colleague support. The goal was to establish if there were statistical differences between the two samples across the variables of interest.

After a correlation analysis between the above-mentioned variables (the Pearson correlation), we performed a hierarchical linear regression in three different models on both samples.

In the first step, we added control variables, in the second step optimism and in the third step colleague support, organizational identification and workload, in order to appreciate their impact on work burnout, student-related burnout, and life satisfaction.

The regression models include the same set of variables for both samples, in order to compare the regression coefficients between the two samples.

# RESULTS

Mean comparison shows that the two burnout measures do not differ between the two samples: teachers from Ticino and teachers from Naples show the same level of work (respectively: *M* = 2.44, SD = 0.73; *M* = 2.45, SD = 0.83, *p* > 0.05) and student-related burnout (*M* = 2.27, SD = 0.70; *M* = 2.17, SD = 0.87, *p* > 0.05).

Swiss teachers show higher levels of life satisfaction than teachers from Naples (*M* = 5.25, SD = 1.00; *M* = 4.64, SD = 1.09, *p* < 0.05). On the contrary, teachers from Naples are on average more optimistic than teachers from Ticino (*M* = 3.62, SD = 0.46; *M* = 2.63, SD = 0.45, *p* < 0.05). Moreover, teachers from Naples show higher levels of organizational identification than teachers from Ticino (*M* = 4.01, SD = 0.97; *M* = 3.70, SD = 0.75, *p* < 0.05). Regarding workload and colleague support, we could not find any statistically significant difference.

In the second step of the analysis, we estimated the Pearson correlation coefficients between all the scales, for each teacher sample (see **Table 1**). Correlations between teacher burnout scales and the other scales are statistically significant in both samples, except for optimism in the Italian sample. Correlations between optimism and work (*r* = −0.48, *p* < 0.05) and studentrelated burnout (*r* = −0.39, *p* < 0.05) are negative and significant only in the Swiss sample. Correlation coefficients between the two burnout scales and the other scales are similar between the two samples, with the exception of the organizational identification scale. In fact, correlation coefficients are *r* = −0.45, *p* < 0.05 for work burnout and *r* = −0.34, *p* < 0.05 for studentrelated burnout in the Italian sample, and *r* = −0.07, *p* < 0.05 and *r* = −0.12, *p* < 0.05 in the Swiss sample.

The Satisfaction with Life Scale correlates significantly with all the other scales in both samples, with the exception of organizational identification in the Italian sample. Moreover, the correlation between satisfaction with life and colleague support is higher for Italian teachers than for Swiss teachers (respectively, *r* = 0.36, *p* < 0.05 and *r* = 0.23, *p* < 0.05).

Regarding the correlation between the other scales, results are similar between the two samples, with some exceptions. First, workload has a negative and significant correlation with colleague support only in the Ticino sample (*r* = −0.21, *p* < 0.05), while there is no significant correlation between the two scales in the Italian sample. Second, organizational identification is significantly correlated with satisfaction with life (*r* = −0.11, *p* < 0.05) and optimism (*r* = 0.08, *p* < 0.05) only in the Swiss sample, while there are not statistically significant correlations in the Italian sample.

All the other scales are positively correlated with each other, with the exception of workload and organizational identification that do not show significant correlations in both samples (see **Table 1**).

In order to better establish the relationship between burnout, life satisfaction, and the other variables under investigation, we performed a hierarchical linear regression in three steps (**Table 2**).

The first model shows that in the Swiss sample, age and gender do not have any statistically significant impact. Instead, having children seems to have a modest but positive impact on work burnout: Swiss teachers with children show a lower level of work burnout than the ones having one child or more (*β* = −0.08, *p* < 0.05). The same result is not found in the Italian sample. Age is a predictor of work burnout for Italian teachers (*β* = 0.02, *p* < 0.05).

A difference is also found for optimism: the effect is quite strong for Ticino's teachers (*β* = −0.51, *p* < 0.05), whereas for the Italian sample, this variable has no effect on the work burnout. Likewise, for colleague support, we did not find an effect for the sample from Naples, but a statistically significant one in the Ticino sample. Namely a high perceived colleague support corresponds, on average, to a high work burnout (*β* = −0.06, *p* < 0.05). At the opposite, organizational identification seems to play a role in predicting work burnout only in the Naples sample (*β* = −0.27, *p* < 0.05). Finally, workload has an effect in both samples: teachers with high values of workload have, on average, high values of work burnout (*β* = 0.51, *p* < 0.05 for Ticino, *β* = 0.67, *p* < 0.05 for Naples).

The first model explains, in total, 45% of the variance for the Swiss sample and 54% for the Italian sample.

#### TABLE 1 | Correlations between scales.


*\*Correlation is significant at the 0.05 level (two tailed).*

*\*\*Correlation is significant at the 0.01 level (two tailed).*

#### TABLE 2 | Regressions results.


*The β coefficients reported are relative to Step 3. Independent variables: gender 0 = male, 1 = female; dependent children 0 = no, 1 = yes. \*p < 0.05; \*\*p < 0.01; \*\*\*p < 0.001.*

Interestingly, for Naples, adding the third block substantially increases the predictive power (Δ*R*<sup>2</sup> = 0.46), as if in this sample, the control variables and optimism barely explain work burnout. For Ticino, the second and the third blocks seem to explain the same share of variance (respectively, Δ*R*<sup>2</sup> = 0.23 and Δ*R*<sup>2</sup> = 0.21).

From the second model, which has student-related burnout as the dependent variable, different results with respect to the first model emerge. In fact, gender is statistically significant for Ticino, whereas age and having children are no longer statistically significant. Women are thus more likely, all other things being equal, to have a lower level of student-related burnout than male teachers (*β* = −0.18, *p* < 0.05). For the Italian sample, the control variables have the same impact as in the first model, particularly with age having a negative impact (*β* = 0.03, *p* < 0.05).

For optimism, we also found the same results as in Model 1, with these variables having an effect on student-related burnout only in the Swiss sample (*β* = −0.43, *p* < 0.05). That is, teachers with high values of optimisms are less likely to have high values of student-related burnout. Regarding the other three scales introduced, we found again an effect of workload in both samples (*β* = 0.35, *p* < 0.05 for Ticino, *β* = 0.59, *p* < 0.05 for Naples) and organizational identification is now significant for Ticino (*β* = −0.06, *p* < 0.05), whereas it is no more significant for Naples. Colleague support is not significant in both samples.

This model explains 29% of the variance for the Swiss sample and 39% for the Italian sample. Values are lower than in the first model, meaning that the set of variables introduced in the models better explains the work burnout dependent variable. Again, for the Italian sample, the salient variables are the ones included in the third block (Δ*R*<sup>2</sup> = 0.29). As in the first model, the second and the third blocks of variables in the Swiss sample seem to contribute in the same proportion (Δ*R*<sup>2</sup> = 0.15 and Δ*R*<sup>2</sup> = 0.11).

Finally, the third model has life satisfaction as dependent variable. For the Italian sample, we found no significant effect of the control variables, while for the Swiss sample, age and having children have an impact on the satisfaction with life. In fact, older teachers seem to have lower values of satisfaction with life (*β* = −0.01, *p* < 0.05), and having children increases the satisfaction with life (*β* = 0.22, *p* < 0.05), other things held constant. Optimism has a strong and positive effect in both samples: teachers with high values of optimism have in mean a high satisfaction with life (*β* = 0.99, *p* < 0.05 for Ticino; *β* = 0.83, *p* < 0.05 for Naples).

Workload has also an impact on life satisfaction in the two samples: the effect seems stronger for the Italian teachers (*β* = −0.42, *p* < 0.05 vs. *β* = −0.21, *p* < 0.05). Colleague support has an impact only on the Swiss sample (*β* = 0.09, *p* < 0.05), whereas in the Italian sample, there is no effect of this on life satisfaction. Finally, organizational identification has no impact on both the samples.

The model explains 32% of the variance for the Swiss sample and 37% for the Italian sample. As in the other two models, the third block is the one that explained mainly satisfaction with life (Δ*R*<sup>2</sup> = 0.37) in the Italian sample. On the contrary, for Ticino, the variable that better explains life satisfaction is optimism (Δ*R*<sup>2</sup> = 0.26).

#### DISCUSSION

Results confirm what has already been found in the literature, namely the impact of contextual and organizational factors on the teachers' well-being and burnout (Castelli et al., 2017).

In contrast to what has been assumed, i.e., higher levels of burnout are expected in the Italian sample than the Swiss sample, burnout, perceived support, and perceived workload are present to a similar extent in the two samples, despite the existing differences in terms of the socio-economic context, wages and working conditions. Therefore, it seems that the macro factors of the work context have a limited impact on the risk of developing burnout at work.

At the same time, results have highlighted some interesting differences between the populations under study, which could also be attributable to the differences already highlighted regarding the working conditions of the two samples: Italian teachers show a higher level of optimism and identification with their school than their Swiss colleagues who are instead more satisfied with their lives.

With respect to the correlations between the variables identified (dispositional, relational, and organizational), they are all related to burnout in both contexts, confirming the initial assumptions.

Compared to what emerges from the regression analysis, where the impact of the variables under investigation on work burnout was studied, it is very interesting to find some differences between the two samples: optimism has a significant effect on teachers in Switzerland, while for the Italian sample, this variable has no effect on work burnout. Similarly, for the support of colleagues, we did not find an effect in the Italian sample, but it is statistically significant in the Swiss sample. This would lead us to think of a difference in professional culture between the two samples considered, where Swiss teachers seem to rely more on social support than Italian teachers.

On the contrary, organizational identification seems to play a role in predicting burnout only in the sample of Naples.

Finally, the workload has a significant effect in both samples examined, confirming previously results from other studies (Castelli et al., 2017): teachers with a high workload have, on average, high values of work burnout.

If the workload therefore has an effect on both samples, another difference can be seen in the effect of optimism and organizational identification In fact, optimism has a significant effect only in the Swiss sample on the three dependent variables, whereas in the Italian sample, optimism has an effect only in the life satisfaction. Thus, it seems that in the Swiss sample, an individual disposition protects teachers from burnout, while for the Italian sample, an organizational variable seems to be more protective. It is interesting to assume, thus, that it is precisely the dimension of organizational identification that compensates for a greater presence of risk factors for Italian teachers, in line with what has already been highlighted in other works previously mentioned (Avanzi et al., 2018).

In conclusion, this study allowed to explore the protective factors of work burnout and student-related burnout in two samples that significantly differ for their working contexts and environments. However, it is important to underline some limitations of the study. The first limitation concerns the sampling procedure: the Italian sample was a convenience sample; thus, the results obtained may not be generalized to the whole population of Italian teachers. As stated before (Wilkinson and Task Force on Statistical Inference APA Board of Scientific Affairs, 1999), the use of convenience sample could be considered as acceptable, but results must be read carefully.

The second limitation regards the interpretation of the results: a causal relation between the variables under investigation has been postulated, while, in fact, the regression model does not exclude the possibility of a reverse causality; also, for this reason, results have to be read carefully.

Finally, it is possible that we omitted variables that could explain our dependent variables and that are also strongly correlated to our independents variables.

The limitations listed above suggest that a more robust sampling procedure and considering further variables in the analysis would represent potential directions for future investigations. Moreover, it would be interesting to explore the existence of mediation effects between the variables under investigation.

# DATA AVAILABILITY

The data sets generated for this study are available upon request from the corresponding author.

# ETHICS STATEMENT

The studies that involved human participants were reviewed and approved by the Ethics Committee for the Psychological

# REFERENCES


Research (CERP) of the Department of Humanities at the University Federico II of Naples. Patients/participants have provided their written informed consent to participate in this study.

### AUTHOR CONTRIBUTIONS

All authors designed the study. AA and II analyzed the data. All authors interpreted the results, wrote up the first draft of the manuscript, and approved the final version.


OECD (2017). *How's life? 2017: Measuring well-being*. Paris: OECD Publishing.

Parrello, S., Iorio, I., Carillo, F., and Moreno, C. (2019). Teaching in the suburbs: a participatory action research against educational wastage. *Front. Psychol.* (in press).


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Copyright © 2019 Parrello, Ambrosetti, Iorio and Castelli. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# Technostress Among University Teachers in Higher Education: A Study Using Multidimensional Person-Environment Misfit Theory

Xinghua Wang1,2 \* and Bo Li<sup>3</sup>

<sup>1</sup> Normal College, Qingdao University, Qingdao, China, <sup>2</sup> Center for Research and Development in Learning, Nanyang Technological University, Singapore, Singapore, <sup>3</sup> Institute of Higher Education, Linyi University, Linyi, China

To investigate the phenomenon of technostress among university teachers in higher education, a multidimensional person-environment misfit framework of technostress was proposed and validated by 343 teachers from universities in China. The findings indicate that person-organization (P-O) misfit, person-technology (P-T) misfit, and person-people (P-P) misfit largely captured how university teachers interact with multiple dimensions of the higher education environment in an imbalanced way that causes technostress. P-O misfit predicted P-T misfit and P-P misfit. Relationships between multidimensional technostress and job performance were investigated. It was found that university requirements related to the use of ICT and the suitability of ICT for university teachers' work were critical factors affecting their job performance. In addition, a comparison was made among university teachers from different grade levels, revealing that university management related to ICT use tended to affect university teachers of higher-grade levels more than those of lower-grade levels in generating technostress.

Keywords: technostress, multidimensional person-environment misfit, higher education, university teachers, information and communication technologies

# INTRODUCTION

Universities worldwide have been advancing their agendas for education enhanced by information and communication technologies (ICT), such as promoting mobile learning, blended learning, and virtual reality-based instruction (Markowitz et al., 2018; Qi, 2019). Admittedly, these efforts are laudable and could potentially benefit learners. However, they could also exert increased pressure on university teachers who are often less technology-savvy than their students (Jena, 2015; Hatlevik and Hatlevik, 2018), but have to constantly adapt themselves to the ever-demanding university requirements related to the use of technologies at work, which is exacerbated by the rapid changes and advancement of ICT. As a result, the incongruence or misfit between universities and teachers may lead to the latter being subjected to technostress, which is defined as a modern maladaptation resulting from the failure to cope with ICT and changing requirements related to the use of ICT (Brod, 1984; Fuglseth and Sørebø, 2014).

Taking the perspective of person-environment (P-E) fit theory, which has been essential in technostress research (Edwards et al., 1998; Ayyagari et al., 2011), technostress is basically the consequence of misfit between a person and the environment surrounding the person. Given that

#### Edited by:

Wong Yau Ho Paul, The Education University of Hong Kong, Hong Kong

#### Reviewed by:

Wan Khairuzzaman Wan Ismail, Sulaiman Al Rajhi Colleges, Saudi Arabia Annie Hu, The Education University of Hong Kong, Hong Kong

#### \*Correspondence:

Xinghua Wang wangxinghua379@gmail.com

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 07 May 2019 Accepted: 18 July 2019 Published: 06 August 2019

#### Citation:

Wang X and Li B (2019) Technostress Among University Teachers in Higher Education: A Study Using Multidimensional Person-Environment Misfit Theory. Front. Psychol. 10:1791. doi: 10.3389/fpsyg.2019.01791

**53**

the person-environment misfit has multiple dimensions (Chuang et al., 2016), technostress is not only related to the ICT that causes it, but it is also related to organizations that set requirements on the use of ICT (Tarafdar et al., 2010) and colleagues within organizations who often have influence on individuals' use of ICT (Avanzi et al., 2018). Consequently, technostress caused by misfits between teachers and multiple dimensions of the university environment could negatively affect university teachers' performance at work, leading to job burnout and even intentions of quitting the job (Al-Fudail and Mellar, 2008; Tarafdar et al., 2015; Pignata et al., 2016).

In spite of negative consequences associated with technostress to university teachers' work, extant research on technostress has primarily been focused on government and business sectors (e.g., Ayyagari et al., 2011; Fuglseth and Sørebø, 2014). Limited research has investigated this issue in education, particularly in higher education settings where university teachers are exposed to varied ICT and rapid advancement of ICT is revolutionizing learning and teaching (Jena, 2015; Ortagus et al., 2018). Moreover, prior research investigating technostress from the P-E fit perspective mostly examined single dimensions of the environment such as organizations or jobs within organizations (e.g., Player et al., 2017), thus producing an incomplete understanding of technostress and subsequently impeding the development of informed solutions of this issue.

In view of the context-dependent characteristic of technostress (Tarafdar et al., 2015) and the multidimensional nature of the environment (Jansen and Kristof-Brown, 2006), it is imperative to examine technostress in higher education from a more comprehensive perspective. Therefore, this study aimed to bridge this gap by developing a multidimensional person-environment misfit framework of technostress, taking into consideration multiple dimensions of the university environment where university teachers are situated.

# THEORETICAL FRAMEWORK

# The Paradox of ICT

The use of ICT in organizations has been fraught with controversy and paradoxes (Tarafdar et al., 2011). With higher education being gradually digitalized, there is little doubt that ICT bring benefits to university teachers' work. ICT such as mobile computing, collaborative software, and learning management system enable teachers to work from anywhere and anytime, access information conveniently, and upgrade their teaching (Jena, 2015; Qi, 2019). On the other hand, ICT also presents challenges to people's physical and psychological well-being and their job performance (Ayyagari et al., 2011). For instance, ICT may drive university teachers to work faster than they can sustain (techno-overload) and invade their personal life (techno-invasion). Frequent changes and upgrades of software and hardware often make university teachers feel incompetent (techno-complexity and technouncertainty). Moreover, fast technological advances such as Massive Online Open Course (MOOCs) may threaten their job security (techno-insecurity). Consequently, university teachers may feel exhausted, anxious, and stressful. This phenomenon is termed as technostress and will likely affect university teachers' job effectiveness (Al-Fudail and Mellar, 2008; Jena, 2015).

# Extant Research on Technostress

Technostress, as a dark side of ICT, is a relatively new and understudied area, in contrast to considerable numbers of topics on benefits associated with ICT to people' work and life (Tarafdar et al., 2010). Currently, research on technostress has been conducted mostly in government and business environments (e.g., Ragu-Nathan et al., 2008; Fuglseth and Sørebø, 2014). Despite the limited number of studies on technostress in the field of education (e.g., Al-Fudail and Mellar, 2008; Joo et al., 2016), the prevalence and severity of this issue in this field, particularly in higher education, may be no less pronounced than that in government and business environments, in view of the huge investment on ambitious agendas of modernizing learning and teaching via ICT (Glenn, 2008).

Nevertheless, there has been a consensus regarding possible consequences of technostress across different fields. Technostress could negatively affect people's productivity and innovation in their tasks, leading to decreased job performance, lower job satisfaction, and higher turnover rates (Ayyagari et al., 2011; Tarafdar et al., 2015). As such, the issue of technostress deserves proper attention from researchers, developers of ICT, and policymakers in organizations, including higher education institutions.

Prior studies investigating technostress have been done mainly from two perspectives: transaction theory of stress (e.g., Fuglseth and Sørebø, 2014) and person-environment fit (e.g., Ayyagari et al., 2011).

Transaction theory of stress describes the issue of stress as a combination of stimulating conditions and individuals' responses to them (Lazarus and Folkman, 1984). According to this theory, the emergence of technostress seems to be a linear process, from stressors and situational factors to strain and outcomes. Studies based on this theory have mostly followed a reductionist approach in which technostress creators and inhibitors are differentiated and singled out to examine this issue (e.g., Ragu-Nathan et al., 2008).

P-E fit theory highlights the extent to which a person and the environment match (Edwards et al., 1998). Compared with the transactional-based approach, P-E fit theory acknowledges the complex characteristics of technostress. It argues that stress neither arises from the person nor the environment alone; instead, it emerges from the interaction of both. In other words, technostress arises when misfit between the person and the environment happens. But, studies following P-E fit theory have often examined the fit/misfit between the person and a single aspect of the environment (Jansen and Kristof-Brown, 2006), such as either organizations, people, or jobs. This is inconsistent with the fact that people are concurrently nested in multiple dimensions of the environment (Chuang et al., 2016).

As a result, studies examining technostress from a single dimension of P-E misfit or treating its emergence as a linear process may generate a limited understanding of technostress and its impact on individuals' job performance and psychological

well-being. Therefore, this study argues for a multidimensional P-E misfit framework to investigate the phenomenon of technostress in higher education. This study proposes that the formation of technostress results from the unbalanced interaction between the person (teacher) and multiple dimensions of the university environment, and that different dimensions of P-E misfits have influence on one another.

# Multidimensional P-E Misfit Framework of Technostress

Contrasting to the conventional P-E fit approach which often focuses on a single aspect of the environment, the multidimensional P-E fit theory highlights the importance of multiple characteristics of the environment (Jansen and Kristof-Brown, 2006). In line with the multidimensional P-E fit theory and the context of this study, we considered three dimensions of P-E fit: person-organization (P-O) fit, person-technology (P-T) fit, and person-people (P-P) fit. "Organizations" (i.e., universities) in P-O fit refers to the management of universities, including demands for teachers to meet university goals such as requirements and regulations, as well as resources available for university teachers to meet these demands, such as technical support, training, and culture (Demerouti et al., 2001; Avanzi et al., 2018). "Technologies" in P-T fit are concerned with varied ICT employed by universities to digitalize teaching, research, and faculty management. "People" in P-P fit refers to colleagues of teachers in universities in this study as supervisors and managers who are often policymakers in universities are subsumed under "organizations" in P-O fit.

Based on the definition of P-E fit theory and research on multidimensional P-E (mis)fit, P-O fit, P-T fit, and P-P fit are defined as the extent to which the person fits organizations, technologies, and colleagues, respectively. Technostress arises when misfits between the person and the multiple dimensions of the environment occur. In line with the above reasoning, P-O misfit, P-T misfit, and P-P misfit form the multidimensional P-E misfit framework of technostress.

While we can analyze technostress as misfits between a person and multiple dimensions of the environment, another important factor is the causes of the misfits. According to Edwards et al. (1998), stress emerges when (a) the environment does not provide sufficient supplies to meet the person's needs; and/or (b) the person's abilities do not meet the environment's demands. As such, a P-E misfit is often investigated in two ways: abilities-demands (A-D) misfit and/or needs-supplies (N-S) misfit (Player et al., 2017).

In addition, P-E fit theory distinguishes from general interactionist models of the person and the environment in that P-E fit theory requires both the person and the environment constructs to be commensurate with each other (Edwards et al., 1998). For instance, A-D misfit on the organizational level (P-O misfit) should involve the comparison between the amount of ICT use in teaching demanded by universities and the amount of ICT use teachers could incorporate in their teaching. Accordingly, abilities-demands and need-supplies in the multidimensional P-E misfit framework of technostress in this study's context are defined in the following ways:

#### Demands of Organizations and Technologies Versus Abilities of the Person

Demands refer to quantitative and qualitative job requirements from universities related to technological use (the organizational level) and requirements of ICT for effective use of them (the technological level). Abilities are concerned with university teachers' skills, aptitudes, and time to meet demands from universities and ICT. When demands exceed university teachers' abilities, the misfit is likely to yield technostress. However, as to the people level misfit in the multidimensional P-E misfit framework, we only focused on N-S interaction as this dimension of misfit only investigated social support from colleagues in the use of ICT at work.

#### Needs of the Person Versus Supplies by Organizations, Technologies, and Colleagues

"Needs" broadly refers to teachers' requirements for universities to support effective use of ICT (the organizational level), teachers' requirements for ICT to assist them in fulfilling job objectives (the technological level), and teachers' requirements for colleagues to help them effectively use ICT at work (the people level).

Supplies are related to resources and support provided by universities to enable teachers to effectively integrate ICT into their work, functions of ICT available in helping teachers achieve job objectives, and social support from colleagues in stimulating effective ICT use. When supplies fall short of university teachers' needs, the misfit tends to generate technostress.

# Operationalization of Multidimensional P-E Misfit Framework of Technostress

In this study, we propose a nested structure by operationalizing the emergence of technostress (A-D misfit and N-S misfit) in the multiple dimensions of person-environment interactions (see **Table 1**), as explained below.

#### A-D Misfit in P-O Misfit (ADO)

The implementation of ICT in universities increases requirements of job scope and skills for university teachers, who are likely to experience higher task difficulty and more ambiguity about performance expectations from universities (Ragu-Nathan et al., 2008), thereby tending to create an incongruence or misfit between abilities of teachers and demands of universities.

#### N-S Misfit in P-O Misfit (NSO)

As integrating ICT in teaching often necessitates new and higher skills from university teachers and changes of work processes (Tarafdar et al., 2010), university support is essential in preparing teachers to adapt to the changes. Support, such as technical support and professional training, is considered important in helping university teachers integrate ICT into their work, while insufficient university support is likely to intensify teachers' stress during ICT integration (Joo et al., 2016).

# A-D Misfit in P-T Misfit (ADT)

Due to constant updates of ICT hardware and software and their increasing complexities, university teachers' skills are subject to be devaluated frequently. In addition, university teachers may

TABLE 1 | Nested structure of multidimensional P-E misfit framework of technostress.


also feel inundated by vast amounts of information from multiple ICT such as learning management systems, social media, and staff management systems. They are compelled to work faster to deal with increased processing demands, consequently creating a gap between abilities of teachers and demands of ICT for a better use of them (Hogan and McKnight, 2007; Ragu-Nathan et al., 2008).

#### N-S Misfit in P-T Misfit (NST)

ICT often needs to be reconfigured and customized before being applied to university teachers' daily work. However, modifications of ICT often lead to problems such as system crashes, data loss, and inadequate technical resources (Ragu-Nathan et al., 2008). Consequently, the ICT to be used may fall short of meeting university teachers' needs. Furthermore, because of the lack of teachers' involvement during ICT purchase and implementation phases, the ICT may turn out to be unable to match teachers' work requirements (Tarafdar et al., 2010). This not only causes unnecessary financial waste, but also negatively affects university teachers' job performance.

#### P-P Misfit (PPF)

Person-people misfit is mainly concerned with the lack of support from colleagues in the use of ICT at work. Colleagues' social support constitutes an important resource to deal with technostress at work (Avanzi et al., 2018). It can induce positive affect among university teachers, as they perceive that they do not cope with high work demands alone. However, the lack of colleagues' social support tends to accentuate people's feelings of helplessness, especially in the face of work challenges (Al-Fudail and Mellar, 2008).

# Hypothesis Development

Overall, this study was guided by the following two research questions: (a) What are the relationships among the multidimensional P-E misfits of technostress? and (b) How do the multidimensional P-E misfits of technostress affect university teachers' job performance? To help answer the research questions, the hypotheses related to the multidimensional P-E misfits of technostress and job performance would be developed.

Organizational management related to ICT directly determines what ICT are needed and how ICT are used in organizations (Tarafdar et al., 2010). Inappropriate organizational management may cause the introduction of unsuitable ICT and/or improper implementation of ICT (Al-Fudail and Mellar, 2008), which may end up either being too complex for teachers to handle or being insufficient to satisfy their needs for teaching.

Organizational management also affects working relationships among university teachers in organizations (Hayton et al., 2012). Effective organizational management can foster prosocial behaviors and increase teamwork, thus increasing social support among university teachers (Avanzi et al., 2018). However, improper organizational management likely distances university teachers from one another and decreases their chances of obtaining sufficient help from colleagues to cope with challenges at work (Luchman and González-Morales, 2013). Thus, we hypothesized that P-O misfit (containing ADO and NSO) is fundamental to P-T misfit (containing ADT and NST) and P-P misfit in the present study. As such, the following hypotheses were developed:


Job performance in this study refers to the degree to which ICT helps university teachers complete their job requirements (Tarafdar et al., 2010). As ICT gradually becomes an integrated part of organizations including higher education institutions, they have an increasing influence on university teachers' performance and the success of organizations. Furthermore, university teachers' job performance plays a significant role in preparing students for the future workforce and is essential in maintaining universities' competitiveness in the global education market (Jena, 2015).

Based on extant research (Tarafdar et al., 2011; Chuang et al., 2016), the fit between university teachers and multiple dimensions of the higher education environment leads to improved job performance, while the misfit results in otherwise. Understanding how university teachers' job performance is affected by the multiple dimensions of the environment is important in addressing technostress in higher

education and in better utilizing benefits associated with ICT. Therefore, we hypothesized that the multidimensional P-E misfits of technostress negatively affect university teachers' job performance.


Based on the analyses above, the research model of this study was developed and is illustrated in **Figure 1**.

# METHODOLOGY

#### Participants

The participants of the study were sampled from five public universities in mainland China. To respond to the call from the Minister of Education (MOE) of China to modernize learning and teaching via ICT (MOE of PRC, 2015), the five public universities, among many others in the country, have been advancing agendas such as digitalizing curricula via MOOC, promoting mobile learning, and experimenting with ICT-enhanced flipped classrooms. To accomplish the agendas, teachers in the universities are required to learn to create video and audio teaching resources of different lengths, utilize learning management systems for routine work, and adjust their teaching practices in some courses toward learner-centered and ICT-enhanced pedagogy. Consequently, many teachers in the universities complain about the technostress generated as a result of these agendas.

To facilitate the implementation of these agendas in the public universities in the country, the MOE sponsored a series of professional development programs aiming to increase university teachers' capabilities of ICT-enhanced learning and teaching. The participants of this study who were from the five public universities were assigned in one of the programs. Among the 400 participants approached, with their informed consent valid responses from 343 participants were obtained, with a response rate of 85.75%. Their demographic information is presented in **Table 2**.



#### Instrument Development and Validation

The initial survey instrument contained six constructs with 37 items (see **Appendix A**) and was rated on a five-point Likert scale, with 1 representing Strongly Disagree and 5 Strongly Agree with the items.

In the initial survey, 31 items were developed to operationalize multidimensional P-E misfits of technostress, comprising of ADO, NSO, ADT, NST, and PPF, based on Edwards et al. (1998), Jansen and Kristof-Brown (2006), Chuang et al. (2016). Specifically, abilities-demand misfit on the organizational level (ADO) was originally described by eight items, for instance, "I find it difficult to meet the high demands of school policies regarding the use of ICT at work" and "I find it hard to adjust my current work pattern so as to comply with school policies regarding the use of ICT at work." Needs-supplies misfit on the organizational level (NSO) contained five items, for example, "My school does not provide me with sufficient professional training to effectively use ICT at work," and "My school does not provide me with sufficient incentives to effectively use ICT at work." Abilities-demands misfit on the technological level (ADT) had seven items, such as "I find it difficult to effectively use ICT due to my limited investment of time and effort," and "I find it difficult to cope with the high demands of ICT with my current capability." Needs-supplies misfit on the technological level (NST) contained seven items, for instance, "The ICT in my school are not effective in helping me increase my productivity at work," and "The ICT in my school are not very relevant for the improvement of my work." Person-people misfit (PPF) was described by four items, such as "I do not have sufficient support from my colleagues for the use of ICT at work," and "My colleagues are not encouraging with regard to the innovative use of ICT at work."

The construct of job performance was adapted from Tarafdar et al. (2010) and contained six items, such as "The ICT in my school improve the quality of my work" and "The ICT in my school enhance my work productivity." The reported Cronbach's alpha of job performance was 0.91.

To ensure the clarity of the original survey items, three participants were invited to check their understanding and the wording of the survey, which was subsequently reworded to deliver clearer ideas related to technostress. Furthermore, to improve the face validity of the original survey, two experts in stress research were approached to obtain their comments on the survey items, which was refined accordingly. In addition, because the original survey was in English, we conducted a back-translation to ensure that there was minimal difference between the English and Chinese versions.

Given that the five constructs (ADO, NSO, ADT, NST, and PPF) measuring multidimensional P-E misfits of technostress were self-developed, it was necessary to examine the internal consistency and validity of this technostress model before proceeding to further analyses. A number of rounds of exploratory factor analysis (EFA) were first performed on the technostress model consisting of the five constructs to extract the preliminary factor structure, which was then examined via confirmatory factor analysis (CFA; see **Appendix B** for detailed criteria of performing EFA and CFA and related results). In the end, the refined technostress model comprising the five constructs with 22 items was attained. Cronbach's alpha values for ADO, NSO, ADT, NST, and PPF were 0.90, 0.89, 0.79, 0.86, and 0.86, respectively. The refined technostress model obtained good model fit (χ 2 /df = 2.06, CFI = 0.95, NFI = 0.91, and RMSEA = 0.06).

Given that this study utilized self-report data, Harman's single-factor test (Podsakoff et al., 2012) was carried out to examine possible common method bias. After entering all variables into an EFA to investigate the unrotated factor solution, the total variance explained by a single factor was 33.29%, which was considerably lower than 50%. This suggests that there is no significant amount of common method bias existing in the data.

# Data Analysis

PLS-SEM was used to analyze the research model as illustrated in **Figure 1**. As a variance-based structural equation modeling, PLS-SEM is prediction-oriented and exploratory with the aim of maximizing the variance explained for the dependent variables (Chin, 1998; Hair et al., 2014). This study's use of PLS-SEM was mainly based on the following two reasons. First, PLS-SEM works effectively with small sample sizes. Because the PLS-SEM algorithm analyzes portions of the PLS-SEM model one by one and iteratively, it is viable if the sample size is sufficient to calculate the single largest regression equation in the model rather than the whole model (Willaby et al., 2015). Second, PLS-SEM is good at prediction and theory development. It can generate stronger or equivalent statistical power than covariance-based structural equation modeling with smaller sample sizes (Reinartz et al., 2009; Astrachan et al., 2014). Considering that this study endeavored to develop a multidimensional P-E misfit framework of technostress and is exploratory in nature, PLS-SEM fits well with the aim of the present study. The PLS-SEM package (Sanchez, 2013) in the R programing language was used for data analysis.

The research model in this study was analyzed via PLS-SEM following two steps: the measurement model and the structural model (Hair et al., 2014). Then, the whole dataset was split based on university teachers' grade levels as higher-grade courses have more complicated knowledge structures and those teaching higher-grade courses are often senior teachers, thus being more likely to be subjected to technostress (McIver et al., 2016). Subsequently, multi-group comparisons were carried out on the sub-datasets to examine whether there were any possible differences regarding technostress among university teachers of different grade levels, the knowledge of which could inform the development of targeted countermeasures against technostress.

Considering that we categorized five grade levels of teaching (Year 1, Year 2, Year 3, Year 4, and Postgraduate schools) and there were very unequal numbers of participants in different grade levels, we combined participants of Year 1 and Year 2 and labeled them as Lower-grade levels (N = 207). Participants of Year 3, Year 4, and Postgraduate studies were added up and labeled as Higher-grade levels (N = 136).

# Measurement Model

The measurement model of this study was investigated from the following criteria: (a) item reliability, (b) convergent validity, and (c) discriminant validity.

#### Item Reliability

Item reliability is assessed by examining the loadings of survey items with their respective latent construct. The standardized loadings of the items should exceed 0.70 (Hair et al., 2014). **Table 3** indicates that the loadings of all items satisfied the requirement.

#### Convergent Validity

This criterion investigates the extent to which survey items that are theoretically related to one another are related in practice (Hair et al., 2011). Convergent validity is examined by checking (a) internal consistency and (b) average variance extracted (AVE) of each latent construct (Fornell and Larcker, 1981). Internal consistency of a given latent construct is assessed through composite reliability. An internally consistent model should have composite reliability of above 0.70 (Nunnally, 1978). As shown in **Table 3**, the research model met the requirement. As to AVE, the minimum AVE value of 0.50 suggests that at least 50% of the variance of the indicators is explained (Hair et al., 2011). As indicated in **Table 3**, AVEs of the latent constructs in the research model satisfied the requirement. Taken together, the convergent validity of the research model in this study was substantiated.

#### Discriminant Validity

The discriminant validity of the research model was examined in two aspects: (a) the square root of the AVE for each latent construct shall exceed the correlation coefficients between that and other latent constructs (Chin, 1998); and (b) survey items should load more on the latent constructs that they aim to measure than on other latent constructs (Chin et al., 2003). As TABLE 3 | Cronbach's alpha, composite reliability, average variance extracted (AVE), and factor loadings of the constructs and items in the research model (N = 343).


shown in **Table 4** and **Appendix C**, the two requirements of discriminant validity were supported.

# Structural Model

This study's structural model was assessed through path coefficients' significance levels and explanatory power (i.e., R 2 ) of endogenous constructs. **Figure 2** illustrates the validation outcomes of the structural model in this study. Bootstrapping analyses were utilized to examine the statistical significance of the path coefficients in the structural model as PLS-SEM does not rely on distributional assumptions and thus the significance levels

#### TABLE 4 | Discriminant validity of the research model (N = 343).


The bold values in the diagonal row are the square roots of the average variance extracted for the constructs in the research model.

are not suitable to be examined through parametric approaches (Hair et al., 2014).

**Table 5** presents the bootstrapping validation outcomes. Overall, P-E misfit on the organizational level (P-O misfit: ADO and NSO) largely served as the fundamental misfit, underlying the functioning of P-E misfit on the technological (P-T misfit: ADT and NST) and people levels (P-P misfit). Specifically, ADO significantly predicted ADT, NST, and PPF, thereby supporting H1, H2, and H3. NSO significantly predicted NST and PPF, except for ADT; thus, H5 and H6 were substantiated while H4 was not. With regards to the effect of multidimensional P-E misfits of technostress on university teachers' job performance, ADO and NST demonstrated strong negative effects. As such, H7 and H10 were supported. However, NSO, ADT, and PPF did not exert significantly negative effects on job performance; therefore, H8, H9, and H11 were not supported.

As PLS-SEM aims to maximize the variance explained in endogenous constructs, R 2 values of endogenous constructs are viewed as the primary criterion for assessing the quality of structural models (Henseler et al., 2009). However, due to the lack of generally agreed-upon values of R 2 , this study followed the research of Cohen (1988) on R2. Cohen (1988) pointed out that R 2 values of 0.02, 0.13, and 0.26 imply small, medium, and large effect sizes, respectively. As shown in **Figure 2**, the R 2 values of ADT, NST, and PPF were 0.40, 0.47, and 0.41, respectively, indicating substantial explanatory power. The R 2 value of JP was 0.17, implying a moderate explanatory power. On the whole, the predictive power of this study's model was acceptable.

A global criterion of goodness-of-fit (0 < GoF < 1) has been developed by Tenenhaus et al. (2004) to assess the overall quality of PLS-SEM analyses. It is computed as the geometric mean of the average communality and average R 2 . The GoF values of 0.10, 0.25, and 0.36 are defined as small, medium, and large, respectively (Wetzels et al., 2009). The research model's GoF value in this study was 0.51, which was considerably large. In sum, the reliability and validity of the proposed research model in this study were confirmed and acceptable according to the analyses above.

# Multi-Group Comparison Based on Grade Levels of Teaching

Multi-group comparison is conducted by comparing differences at structural levels of research models. Specifically, path coefficients of research models based on different groups of participants are examined (Sanchez, 2013; Hair et al., 2014), as the aim of path modeling with latent constructs is to estimate linear relationships among the constructs. The approach of bootstrap t-test was used by following a three-step procedure: (a) the whole dataset is split into groups; (b) bootstrap samples are carried out with replacements for each group; (c) subsamples are compared through t-tests in terms of standard error estimates of path coefficients.

**Table 6** reveals that there were significant differences between university teachers of different grade levels regarding (a) the path coefficient of ADO on ADT, indicating that ADO exerted a greater influence on ADT for university teachers of higher-grade levels than those of lower-grade levels; and (b) the path coefficient of NSO on PPF, implying that NSO had a more significant effect on PPF for teachers of higher-grade levels than those of lowergrade levels.

# DISCUSSION AND CONCLUSION

This study investigated the phenomenon of technostress among university teachers in higher education. A research model containing a multidimensional P-E misfit framework of technostress was proposed to examine technostress from a more comprehensive perspective: person-organization (P-O) misfit, person-technology (P-T) misfit, and person-people (P-P) misfit, and how the multidimensional P-E misfits of technostress negatively affected university teachers' job performance.

The findings indicate that the proposed research model demonstrated high reliability and validity. P-O misfit basically played a fundamental role, greatly predicting P-T misfit [with the R 2 values of 0.40 and 0.47 for abilities-demands misfit (ADT) and needs-supplies misfit on the technological level (NST), respectively] and P-P misfit [with the R 2 value of 0.41 for person-people misfit (PPF); see **Figure 2**]. This implies that organizational management, including organizational demands of ICT use and supplies to university teachers to meet the demands, largely determines the emergence of technostress and thus, is essential in preserving university teachers' well-being. In this regard, although technostress is associated with technologies TABLE 5 | Bootstrap validation outcomes for the research model (N = 343).


∗∗p < 0.01; ∗∗∗p < 0.001; ADO, abilities-demands misfit (in P-O misfit); NSO, needs-supplies misfit (in P-O misfit); ADT, abilities-demands misfit (in P-T misfit); NST, needs-supplies misfit (in P-T misfit); PPF, person-people misfit (P-P misfit); JP, job performance; #The bootstrap analysis did not support the statistical significance of the path coefficient (see Appendix D) due to the possible risk of overfitting issue; The bold rows highlight the hypotheses that were supported in this study.

TABLE 6 | Comparison between teachers of lower- and higher-grades (N = 343).


ADO, abilities-demands misfit (in P-O misfit); NSO, needs-supplies misfit (in P-O misfit); ADT, abilities-demands misfit (in P-T misfit); NST, needs-supplies misfit (in P-T misfit); PPF, Person-people misfit (P-P misfit); JP, job performance; diff.abs, absolute difference; #The comparisons make sense only when the global path coefficients are statistically significant; The bold row indicates the paths where teachers of lower-grades significantly differed from those of higher-grades.

in the first place, it is more likely to be caused by organizational management related to technological use.

Moreover, among the five constructs [abilities-demands misfit on the organizational level (ADO), needs-supplies misfit on the organizational level (NSO), abilities-demands misfit on the technological level (ADT), needs-supplies on the technological level (NST), and person-people misfit (PPF)] in the framework of technostress, ADO (with a path coefficient of −0.21) and NST (with a path coefficient of −0.35; see **Table 5**) had a significantly negative influence on university teachers' job performance. This suggests that, when keeping university teachers' abilities and needs constant, organizational demands of ICT use and suitability of ICT for teachers' work are crucial factors in determining the effectiveness of teachers' work.

With regards to the comparison between university teachers of lower and higher grades, ADO had a more significant effect on ADT for university teachers of higher-grade than those of lower-grade; NSO also exerted a greater influence on PPF for university teachers of higher-grade than those of lower-grade. This implies that for higher-grade university teachers, the impact of organizational management on other dimensional misfits is more significant, thereby contributing more to technostress.

For the sake of delivering coherent themes based on the statistical findings, hypotheses of this study are better discussed collectively, instead of one by one. Overall, four themes were generated, as follows:

(1) P-O misfit, P-T misfit, and P-P misfit may largely capture how university teachers interact with the multiple dimensions of the higher education environment in imbalanced ways that generate technostress, with P-O misfit underlying the functioning of the other two misfits.

The validated multidimensional P-E misfit framework of technostress suggests that the use of P-O (person-organization) misfit, P-T (person-technology) misfit, and P-P (person-people) misfit to describe the formation of technostress is largely aligned with incongruent interactions between university teachers and the multiple dimensions of the higher education environment (Chuang et al., 2016). Although technostress is a type of stress related to ICT, it is also associated with universities that require ICT-enhanced learning and teaching (Ragu-Nathan et al., 2008), as teachers themselves usually do not have sufficient motivation to volunteer to change their established teaching practices by integrating ICT (Joo et al., 2016).

Furthermore, it is also associated with the availability of support from colleagues in universities, which is considered a coping strategy and one of the most important resources for employees to deal with work stress (Halbesleben, 2006). Insufficient support from colleagues may cause great stress when job demands are high (Pignata et al., 2016; Avanzi et al., 2018), as it decreases university teachers' confidence and abilities to cope with challenges related to the use of ICT.

In addition, universities are structured organizations with a clear hierarchy of management and collective goals, as have been identified in other occupational settings (Avanzi et al., 2018). Thus, the fundamental role played by P-O misfit in triggering P-T misfit and P-P misfit is largely in line with the characteristic of universities as a form of organization, as evidenced by Hypotheses 1–6 except for H4 (see **Figure 2** and **Table 5**).

Finally, with regard to the effect size of job performance (R <sup>2</sup> = 0.17) in the whole research model, it is understandable that the multidimensional P-E misfits of technostress negatively contributed a moderate amount of variance in job performance. Although ICT are considered essential to modernization agendas of many universities and play an increasingly important role in teachers' work, they have not yet dominated teachers' work. As such, the R 2 value of 0.17 of job performance was already sufficient to consider technostress hazardous.

(2) Requirements of universities related to ICT use at work are critical in affecting university teachers' job performance.

As shown in **Figure 2** and **Table 5**, within the multidimensional P-E misfits, ADO (abilities-demands misfit on the organizational level) was found to exert a significantly negative influence on university teachers' job performance (H7) while ADT (abilities-demands misfit on the technological level) did not (H9). This may be due to the fact that while being able to use ICT is one thing, being able to satisfy universities' requirements related to ICT use at work may be another. This is because ICT-enhanced learning, such as flipped classroom, mobile learning, and blended learning, involves new teaching paradigms and philosophies which require changes in university teachers' roles in classroom, content delivery, pedagogical design, and assessment (Hogan and McKnight, 2007), instead of simply possessing or applying ICT knowledge. Consequently, teachers may struggle to meet universities' requirements of integrating ICT into work in spite of their mastery of ICT.

However, despite the fundamental role played by P-O (person-organization) misfit in the multidimensional P-E misfit framework, NSO (needs-supplies misfit on the organizational level) in P-O misfit failed to predict ADT (abilities-demands misfit on the technological level; H4) and did not significantly jeopardize university teachers' job performance (H8). These findings about NSO may seem surprising, as previous studies consider university support imperative in preparing teachers to integrate ICT into work (e.g., Al-Fudail and Mellar, 2008; Luchman and González-Morales, 2013). This may be due to the fact that teachers often avoid university supplies such as professional development workshops and technical seminars as they do not want to be overwhelmed by various computing requirements or made to feel ignorant (Shedletsky and Aitken, 2001). Moreover, the ICT-enhanced learning paradigm often consumes more time and effort than onsite teaching (Hogan and McKnight, 2007; Joo et al., 2016) and has few relations to university teachers' promotion, tenure, and salary (Glenn, 2008). Consequently, teachers tend to perceive the university supplies limitedly relevant to their career development (Shedletsky and Aitken, 2001). Therefore, even if needs-supplies misfit occurs on the organizational level (NSO), university teachers' job performance is not likely to be significantly compromised. Likewise, whether university teachers are capable of utilizing ICT has little relevance to NSO.

(3) Suitability of ICT for university teachers' work is significantly important to their job performance.

As revealed in **Figure 2** and **Table 5**, NST in P-T (needssupplies misfit on the technological level) misfit exerted a significantly negative effect on university teachers' job performance (H10), while ADT (abilities-demands misfit on the technological level) and PPF (person-people misfit) did not (H9 and H11). This may be because university teachers in higher education are often professionals in their fields and tend to have higher capabilities in developing their ICT knowledge and skills (Tarafdar et al., 2011). Even if acquiring mastery of ICT involves steep learning curves and/or sufficient social support from colleagues is not available, teachers could spend more time and effort to master the ICT by themselves. As a result,

even if ADT and PPF may affect university teachers' work (Tarafdar et al., 2010; Avanzi et al., 2018), they are not likely to significantly undermine their job performance. Nevertheless, once the ICT available to university teachers cannot satisfy their work needs, the resultant needs-supplies misfit on the technological level (NST) would easily cause a significantly negative effect on university teachers' job performance, regardless of how capable teachers are. Specifically, the ICT should have necessary functional properties that are suitable for and support university teachers' work, instead of setting limits for their work. For instance, if a learning management system stipulates a variety of requirements of formats and properties of teaching resources that can be uploaded, it will restrict university teachers' choices of course design and implementation, eventually negatively affecting their job performance.

(4) University teachers teaching higher-grade classes are more likely to be significantly affected by universities' requirements and supplies related to the use of ICT.

The comparison outcomes between university teachers of lower and higher grades (see **Table 6**) can be discussed from characteristics of senior and junior teachers, as grade levels of teaching often represent different levels of skillsets and abilities; in the context of this study, senior teachers often teach and supervise higher-grade students.

While ICT are revolutionizing education of all kinds, policymakers in educational systems are placing increasing levels of pressure on teachers to digitalize their work (Orlando, 2014). Senior teachers who have taken years in establishing their teaching practices are under greater burdens to change them than their younger counterparts who are mostly in the process of developing their teaching practices (Day and Gu, 2009; Orlando, 2014). Consequently, senior teachers may find it more difficult to cope with the integration of ICT into their work.

As organizational supplies such as professional training and technical support are often of general types and apply to every university teacher (Shedletsky and Aitken, 2001), they may be eventually insufficient and improper for teachers of higher grades who need specialized training programs and support to digitalize their established traditional teaching practices. In addition, higher-grade courses often involve more complex knowledge structures (McIver et al., 2016). Thus, integrating ICT into higher-grade courses tend to involve greater workloads and may be more challenging for these teachers. As a result, the unsupportive work environment is more likely to reduce senior teachers' possibilities of attaining their desired help from colleagues to deal with challenges of adjusting to the ICTenhanced learning paradigm (Luchman and González-Morales, 2013), thereby leading to higher possibilities of P-P misfit.

#### Contributions

This study has the following contributions. First, this study developed a multidimensional P-E misfit framework to investigate technostress among university teachers in higher education. The multidimensional P-E misfits consist of P-O misfit, P-T misfit, and P-P misfit, among which P-O misfit holds a fundamental position and underlies the functioning of the other two dimensions of misfits. In view of the dearth of studies on technostress in higher education settings and the reductionist approach utilized by extant studies in examining this issue, the multidimensional P-E misfit framework of technostress provides a more comprehensive theoretical perspective in understanding the formation of technostress.

Second, to facilitate future research on technostress in higher education, this study also designed and validated an instrument to measure technostress from multiple dimensions. The instrument has demonstrated high psychometric properties and can be utilized in future theoretical and empirical efforts for fuller knowledge of technostress. In the finalized multidimensional P-E misfit scale of technostress with 22 items (the total score ranging from 22 to 110), the higher the score, the greater the levels of technostress. Specifically, a score of 22 indicates the absence of technostress; 23–65 corresponds to a mild level of technostress; 66–87 implies a moderate level of technostress; and ≥ 88 corresponds to a severe level of technostress.

Third, this study revealed differences among university teachers from different grade levels in the causes of technostress through the multidimensional P-E misfit framework, thus highlighting possible differences between senior and junior teachers in their experiences of technostress and possibilities of suffering from technostress due to varying workloads associated with integrating ICT in different grade levels of courses.

Fourth, among the A-D and N-S misfits in the multidimensional P-E misfit framework of technostress, the A-D misfit on the organizational level (ADO) and the N-S misfit on the technological level (NST) were identified to have significantly negative effects on university teachers' job performance. Therefore, these findings could effectively inform the development of targeted strategies for addressing the issue of technostress among teachers in higher education.

#### Implications

This study carries the following implications for future research and practice on technostress in higher education. First, in the multidimensional P-E misfit framework developed in this study, organizational management plays an important and fundamental role in the formation of technostress, which is more likely to arise when organizational demands are high and organizational supplies are insufficient (Griffiths, 2014). Organizational management largely determines what ICT should be introduced, how ICT should be implemented, and the availability of support university teachers may get from colleagues (Tarafdar et al., 2010; Avanzi et al., 2018). In addition, ADO and NST exerted significant negative effects on university teachers' job performance. Therefore, the priority of developing countermeasures against technostress should start from adjusting university management related to the use of ICT.

Universities with cases of teachers reporting the issue of technostress should reconsider their primary objectives of ICT use and align university objectives with ICT affordances and teachers' real needs related to ICT. Teachers should be involved in ICT planning, purchase, and implementation phases

(Tarafdar et al., 2010). Universities can identify teachers' real needs of ICT for their work if teachers are allowed to participate in discussions related to ICT adoption. Also, incorporating teachers' real needs and suggestions into universities' decisionmaking from the beginning can help improve the match between university demands and what teachers can provide and between teachers' needs and what ICT can provide, thereby decreasing possible technostress yielding from the mismatches and improving university teachers' job performance.

Second, as university teachers of higher-grade levels are more likely to be negatively affected by improper organizational management related to ICT use, universities advancing their digitization agendas should not adopt the one-size-fits-all strategies. Instead, considering that university teachers of highergrade levels have often spent years developing their work practices and higher-grade courses often involve more complex knowledge structures, university demands of ICT use for them should be modest and university supplies should be tailored specifically to their characteristics.

Third, as technostress is basically a psychological reaction (Ragu-Nathan et al., 2008), it is imperative that teachers themselves learn to regulate their emotional and psychological responses to external challenges. More importantly, they should also seek to develop their capabilities and skills to effectively cope with challenges associated with ICT-enhanced learning and teaching paradigm.

#### Limitations and Future Research

Certain limitations should be considered when interpreting this study's findings. First, this study adopted a cross-sectional design, thus making it difficult to attain causal relationships among constructs related to technostress. To strengthen the explanatory power of the multidimensional P-E misfit framework of technostress, a longitudinal design is necessary in future research. Second, in view of possible influence of different cultures on technostress, the generalization of the study's findings should be cautious. Future research is suggested to validate the multidimensional P-E misfit framework of technostress in Western universities to obtain a more comprehensive understanding of technostress in higher education. Third,

# REFERENCES


the participants of this study may not optimally represent university teachers in higher education, as the number of female participants doubles that of males. Researchers in the future are advised to balance the gender distribution of participants to further validate the findings of this study. Fourth, this study's findings were attained through self-report data, thereby subjecting to report errors related to human perceptions. As such, future studies may consider triangulating self-report data with possible clinical diagnostic tests of technostress and indepth interviews with university teachers, administers, and other stakeholders in higher education to provide stronger arguments for the phenomenon of technostress and inform the development of solution strategies.

# DATA AVAILABILITY

The datasets generated for this study are available on request to the corresponding author.

# ETHICS STATEMENT

This study was conducted in accordance with the recommendations of the Institutional Review Board (IRB), Nanyang Technological University with written informed consent from all participants.

# AUTHOR CONTRIBUTIONS

XW designed the study, analyzed the data, and wrote the manuscript. BL contributed to the design of the study and data collection.

# SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyg. 2019.01791/full#supplementary-material

for Business Research, ed. G. A. Marcoulides (Mahwah, NJ: Lawrence Erlbaum Associates Publishers), 295–336.



**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Wang and Li. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Violence at School and the Well-Being of Teachers. The Importance of Positive Relationships

#### Federica De Cordova, Sabrina Berlanda\*, Monica Pedrazza and Marta Fraizzoli

Department of Human Sciences, University of Verona, Verona, Italy

Edited by: Caterina Fiorilli,

LUMSA University, Italy

#### Reviewed by:

Cristina Zucchermaglio, Sapienza University of Rome, Italy Elrie Botha, North-West University, South Africa

> \*Correspondence: Sabrina Berlanda sabrina.berlanda@univr.it

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 06 June 2019 Accepted: 22 July 2019 Published: 07 August 2019

#### Citation:

De Cordova F, Berlanda S, Pedrazza M and Fraizzoli M (2019) Violence at School and the Well-Being of Teachers. The Importance of Positive Relationships. Front. Psychol. 10:1807. doi: 10.3389/fpsyg.2019.01807 In the discipline of positive psychology, "well-being" is considered a fundamental aspect of "human flourishing." Inherent to this multidimensional model are emotional, psychological, and social forms of well-being, which can be grounded in positive relationships in the work environment. By adopting an interpretive framework that emphasizes these dynamics, researchers are able to focus on elements that actively help sustain the process of flourishing, rather than on negative environmental features that should be avoided if possible. Within this broader discussion, the specific question of teachers' well-being is one that has global relevance and merits greater academic attention. After all, it has significant consequences for the educational environment and students' ability to learn. The literature suggests that teachers are increasingly exposed to violence on the part of students and/or their parents. Inappropriate and/or aggressive behavior like this can reduce a teacher's occupational well-being and make it more difficult to build positive relationships in the classroom. Ultimately, it is one of the most serious work-related stress factors affecting the profession today. Previous studies have attempted to characterize the contexts in which violence occurs, and its negative impact on both the individuals involved and the broader educational climate. Less attention has been devoted to the capacity of teachers to deal with violence and develop a more resilient mindset. The positive psychology perspective focuses on well-being as a multidimensional construct wherein psychosocial and physical comfort does not simply arise provided there is an absence of suffering and violence. Rather, according to this model, such experiences can be counteracted by a capacity to endure and build positive environments. With these considerations in mind, our study presents data gathered in primary and secondary schools in northern Italy. A total 475 teachers completed an online, self-report questionnaire. The results indicate that teachers can experience occupational well-being even if they are subjected to aggressive behaviors. Supportive leadership and good relationships with colleagues may be considered valuable resources for fostering well-being among teachers.

Keywords: positive psychology, well-being, teacher, violence at school, social support

# INTRODUCTION

fpsyg-10-01807 August 5, 2019 Time: 13:53 # 2

Taking their lead from the field of positive psychology, for nearly 20 years, academics have been working to understand the concept of human well-being in terms of "human flourishing" (Fredrickson and Losada, 2005). This multidimensional model of positive functioning incorporates elements of emotional, psychological and social well-being and seeks to pinpoint those factors that help sustain the process of flourishing, in contrast to approaches that highlight negative environmental features that should be avoided where possible (Seligman and Csikszentmihalyi, 2000). Indeed, "psychology should be able to help document what [. . .] work settings support the greatest satisfaction among workers [. . .] and how our lives can be most worth living. Yet psychologists have scant knowledge of what makes life worth living. They have come to understand quite a bit about how people survive and endure under conditions of adversity. [. . .] This almost exclusive attention to pathology neglects the fulfilled individual and the thriving community. The aim of positive psychology is to begin to catalyze a change in the focus of psychology from preoccupation only with repairing the worst things in life to also building positive qualities" (Seligman and Csikszentmihalyi, 2000, p. 5). This sort of inversion in the theoretical standpoint seems, to us, to offer a particularly fruitful way to approach the issue of well-being among teachers, which has arisen as a common concern around the world, given that it has a bearing on the educational environment and students' learning skills (Zeinabadi, 2010; Skaalvik and Skaalvik, 2011a,b; McMahon et al., 2014; Benevene et al., 2018; Liu et al., 2018).

Educational contexts are emotionally demanding (Buonomo et al., 2017; Pedrazza et al., 2018). Teachers, in particular, are frequently faced with significant challenges and a lack of adequate resources, both material and immaterial (Hakanen et al., 2006; Bakker and Demerouti, 2017). In such conditions, the teacherstudent relationship has a significant bearing on the risk of burnout. In many cultures, teachers are expected both to act as models for their students and to provide them with some form of protection, in a relationship of care. Failure to meet these key duties (Aldrup et al., 2018) is likely to cause stress and harm teachers' well-being. Recent studies have identified the phenomenon of violence inflicted on teachers in the school setting by students and/or their parents as a serious issue. Indeed, misbehavior and aggression on the part of students can be considered one of the most serious work-related stress factors for the teaching profession, one that is capable of seriously reducing occupational well-being among teachers and preventing them from building positive relationships in the classroom. Violence against teachers at school is therefore an issue with ramifications for societies all over the world, but it remains an under-researched phenomenon. Teacher-directed violence is not defined consistently in the literature (McMahon et al., 2017; Reddy et al., 2018). Existing studies have attempted to characterize the contexts in which violence occurs, and its negative impact on both the individuals involved and the broader educational climate. Less attention has been devoted to the capacity of individuals and groups to deal with violence at school and develop a more resilient mindset, or to describing how such experiences can be endured and counteracted by building a more positive environment. However, if possessing "positive resources" can protect teachers from risk factors and enhance their ability to cope with work-related stressors (Buonomo et al., 2017), it is worth investigating which factors can contribute to and maintain well-being among teachers, even in the face of violence in the school setting.

Looking at the international literature, there are data indicating that 80% of teachers in the United States reported at least one form of victimization at the workplace during the same or previous year in which they completed the research survey (Galand et al., 2007; Espelage et al., 2013; McMahon et al., 2014; Longobardi et al., 2018; Reddy et al., 2018). Although partial, this picture reflects a growth in numbers of aggressive acts directed toward teachers, across different school types and locations and in forms ranging from verbal threats or intimidation to acts of physical and/or sexual violence (Dinkes et al., 2009; Wilson et al., 2011).

Violence is a multidimensional phenomenon produced through various isolated or simultaneous behaviors (Jevtic et al., 2014; Padurariu et al., 2016; Berlanda et al., 2019), which can operate at one or more different levels from the physical to the emotional, psychological and relational. Whereas physical aggression involves the infliction of physical harm, relational aggression is typically characterized by verbal abuse (e.g., shouting and screaming), spreading rumors, and exclusionary behaviors (Lansford et al., 2012). Some researchers have used the terms "bullying" and "victimization" to cover these forms of abuse and violence (Kauppi and Pörhölä, 2012; Anderman et al., 2018); others have applied terms such as "harassment" (Kauppi and Pörhölä, 2009) and "violence against teachers" (Chen and Astor, 2008). Regardless of the exact terminology and categorization of the violence – "student violence," "school violence," and "workplace violence," etc. (Kauppi and Pörhölä, 2012; Skåland, 2016; Anderman et al., 2018) – a picture has emerged in which a significant number of teachers are victimized by their students (Dzuka and Dalbert, 2007; Chen and Astor, 2008; Kauppi and Pörhölä, 2012), with similar features and outcomes all over the world (Van de Vliert et al., 2013; Nielsen and Einarsen, 2018). In the present study, we have chosen the expression "violence directed against teachers" to describe aggressive behaviors perpetrated by students and/or their parents with the intention of harming the teacher. These include harmful verbal and/or physical behaviors (Anderman et al., 2018) and/or damage to personal property (Dzuka and Dalbert, 2007) that violate school rules, negatively affect the school environment, and put the well-being of those involved in the school at risk (Espelage et al., 2013; McMahon et al., 2017).

The well-being of teachers has been widely studied, with research focusing variously on personal, environmental and relational factors (Brouskeli et al., 2018). Aelterman et al. (2007), p. 286) defined teachers' well-being as "a positive emotional state, which is the result of harmony between the sum of specific environmental factors on the one hand, and the personal needs and expectations of teachers on the other." Many studies have shown that, in different contexts around the world and regardless of significant differences in education systems

(Benevene et al., 2018), violence has a negative impact on the well-being of the teachers affected (McMahon et al., 2014; Aldrup et al., 2018) as well as on the quality of their teaching (Kauppi and Pörhölä, 2012; Montuoro and Mainhard, 2017).

Like violence, well-being is a multidimensional construct (Benevene et al., 2018) that is subject to a range of personal and external factors and that correlates with indicators of resilience (Brouskeli et al., 2018). Occupational well-being among teachers – which is associated with optimal psychological function and their positive experience of work (Ryan and Deci, 2001) – has been defined in terms of both the presence of positive aspects, such as job satisfaction and enthusiasm for work (Benevene et al., 2018), and the absence of negative factors, such as stress caused by difficult relationships with students and their parents (Aldrup et al., 2018). Positive relationships with pupils and parents, and colleagues and school management appear to foster a positive attitude toward the profession (Demirtas, 2010; Berlanda et al., 2017; Buonomo et al., 2017) improve teachers' level of motivation and engagement with their work, and improve their inclination to apply themselves to developing their professional abilities (Schaufeli et al., 2008; Buonomo et al., 2017).

A positive teacher-student relationship is typically characterized by respect, warmth and trust, as well as low levels of interpersonal conflict (Mansfield et al., 2016; Aldrup et al., 2018). Today, teachers are increasingly dependent on the cooperation of parents and need to maintain a positive relationship with them (Skaalvik and Skaalvik, 2011b). Moreover, establishing positive professional relationships with co-workers and senior members of staff may help in establishing networks of emotional support. Social support from other staff at school can foster well-being and protect against the negative effects of violence at school (Galand et al., 2007; Kauppi and Pörhölä, 2012). Mutually supportive collegial relationships between teachers and school management are useful in developing a working community where problems can be discussed and shared, and potential solutions considered collaboratively (Acton and Glasgow, 2015).

Previous research (Galand et al., 2007) suggests that wellbeing depends on the presence of positive aspects – such as job satisfaction, good relations with students and their parents and support from school management and colleagues (Independent Variables) – and the absence of negative experiences, such as teacher-directed violence (Dependent Variables). Investigating which factors affect well-being may offer further and deeper insights, which in turn may contribute to the formation of strategies designed to transform possible risk factors into future opportunities to flourish (Brouskeli et al., 2018). While job satisfaction and good relationships can be seen as elements that contribute to well-being at work, many teachers are exposed to violent incidents that impair their well-being. Support from colleagues and school management can mitigate these difficulties and foster well-being.

H1 = We expect a negative correlation between job satisfaction and perceived levels of violence;

H2 = There is a significant, negative correlation between social support and perceived violence;

H3 = Compared to responses from male teachers, responses from female teachers will trend to indicate a greater level of perceived exposure to violence (in terms of the frequency of incidents of perceived violence);

H4 = Longer-serving teachers can be expected to exhibit higher levels of satisfaction than less long-serving colleagues. H5 = It is expected that job satisfaction, social support and length of service have an impact on perceived levels of student-and-parent-perpetrated violence.

# MATERIALS AND METHODS

An online, self-report questionnaire was administered to a convenience sample of teachers working in primary and secondary schools in north-east Italy. The email addresses used to circulate the questionnaire were provided by the headteachers/directors of the schools involved. Ethical approval for the study was obtained from the Ethics Committee at the Department of Human Sciences at the University of Verona. The questionnaire included a section that explained the nature and the purpose of the study, and a consent form. Written informed consent was obtained from all participants, and participation was entirely voluntary. Participants were informed that they could withdraw from the study or refuse to give information at any time without negative consequences of any sort. We protected the privacy and anonymity of the teachers involved in our research in line with Italian law and the ethical principles of the Italian Psychological Association.

# Participants and Procedure

The study was carried out in March–April 2019. In total, 943 teachers were contacted by email. The research participants were selected on a voluntary basis and a total of 475 questionnaires were completed (response rate of 50.37%). These participants were asked to complete a questionnaire designed to provide measures of job satisfaction, satisfaction with teacher-student and teacher-parent relationships, levels of social support from school management and colleagues, and perceived levels of violence directed against teachers (experienced or witnessed). The large majority of participants were female (345, 72.63%), 127 were male (26.73%), while 3 participants did not indicate a gender (0.64%). The mean age of participants at the time of response was 45.42 years (SD = 9.77; range = 25–66; 8 missing data, 1.68%), and the mean length of service was 15.21 years (SD = 10.75; range = 1– 43; 2 missing data, 0.42%). Most of the participants were working in upper-secondary school (296, 62.32%), 90 participants were teaching in lower-secondary school (18.95%) and 89 in primary school (18.73%).

# Measures

#### Job Satisfaction

There is little agreement regarding the measurement of job satisfaction (Skaalvik and Skaalvik, 2010, 2013). Locke (1976) defined job satisfaction as a pleasurable or positive emotional state resulting from the appraisal of one's job (Ho and Au, 2006; Aldrup et al., 2018). As such, we conceptualized teacher job

satisfaction in terms of the participants' affective reaction to their work or to their role as teachers. We therefore designed the study to measure the teachers' feelings about their work, both in terms of an overall sense of job satisfaction and, more specifically, in relation to a particular aspect of the job, namely the quality of their relationships with students and students' parents. This area of focus was selected in the belief that positive relationships can contribute to a sense of fulfillment.

Overall job satisfaction was measured using a four-item scale (Skaalvik and Skaalvik, 2011a, 2014). The items were: "I enjoy working as a teacher," "I look forward to going to school every day," "Working as a teacher is extremely rewarding," and "When I get up in the morning I look forward to going to work." Responses were given on a 4-point Likert scale ranging from 1 (complete disagreement) to 4 (complete agreement). Cronbach's alpha for the scale was 0.823.

Satisfaction with relationships with students' parents was measured using a three-item scale (Skaalvik and Skaalvik, 2011b) designed to evaluate the extent to which the teachers felt they were trusted by parents. The items were: "I feel that parents have faith in my teaching," "Parents are easy to work with," and "Parents trust and accept my decisions." Responses were given on a 4-point Likert scale ranging from 1 (complete disagreement) to 4 (complete agreement). Cronbach's alpha for the scale was 0.859.

Satisfaction with teacher-student relationships was measured using a three-item scale (with items similar to those on the previous scale). The items were: "I feel that my students have faith in my teaching," "My students are easy to work with," and "My students trust and accept my decisions." Responses were given on a 4-point Likert scale ranging from 1 (complete disagreement) to 4 (complete agreement). Cronbach's alpha for the scale was 0.779.

#### Social Support From School Management and Colleagues

To evaluate perceived levels of social support from colleagues and management figures, we used the six-item Workplace Social Support scale from the Job Content Questionnaire (JCQ; Karasek et al., 1998; Hoang et al., 2013) in the Italian version (Baldasseroni et al., 2001). The short version of this scale measures coworker support (3 items, e.g., "My co-workers are friendly") and supervisor support (3 items, e.g., "My supervisor is successful in getting people to work together"). Participants' responses were recorded on a 4-point Likert scale that ranged from 1 (completely disagree) to 4 (completely agree). Cronbach's alphas were 0.708 (Co-worker support), and 0.680 (Supervisor support).

#### Violence Toward Teachers

To measure perceived levels of teacher-directed violence, we used the eleven-item scale developed by McMahon et al. (2014). This scale assesses three major forms of violence toward teachers: harassment, property offenses, and physical attacks. Harassment included 5 items (e.g., verbal threats, obscene gestures, and internet victimization). Property offenses included 2 items (theft of property and damage to personal property). Physical attacks included 4 items (e.g., physical attacks and objects thrown). We asked the participants to estimate how often they had been subjected to or witnessed any form of violence directed at a teacher in the 12 months prior to completing the questionnaire, first in relation to incidents perpetrated by students and, second, in relation to incidents perpetrated by students' parents. Responses were given on a 4-point Likert scale ranging from 1 (never) to 4 (frequently). Cronbach's alphas were 0.856 (perpetrated by a student), and 0.787 (perpetrated by a parent).

# Data Analysis

Data analysis was performed using IBM SPSS Statistics 21.0. First, for each variable, a composite score was computed by averaging the respective items. Pearson correlation was used to examine the level of association between variables. To test whether different levels of job satisfaction, social support and perceived violence were reported by male vs. female respondents, and less long-serving vs. longer-serving teachers, independent t-tests were applied. To explore the potential differences between primary, lower-secondary and upper-secondary teachers, we applied a One-way ANOVA with post hoc Tukey. Finally, multiple linear regression analyses were conducted. The regression models included job satisfaction, social support and length of service as predictors, and violence perpetrated by students or students' parents as dependent variables.

# RESULTS

# Descriptive Statistics and Correlations

**Table 1** presents the mean score for the research variables with a note of standard deviations and correlations between variables. In line with our assumptions, the results reveal that more satisfied teachers perceive lower levels of student-and parent-perpetrated violence. These results support our first hypothesis. In regard to social support, the data analysis shows that student-and-parentperpetrated violence is negatively correlated with perceived levels of support from co-workers, in the first instance, and support from school management in the second. These results support our second hypothesis. There is a positive correlation between teachers' length of service and their level of satisfaction with their relationships with students. An unexpected result is that the perceived level of support received from colleagues decreases as the length of service increases.

# Independent T-Test and One-Way ANOVA

There were significant differences in the results when broken down by gender or length of service (**Table 2**). Female teachers reported a higher level of perceived parent-perpetrated violence (M = 1.12, SD = 0.18) than male teachers (M = 1.07, SD = 0.18; p < 0.020). Female teachers received less support from school management (M = 2.50, SD = 0.69) than male teachers (M = 2.71, SD = 0.72; p < 0.005). Female teachers derived less satisfaction from their relationships with students (M = 3.14, SD = 0.55) and students' parents (M = 3.05, SD = 0.58) than male teachers (M = 3.25, SD = 0.53; p < 0.050; M = 3.18, SD = 0.59; p < 0.050). Longer-serving teachers may be expected to be better than less long-serving colleagues at managing their classrooms and building good relationships with their students. The results support this, with longer-serving teachers reporting greater levels of satisfaction with student-teacher relationships

#### TABLE 1 | Descriptive Statistics and Intercorrelations (N = 475).


<sup>∗</sup>p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

TABLE 2 | Differences in the sample means (N = 475).


<sup>∗</sup>p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

(M = 3.22, SD = 0.55) than new teachers (M = 3.10, SD = 0.55; p < 0.020). Moreover, longer-serving teachers reported a lower risk of student-perpetrated violence (M = 1.44, SD = 0.41) than new teachers (M = 1.53, SD = 0.43; p < 0.050). These results support our third and fourth hypothesis. Less longserving teachers reported receiving more support from colleagues (M = 3.05, SD = 0.67) than longer-serving teachers (M = 2.89, SD = 0.71; p < 0.020).

Breaking down the results by school level revealed additional significant differences (**Table 2**). Primary school teachers are more satisfied with the student-teacher relationships (M = 3.31, SD = 0.51) than lower-secondary teachers (M = 3.11, SD = 0.56; p < 0.050) and upper-secondary teachers (M = 3.14, SD = 0.55; p < 0.050). Teachers at the primary school level perceived a greater frequency of parent-perpetrated violence (M = 1.19, SD = 0.26) than teachers in both lower-secondary (M = 1.10, SD = 0.16; p < 0.005) and upper-secondary school (M = 1.08, SD = 0.16; p < 0.001). Furthermore, upper-secondary teachers perceived lower levels of support from colleagues (M = 2.90, SD = 0.68) than primary (M = 3.11, SD = 0.75; p < 0.050) and lower-secondary teachers (M = 3.10, SD = 0.63; p < 0.050).

# Multiple Regression Analysis of Variables on Violence

**Table 3** shows the results of multiple regression analysis. The results indicate that student-perpetrated violence against teachers is negatively associated with perceived levels of support from school management, while parent-perpetrated violence against teachers is positively associated with satisfaction with student-teacher relationships, and negatively associated with both perceived levels of support from school management and the teachers' level of satisfaction in relation to their relationships with students' parents. These results support our fifth hypothesis.

# DISCUSSION

#### Resource Identification Initiative

The results of our study show that, in order to improve wellbeing at work and decrease perceived violence, it is necessary to promote job satisfaction, social support and positive relationships at school. With regard to violence perpetrated against teachers by students, the results show that support from school management operates as a protective factor. Regarding perceived levels of violence perpetrated against teachers by students' parents, the results indicate that this variable is positively associated with the teachers' satisfaction with student-teacher relationships, and negatively associated with both the level of support they feel they receive from school management and their level of satisfaction with their relationships with students' parents. These data are consistent with the existing literature on this topic, according to which (Skaalvik and Skaalvik, 2011b; Mansfield et al., 2016; Benevene et al., 2018), more satisfied teachers are better able

#### TABLE 3 | Multiple regression analysis.

fpsyg-10-01807 August 5, 2019 Time: 13:53 # 6


The 95% bootstrap CIs were computed for unstandardized regression coefficients (1,000 resamples). CI, confidence interval.

to establish positive relationships, resulting in lower levels of perceived violence. A sense of satisfaction with the work environment can foster positive relationships at work, which in turn promotes self-confidence and the development of skills for dealing with aggressive behaviors on the part of students or their parents (Buonomo et al., 2017; Fiorilli et al., 2017).

Breaking down the results by gender: women seem more strongly affected by aggressive behaviors (experienced directly or Witnessed) than their male counterparts. This would appear to be in line with finding -supported by validated data- that, in comparison to men, women tend to perceive themselves as more vulnerable when exposed to threating contexts (Harris and Miller, 2000; Berg and Cornell, 2016), and report a greater number of experiences of aggression. A similar variance emerges in our data in relation to the variable "length of service" (Berg and Cornell, 2016; Reddy et al., 2018), with less long-serving teachers more inclined to feel vulnerable when threatened with aggressive behavior. Interpreting this result is not a straightforward matter, as there are multiple variables that have not been considered in this study. The perception of a lower frequency of violent behaviors among longer-serving teachers might be explained, at least in part, by a greater ability in dealing with potential conflicts with pupils (Reddy et al., 2018). We might also hypothesize that longer-serving teachers have been subject to some sort of selection effect (Berg and Cornell, 2016). One unexpected outcome to emerge in our analysis – again in connection with length of service– is that the perceived level of support from colleagues tends to decrease as length of service increase. When we take into account the school grade, the lowest rate of support by colleagues is highest rate in upper-secondary school. One possible explanation is that longer-serving teachers tend to be more capable at managing problems, which leads their colleagues to offer less support. Whether or not this is the case, it is result that raises some challenging questions and one that merits further, more detailed analysis focusing on contextual variables. Another interesting result is that teachers who perceive more violence on the part of their students' parents tend to be more satisfied with their relationship with their students but less satisfied with their relationship with the parents. In particular, the primary teachers in our study reported both the highest levels of parent-perpetrated violence and the greatest levels of satisfaction in relation to student-teacher relationships. Having a good relationship with students is usually associated with the perception of reciprocated trust between student and teacher. This may place the teacher, in a sense, in competition with the parents. Furthermore, a teacher who enjoys a positive relationship with a student or students may feel more driven or entitled to make demands on parents in terms of expecting them to cooperate in their child's education and share the teacher's mindset, which could be a trigger for confrontation or situations of conflict.

Positive relationships encourage positive emotions and support professional flourishing (Acton and Glasgow, 2015). There is wide recognition in the literature (supported by our own data) that social support has a vital role in teacher's coping processes (Kauppi and Pörhölä, 2012), because it offers a valuable resource for dealing with stressful experiences. Other studies indicate that colleagues and management figures can offer practical support with problems (Fiorilli et al., 2019) thus fostering emotional well-being and protecting against violent events (Chen and Astor, 2008; Skaalvik and Skaalvik, 2009; Fiorilli et al., 2017, 2019; Aldrup et al., 2018; Anderman et al., 2018; Bounds and Jenkins, 2018; Nielsen and Einarsen, 2018). Moreover, the results of existing studies suggest that higher levels of support from colleagues and school management are negatively associated with exposure to student-and-parentperpetrated violence (Galand et al., 2007; Reddy et al., 2018). Generally speaking, a school management structure that favors team-working and participative decision-making can help teachers cope with the emotional and psychological activation caused by violent events (Galand et al., 2007; Kauppi and Pörhölä, 2012) and have a demonstrable, direct protective effect on subjective well-being (Reddy et al., 2018). The results of our study highlight the importance of support from head teachers/school

directors, which appears to have a direct protective effect that can lead to an overall increase in occupational well-being (Galand et al., 2007; Reddy et al., 2018) and underpin the quality of the educational process (Kauppi and Pörhölä, 2012; Montuoro and Mainhard, 2017).

In spite of initiatives implemented to address the serious issue of violence in schools, it remains an emerging concern. This leads to the conclusion that strategies designed to prevent violence have not yet proven entirely effective, leaving education professionals at risk of some form of aggression. In particular, parents are more involved than ever in the school environment and, despite the importance of positive school-family relationships for the educational process, students and their parents increasingly display a negative attitude toward school staff. However, our data indicate that aggressive behaviors do not always have a profound negative impact on the recipient. Indeed, under certain conditions, professionals can display and leverage personal qualities that help reduce the impact of violence and draw on capacities associated with resilience. These findings can be more clearly understood in terms of the broaden-and-build theory (Fredrickson, 2000, 2001, 2004), which sets negative and positive emotions apart as two separate systems that respond to different psychological needs. While negative emotions are useful in terms of self-preservation, for instance for focusing on potential threats, positive emotions can help in broadening thought–action repertoires and developing personal resources and creative actions. As such, negative and positive emotions work as complementary systems, in which one of the principal functions of positive emotion lies in deactivating the "negative setting" engendered by negative emotion in order to open up resources for resilience (Fredrickson et al., 2000). In these terms, well-being entails the fulfillment of profound psychological needs and should not be understood as the mere absence of negative emotions (Buss, 2000; Ryan and Deci, 2000). Relationships play a crucial role in the process of building resilience (e.g., Le Cornu, 2013; Gu, 2014; Mansfield et al., 2016) and efforts to engender positive emotions by working to develop better relationships should be a keystone of strategies for establishing and sustaining professional flourishing (Acton and Glasgow, 2015).

# REFERENCES


Our study has certain limitations, which we can acknowledge here. First, it involved a small number of schools, which makes it difficult to draw generalizable conclusions from the data. All the same, our data are in agreement with the existing literature, and we have no evidence that the situation is different in other schools. Second, the crosssectional nature of our study constitutes another limitation and a longitudinal study would be desirable at this point. Third, we used an online questionnaire, which was issued to teachers in schools selected on a convenience basis rather than through systematic sampling. This choice meant both that teachers who feel less comfortable with web-based data collection were less likely to participate, and that teachers working in schools other than those selected were excluded altogether. Furthermore, the study used a retrospective, selfreporting approach for data collection, which could be affected by recall bias and the subjectivity inherent to self-reported measurements. To minimize these sorts of problems, we used validated questionnaires that have been shown to have good reliability. Finally, to avoid making the research instrument unwieldy, and thus potentially limiting the total number of returned completed questionnaires, we decide to limit the number of the items in the questionnaire (see Phellas et al., 2011). This choice limited the capacity of the study to consider a number of contextual variables, particularly in regard to differences between school environments. As we have suggested already, these would merit further investigation.

# DATA AVAILABILITY

The dataset generated and analyzed for this study is available from the corresponding author upon request.

# AUTHOR CONTRIBUTIONS

All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication.




**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 De Cordova, Berlanda, Pedrazza and Fraizzoli. This is an openaccess article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Italian Teachers' Well-Being Within the High School Context: Evidence From a Large Scale Survey

#### Barbara Barbieri <sup>1</sup> \*, Isabella Sulis <sup>1</sup> , Mariano Porcu<sup>1</sup> and Michael D. Toland<sup>2</sup>

<sup>1</sup> Department of Social and Political Sciences, University of Cagliari, Cagliari, Italy, <sup>2</sup> Department of Educational, School and Counseling Psychology, University of Kentucky, Lexington, KY, United States

This paper aims to investigate the relationship between Italian teachers' well-being, socio-demographic characteristics and professional background. Using data from the 2015 wave of the Program for International Student Assessment (PISA) we considered information collected by the questionnaire completed by a total of 6,491 teachers in the sampled schools. Moving from existing literature on teachers' well-being, we investigate several aspects related to the teachers' working environment, career motivation and investment, and job satisfaction. We assess the variability in the observed outcomes attributable to school factors and heterogeneity between disciplines. Measurement models are combined in a multilevel setting in order to define teachers' well-being on a broad perspective while accounting for the multiple sources of heterogeneity due to several factors (e.g., discipline, teacher professional background, and individual differences) occurring at different levels of the data structure. In general, results show that the teachers' positive perception of the working environment in terms of availability of adequate human and physical resources, and professional development opportunities, provide a substantial state of well-being at work, and are related to teachers' job satisfaction. Moreover, results highlight the key role of transformational leadership in defining a teacher's well-being. Findings and implications are discussed.

#### Edited by:

Paula Benevene, Libera Università Maria SS. Assunta, Italy

#### Reviewed by:

Michela Cortini, Università degli Studi G. d'Annunzio Chieti e Pescara, Italy Alessandro De Carlo, Giustino Fortunato University, Italy

#### \*Correspondence:

Barbara Barbieri barbara.barbieri@unica.it

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 07 June 2019 Accepted: 05 August 2019 Published: 21 August 2019

#### Citation:

Barbieri B, Sulis I, Porcu M and Toland MD (2019) Italian Teachers' Well-Being Within the High School Context: Evidence From a Large Scale Survey. Front. Psychol. 10:1926. doi: 10.3389/fpsyg.2019.01926 Keywords: well-being, job satisfaction, teachers, PISA, measurement models, multilevel

# INTRODUCTION

Research in different cultures indicate that school teachers are among those professionals with the highest levels of job stress (Stoeber and Rennert, 2008; Fiorilli et al., 2015a), often attributed to an excessive workload, poor relationships with colleagues, lack of suitable resources, constant changes within the profession, and inadequate salary (Santavirta et al., 2007; Simbula et al., 2012; Falco et al., 2013).

Italian teachers are in the last positions for the perceived social recognition of the profession (OCDE, 2014). Furthermore, in 2015 the Italian National Observatory on Health and Wellbeing of Teachers (ONSBI—Osservatorio Nazionale Salute e Benessere degli Insegnanti) research study highlighted the low satisfaction of the category for the physical (Guglielmi et al., 2012) and organizational conditions of schools (Fiorilli et al., 2015b). Although the OBSBI study was able to assess the risk levels of teachers' physical and psychological health and the factors that influence them, it was not very informative with respect to the factors that contribute to generating a condition of well-being among teachers. Studies on the school environment, and in particular on the teachers' role, have been focused primarily on the nature of the stress process, assuming that the context itself is overly demanding and characterized by lack of support toward teachers (Kyriacou, 2001). A more recent line of studies on teacher well-being at work, conversely, analyzes the working environment from a resource perspective, emphasizing how work environment and job characteristics explain motivation or work engagement despite high job demands (Bakker et al., 2007; Magnano et al., 2014). Moreover, literature highlights how job resources may be located at the level of the organization at large (e.g., career opportunities, job security), interpersonal and social relations (e.g., supervisor and co-worker support, team climate), or organization of work (e.g., participation in decision making; Bakker et al., 2007).

Thus, the central aim of the present study is to investigate, in a sample of 6,491 Italian high school teachers, how some professional and organizational factors (i.e., professional choice; professional development; principal leadership and colleagues support; perception of rich vs. poor work environment; management practices and polices) could be perceived by teachers as potential resources of well-being in terms of job satisfaction at work and satisfaction with respect to their profession. Understandably, a teacher's well-being at school is not only related to the professional context, but a teacher's personality or personal background can influence the way in which a teacher deals with factors within the school and with the demands of the profession. Consequently, we not only measure the individual well-being of teachers, but assess the school's contribution to overall teachers' well-being.

# THEORETICAL FRAMEWORK

As it is well-known in literature, teaching is a highly demanding occupation (Travers and Cooper, 1996; Guglielmi and Tatrow, 1998; Kyriacou, 2001; Zurlo et al., 2007). Research suggests that the main stressors among teachers may include, low salaries, disagreement with colleagues or conflict with students' parents, redefinition of job conditions as a consequence of school reform, workload, role conflict and ambiguity, and poor working conditions (Travers and Cooper, 1996; Kyriacou, 2001; Zurlo et al., 2007; Benevene and Fiorilli, 2015; Guidetti et al., 2015; Falco et al., 2017; Girardi et al., 2018). Although many studies indicate antecedents of discomfort, stress and burnout, it's not until more recently that studies have observed the link between organizational health and teacher well-being (Hakanen et al., 2006; Skaalvik and Skaalvik, 2011; Miglioretti et al., 2013). Researchers have also shown how teacher well-being could be an important determinant of student learning outcomes and students' well-being, highlighting the importance to establish the determinants not only for teachers' negative outcomes, but also positive ones that are associated to their work (Denny et al., 2011; Converso et al., 2014; Guidetti et al., 2015). Form this perspective, the present study aims to investigate teachers' well-being within the theoretical framework of the Job Demand –Resource model (JD-R model), in which job resources are all those factors that can help employees in reducing job demands and related costs (Bakker and Demerouti, 2007; Barbieri et al., 2016; Farnese et al., 2017; Falco et al., 2018; De Carlo et al., 2019). In fact, job resources include factors capable of fostering personal growth, learning and development by providing feedback, enhancing task significance, and stimulating employees' motivation (Bakker et al., 2005; Bakker and Demerouti, 2007, 2014).

In light of the above, our goal in this study was to assess the school's contribution to teachers' well-being. More specifically, some professional and organizational dimensions have been considered as distinctive resources of the job, able to positively influence teachers' well-being. To this end, we considered professional choice (i.e., did the teacher actually want to be a teacher), and professional development as characteristics of job resources at the professional level; while the perception of rich vs. poor work environment (i.e., shortage of teaching resources, shortage of educational resources), practices and policies management (i.e., to take part in professional development activities) and principal leadership and colleagues cooperation as characteristics of job resources at the organizational level which are assumed to be related to teachers' well-being. We also considered teacher demographic factors as they relate to well-being [i.e., gender, work experience (i.e., years working as a teacher), and the type of subject taught (i.e., Humanities, Math-Sciences, Other)].

# Professional Choice

In educational literature it is recognized that teacher identity is a key influencing factor on teachers' sense of purpose, self-efficacy, motivation, commitment, job satisfaction and effectiveness (Day et al., 2006; Avanzi et al., 2012; De Stasio et al., 2017). A strong identification with social self (i.e., being a teacher), produces a positive effect on relationships and perception of effectiveness of their own role. Moreover, a perception of effectiveness of the own role is positively related to commitment to and a passion for teaching (Day, 2004) and to stay involved in school activities. According to self-determination theory (Deci and Ryan, 1985), people who carry out a chosen profession are more intrinsically motivated to engage in the activities of their job. Ryan and Deci (2000) also have described how extrinsically motivated activities can be internalized if people value the outcomes of these efforts and if people have access to supportive contextual conditions. In this study we considered the professional choice as a personal resource. An important extension of the JD-R model includes personal resources, as a further protective factor that can trigger positive outcomes (Bakker and Demerouti, 2014; Bakker et al., 2014). Some scholars (Xanthopoulou et al., 2007) observed that job resources influenced personal resources conceived as the individuals' sense of ability to successfully control their environment—and this, in turn, had a positive influence on burnout reduction (Bakker et al., 2014). Following this observation, we considered the personal choice of being a teacher as a personal resource that is useful to understand and control the work environment. Ultimately, this resource can be strengthened and enhanced through organizational strategies (e.g., school's practices and policies oriented to a more participatory and inclusive processes management could have a positive influence on the teachers' sense of ability to successfully manage the school environment).

# Professional Development

Teacher satisfaction at work is related to the ways in which school policies encourage or facilitate career aspirations and development (Butt and Retallick, 2002). Several researchers have highlighted that professional development paths can have positive influences on outcomes such as teachers' satisfaction (Ullah and Jundran, 2014), attitude change, commitment to innovation (Desimone, 2009), and self-efficacy (Tzivinikou, 2015). Professional development can either be hindered by the school's organizational context or on the contrary, can contribute toward commitment to learning goals and collaboration in school (King, 2002). Different research on employees in different types of working environments have shown that personal development opportunity (Liu and Wang, 2001), training (Long et al., 2002), and opportunity for learning (Ng et al., 2006) explain employees' commitment to their organizations and occupational well-being (Weng et al., 2010; Barbieri et al., 2018).

# Work Environment

Researchers in positive psychology have identified the critical role that satisfying work plays in psychological well-being across various domains of human functioning (Lent, 2004; Doest et al., 2006). Numerous studies (see Hoy and Miskel, 1982; Huberman and Vandenberghe, 1999; Hallinger, 2003) have also underlined how workplace-related factors can positively influence job satisfaction and well-being. Such as content of the job, physical and material working conditions, school management, school climate, and interpersonal relationships (e.g., Hakanen et al., 2006). Ultimately, the intensity and frequency with which certain conditions are realized or missing will have consequences for the employee's well-being.

# Management Practices and Policies

The organizational domain of the teacher functioning at school has a markable influence on the teacher as an employee (Cherniss, 1993). For instance, the extent to which teachers are involved in school-based decisions determines the scope in which they sense control over expressing their values through their work (Friedman, 2003). Moreover, several researchers have emphasized that an important element of teachers' identities is related to their experiences in the school (Galloway et al., 1982; Mortimore et al., 1988; Woods et al., 1997). The culture of the school, its internal dynamics and organization, enable or hamper teacher "satisfaction," "commitment," and "motivation," and ultimately the culture of the school has an impact upon a teacher's construction of their own teacher identity.

# Principal Transformational Leadership and Colleagues Cooperation

An important resource for teachers, as for most helping service professions, is to recognize and utilize a social support system to deal with work stress, prevent burnout (Kyriacou, 2001), and increase job satisfaction (Brough and Pears, 2004). Also, theoretical arguments suggest that transformational leadership will be positively associated with perceiving work as meaningful (Arnold et al., 2007; Benevene et al., 2018; Magnano et al., 2019). For instance, Piccolo and Colquitt (2006) have demonstrated a positive link between transformational leadership and employee perceptions of meaning in terms of these job characteristics. Some researchers have argued that this kind of leadership "gives meaningfulness to work by infusing work [. . . ] with moral purpose and commitment" (Shamir et al., 1993, p. 578). Also, the individual respect and support that a transformational leader exhibits for each employee should also apply to the actual work in which each employee is engaged. Transformational leaders inspire employees to transcend self interest and perceptions of their own limitations to become more effective in pursuing collective goals (Bass et al., 2003). They support employees in working toward the goals, such as by acting as a role model, stimulating them to engage in analysis, showing concern for them as individuals, and encouraging teamwork (Podsakoff et al., 1990). Studies have consistently shown that in the general population social support from a leader/manager and colleagues is positively associated with work engagement and well-being (Schaufeli and Bakker, 2004). In addition, researchers have shown the importance of the support provided by a leader not only in generating a sense of meaningfulness (Kahn, 1990), resilience, security, and general motivation (Ryan and Deci, 2001), but also in enhancing an intrinsic or extrinsic motivational role (Bakker and Demerouti, 2007). Laschinger et al. (2004) found that if a leader provides a more supportive environment for their employees, employees will adopt better work attitudes. Also, leaders who show appreciation and support may aid the worker in coping with the job demands, facilitate performance, and act as a protector against ill health (Väänänen et al., 2003). For instance, a study among Finnish teachers (Bakker et al., 2007) showed how leader/supervisor support maintains a high work engagement over time among employees. Other studies have shown that a high quality relationship with colleagues may alleviate the influence of job demands (e.g., work overload, emotional and physical demands). Consistent with studies on subjective wellbeing (e.g., Diener et al., 1999), the way in which teachers perceive working conditions and school relationships affects well-being at school (Kinman et al., 2011).

# PURPOSE OF THE CURRENT STUDY

Based on the above described theoretical analysis, the main objective of this study is to assess the school's contribution to the teachers' well-being. Therefore, we concentrated on professional and organizational factors, which could have an influence on job and professional satisfaction and in turn on teachers' well-being in the work environment. Our study addresses the following exploratory research questions:

Q1: To what degree does the teaching working environment, management practices and policies and transformational leadership explain differences in teachers' well-being? Do these factors have varying influences on the two response variables used to tap into well-being: namely Satisfaction Barbieri et al. Italian Teachers' Well-Being

with the current job environment and Satisfaction with the teaching profession?


#### METHOD

#### Sample and Variables

In order to investigate the influence of the above mentioned factors (teaching working environment, management practices and policies and transformational leadership) on teachers' wellbeing, controlling for heterogeneity, we used data collected in the large scale assessment survey named Program for International Student Assessment (PISA) carried out by the Organization for Economic Co-operation and Development (OECD) in 2015. The PISA is carried out every 3 years (since 2000) to yield comparisons of students' achievement in reading, mathematics, and science among participating countries. In 2015, for the first time, in 19 countries, the PISA surveyed teachers to inquire about different aspect of their working experience.

In this study we considered data collected in Italy where a sample of 9,738 teachers were selected. As stated in the OECD-PISA technical reports (OECD, 2019) in Italy the proportion of non-responding teachers was one of the highest, with 28.1% of sampled teachers not answering the survey<sup>1</sup> .

Therefore, we had a total of 7,005 completed questionnaires, and we retained for the analysis only those teachers that identified having a formal education of at least of ISCED (International Standard Classification of Education) Level 5. So, we considered only teachers with a tertiary degree. This is because we wanted to analyze exclusively the working context of Italian high school teachers with a similar training background. In the end, a total of 6,491 teachers were considered herein.

Datasets made available by OECD include several information, the variables considered in the analysis are described below and classified according to their characteristics. Detailed information on the survey design and variables can be found in documents and reports published at www.oecd.org/ pisa/data/. Some descriptive statistics about the selected variables are reported in **Table 1**.

#### Response Variables

The two response variables have been scaled using teacher answers to different statements related to the two questions "We would like to know how you generally feel about your job. How strongly do you agree or disagree with the following statements?" and "We would like to know how you generally feel about your profession. How strongly do you agree or disagree TABLE 1 | Descriptive statistics.


(Continued)

<sup>1</sup>Countries with the highest percentage of non-respondents were Australia, Brazil, Germany and Italy, each with about 30% of the total. For these countries, results should be interpreted with caution as they might be affected by possible nonresponse bias (OECD, 2019).

TABLE 1 | Continued


with the following statements." Answers have been collected using a four-point Likert scale that ranges from "strongly agree," "agree," "disagree" to "strongly disagree." The derived variables were scaled by OECD using an item response theory (IRT) scaling procedure.

SATJOB. Satisfaction with the current job environment. Higher values of the variable point at a higher job satisfaction about the job environment. The statements considered were:


SATTEACH. Satisfaction with teaching profession. Higher values of the variable point at a higher satisfaction about teaching profession. The statement considered were:


The polarity of negative items has been reversed.

# Manifest Variables

PROPDT20. Proportion of professional development. Teachers were asked to respond to the following question: "During the last 12 months, did you participate in any of the following activities?" Activities included were participation in a "qualification programme," a "network of teachers focusing on professional development," "individual or collaborative research on a topic of interest," "mentoring and/or peer observation and coaching," "reading professional literature" and "engaging in informal dialogue with colleagues." The derived binary variable PROPDT20 indicates whether a teacher took part in any of these activities in the past 12 months.

REQPROFDEV. Required to take part in professional development activities. Teachers were asked to respond to the Yes/No question: "Are you required to take part in professional development activities?."

TCEDUSHORT. Educational material shortage. Teachers were asked to provide information if their school's capacity to provide instruction was hindered due to lack of educational resources (such as textbooks, libraries, laboratory materials, etc.). The fourpoint Likert-type scales ranged from "not at all," "very little," to "to some extent," and "a lot." Higher values indicate a higher educational material shortage.

TCSTAFFSHORT. Educational staff shortage. Teachers were asked whether their school's capability to operate is hampered by a lack of educational staff (such as textbooks, libraries, laboratory materials, etc.). The four-point Likert scales ranged from "not at all," "very little," to "to some extent," and "a lot." Higher values indicate a higher educational staff shortage.

TCLEAD. Transformational leadership. The index has been derived with an IRT scaling procedure to collect information about teachers' view on school leadership. Higher values point at a higher transformational leadership in the teachers' perception. The question considered was "To what extent do you disagree or agree with the following statements regarding your school?"; statements related to this question were (teachers provided their answers in a four-point Likert-type scale ranged from "strongly agree," "agree," "disagree" to "strongly disagree"):


EXCHT. Exchange and co-ordination for teaching. A question addressed to evaluate teaching-related co-operation using statements like "teaching jointly" or "exchanging teaching materials" has been considered to scale the indicator (via a IRT model). Teachers were asked to rate these activities with the following answering categories "never," "once a year or less," "2–4 times a year," "5–10 times a year," "1–3 times a month," and "once a week or more."

WANNABETEACH. After completing ISCED 3-level, was your goal to pursue a career in the teaching profession? Yes/No answers to this question have been considered.

#### Control Variables

The following control variables were considered for each respondent: gender (SEX), age (AGE), years working as a teacher at her/his school (WKEXPSCH), and the employment status as a teacher (EMPLSTATUS: Full-time >90%, Part-time 71– 90%, Part-time 50–70%, Part-time <50%). Moreover, the subject taught at school has been recorded (TCWHAT: Humanities, Math-Sciences, Other); note that with reference to the latter two variables we recorded a high number of missing observations, thus the variable EMPLSTATUS has not been used in the modeling approach as associated with the years of working experience and with age.

#### Data Analysis Plan

In order to explore the relationship between teachers' wellbeing and the teachers' working environment, considering the

heterogeneity between disciplines and teachers' characteristics, a bivariate regression model has been adopted. Specifically, the use of a bivariate approach allows us (Leckie and Charlton, 2013; Sulis and Porcu, 2015; Grilli et al., 2016; Leckie, 2018) to consider the effect of teacher working environment and management practices and policies on both outcome measures selected to monitor teachers' well-being: namely Satisfaction with the current job environment (SATJOB) and Satisfaction with teaching profession (SATTEACH). Both variables are treated as response variables in a regression framework with the aim to shed light on which management practices and policies may have the greatest impact on improving teachers' well-being. The use of a bivariate approach allows us to assess the level of association between the two outcome measures adjusting for the heterogeneity in teachers' characteristics (e.g., disciplines, teacher professional background and professional choices and more generally individual differences, etc.). Moreover, the explanatory model is then combined in a multilevel framework in order to assess (i) if there is a "school effect" in the heterogeneity of the results (Leckie and Goldstein, 2009), (ii) if this effect can be explained by management policies and practices in terms of leadership, poor vs. rich environment and cooperation, and (iii) the impact of these factors at different levels of the data structure (i.e., individual teacher perception vs. perception of teachers in the same school).

Specifically, we fit a multilevel bivariate regression model which considers at level-1 teachers and at level-2 schools and treats the two response variables as two related dimensions of a single underlying latent variable "teacher well-being." Namely, the satisfaction of teacher i (I = 1,. . . , N) belonging to school j (j = 1,. . . , J) with respect the dimension y d ij is specified as function of the vector of respondent (**x**) and school (**z**) covariates. Moreover, two bivariate normally distributed residual random terms θ d <sup>j</sup> ∼ N(0, 2) and ε d ij ∼ N(0, 6) are considered at school (Level-2) and teacher (Level-1) level to take into account for differences between schools and between teachers within the same school

$$
\gamma^d\_{\vec{i}\vec{j}} = \alpha^d + \varkappa^{'d}\_{\vec{i}\vec{j}} \beta^d + z^{'d}\_{\vec{j}} \gamma^d + \theta^d\_{\vec{j}} + \varepsilon^d\_{\vec{i}\vec{j}}.\tag{1}
$$

As a result, the residual variability in the observed outcomes is split in the between-school and within-school components. The random term θ d j , which is shared by teachers belonging to the same school, captures the school "net" effect on the two measures of teachers' well-being (Goldstein and Spiegelhalter, 1996; Goldstein, 2011). The bivariate multilevel model relationship described in Equation 1 is displayed in **Figure 1**. The random term ε d ij captures the residual variability between the responses of teachers belonging to the same school. The model has been estimated with STATA using the runMLwiN routine implemented by Leckie and Charlton (2013), by specifying as the estimation method the Iterative Generalized Least Squares. The bivariate approach allows us to assess the degree of correlation between the two dimensions of teachers' well-being: teachers' satisfaction with respect to their current job and their profession, to assess the effect of covariates on both components and to control the effect of the confounding factors (control variables) on both dimension of well-being.

# RESULTS

We adopted a model building strategy to assess the determinants of teachers' well-being extending a bivariate null model (M0) in several directions. Namely, the bivariate model has been generalized considering the clustering of teachers in schools and the heterogeneity between teachers and schools in terms of their characteristics. Predictors have been clustered in three groups: (i) teacher personal and professional details (SEX, AGE, WKEXPSCH, TCWHAT, PROPDT), including the teacher's self-stated motivation toward the profession (WANNABETEACH); (ii) teacher's perception of the school working environment in terms of resources, organizational details, practices and polices (REQPROFDEV, TCEDUSHORTAGE, TCSTAFFSHORTAGE, TCLEAD,

EXCHTEACH); (iii) compositional variables at school level obtained as the average at the school level of predictors mentioned in point (ii).

To manage the high rate of missing values in some variables, the information provided by variables TLEAD and EXCHTEACH has been considered only at the school level, taking for both variables the average value over teachers belonging to the same school. Moreover, missing values observed for the variable TCWAHT have been treated as a further category of the variable, defining four possible response categories (not declared, mathematics, language, others). We have taken this decision for two main reasons: (1) to inspect if there were differences in the two dimensions of perceived well-being related to belonging to one of the two clusters of teachers (math and language) with respect to other clusters and (2) to assess if there were any differences between observations with missing values and the others. The decision was also supported by the fact that language and maths are the two main disciplines which provide students the competences investigated in the main large scale assessment surveys carried out at both national and international level. So we want to inspect if the was a kind of "main disciplines" effect in teachers' perception of the two dimensions of well-being. Also some studies have suggested that teaching subjects could have an effect on job satisfaction and competence that in turn affect teacher well-being (see for example Pillay et al., 2005; Kunter et al., 2011; Ilgan et al., 2015). The variable AGE has been split into three intervals: <36, 36–50, >50. Although relevant literature for the Italian context, uses a different way to categorize age (namely, <39, 40–54 and >55) (see, for example James et al., 2011; Avanzi et al., 2012), this clustering of the variable has allowed us to maximize differences across observations in the effect of age on the two dimensions of perceived well-being. We did not consider the variable as continuous and we applied this categorization in order to better differentiate the effect of the predictors on the response variable. Finally, given the rate of missingness in almost all variables, 5,973 complete observations (92%) have been retained for the final analysis. Overall, five competing models (Models 0–4 [M0-M4]) were estimated and results compared in terms of goodness of fit statistics (Bayesian Information Criterion [BIC] and Akaike information Criterion [AIC]), which are summarized in **Table 2**. The model comparisons included comparing the variance component model (M1) to the null model (M0) to assess the relevance of the clustering of students in schools. We also compared the multilevel model with teacher's predictors (M2) with the variance component model (M1) to assess the relevance of teachers' characteristics defined at point (ii) in terms of personal and professional details and perception of the school working environment in terms of resources, organizational details, practices and policies on the two outcomes. Furthermore, the multilevel model with compositional variables at school level (M3), has been estimated to assess the effect of school compositional variables in terms of organizational aspects, practices and policies on teachers' well-being. The comparison between M3 and M2 shows that the introduction of compositional variables at the school level (M3) to determine if a greater improvement of goodness of fit statistics with respect to the introduction of predictors at the teacher level (M2) exists. This finding suggests that school management in terms of transformational leadership (the only significant predictor) has the greatest impact on teacher well-being. Finally, M4 which assesses the effect of relevant predictors at teacher and school level considering both sets of covariates, combines findings from M2 and M3. Comparisons between bivariate models in terms of goodness of fit statistics in **Table 2** show that M4 is the model which the data better than all other models. Thus, it is championed for analyzing differences in well-being between teachers and to shed some light on its determinants.

#### Key Findings

Results depicted in **Table 2** for M4 display a medium to low level of correlation between the two outcome variables (0.48) suggesting that they play a key role in assessing teaching wellbeing. Also older teachers have on average a general level of satisfaction lower than younger teachers. Particularly strong is the effect of belonging to the highest age group on satisfaction toward the teaching profession (on average the oldest age group of teachers are 0.4 less satisfied than the youngest group of teachers). Teachers who had the ambition to pursue a career in the teaching profession are on average more satisfied with respect to both dimensions of well-being as it has been operationalized, with a greater impact on the satisfaction toward the teaching profession. Moreover, the number of years of working experience in the school where they currently work seems to have a positive influence only on satisfaction toward the profession. With respect to the disciplines, results suggest that science teachers show on average a greater level of satisfaction with respect to their profession than the other groups, while satisfaction with respect to the working environment seems to be higher for teachers in minor subjects. Teachers who invested in professional development in the last 12 months are on average both satisfied with respect to both dimensions. The main finding concerning organizational aspects, management policies and practices provided by M4 (see **Table 2**) suggest that the teachers' perception of a poor working environment, in terms of shortage of resources and teaching staff, has a negative impact on both dimensions of teachers' well-being, but a stronger impact on satisfaction toward the current job environment (TCEDUSHORTAGE = −0.13 and TCSTAFFSHORTAGE = 


(Continued) Italian Teachers' Well-Being

#### TABLE 2 | Continued


Standarderrorsinparentheses.\*\*\*p<0.01.\*\*p<0.05.

 \*p < 0.10.

Barbieri et al. Italian Teachers' Well-Being

−0.16). Also, the shortage of teaching resources has the greatest impact on satisfaction with the current job environment, whereas the shortage of educational resources has a greater influence on the satisfaction with teaching profession. The school policy to require teachers to take part in professional development activities (REQPROFDEV) provides an expected increase of +0.11 of the score of the teacher's satisfaction with the current job environment and positively affect the satisfaction also with teaching profession (+0.08). Looking at the role of the school principals, it arises that the transformational leadership teachers view is a management practice that has a relevant impact on teachers' well-being. Namely if we compare the value of two schools with two extreme positions (i.e., minimum = −1.03 vs. maximum value = 0.64) with respect the transformational leadership factor it arises that this practice explains about 0.4 standard deviation in the difference in teachers' satisfaction with respect their job, weaker is the effect on satisfaction with respect to their profession. Teachers' perception of a cooperative environment, which promotes exchange and co-ordination for teaching, does not show a relevant effect. This could be explained by the association of this aspect with other practices. The comparison between M0 and M4 also shows that the introduction of teachers and school covariates among the predictors explain almost all the between school variability and produces a decrease of about 11% of the within school variability. To sum up, after controlling for teacher heterogeneity in age, disciplines, working experience and career choices (i.e., WANNABETEACH and PROPDT), it arises that the perception that teachers' have of school resources together with the implementation of policies which support professional development can have a great impact on both aspects of teaching well-being. Moreover, as expected, teachers' well-being benefits of policies and practices that school can implement in terms of supporting transformational leadership, whereas seems not been related to the perception teachers' have of the coordination for teaching. It is worth to highlight that almost all variability in the expected outcomes is due to unexplained individual differences and that the residual variability at school level is very low compared with the one observed at teacher level. However, the residual variability at school level is significant in model M1-M4. Moreover, the value added of a multilevel approach instead a simple bivariate approach is that it allows us to deal with covariates at school-level and to assess their impact on individual well-being.

# DISCUSSION

The focus of this study was to investigate factors influencing and enhancing well-being among teachers in Italy. The modeling approach here adopted allows us to shed light on determinants of the differences in teachers' well-being and to differentiate on their impact on the two outcome measures (Research Question –RQ– 1). Specifically results provide evidence that a teacher's professional background together with their experience, organizational aspects and school practices and policies have a relevant impact on both outcome measures as they relate to wellbeing. The impact of teacher personal and professional details is stronger on satisfaction toward the teaching profession whilst, as expected, the effect of organizational aspects, practices and policies is stronger on the satisfaction toward the current job.

Concerning personal and professional details that mainly affect observed differences in teachers' well-being (RQ 2) results suggest that younger teachers, with a greater working experience, who pursued a specific training to become a teacher and invested in professional development experienced on average higher levels of well-being than older teachers. Question regarding number of years of teaching experience is however controversial, in fact some studies reported that longer teaching experience is associated with higher levels of job satisfaction (Taylor and Tashakkori, 1995), on the contrary others reported that teachers who have taught longer are less satisfied (Xin and MacMillan, 1999), thus continuing the debate and encouraging further research.

Moreover, findings highlight that policies and practices that school principals can support to improve teachers' wellbeing (RQ 3) are those which support a transformational leadership teachers' view involve the investment in professional development and the availability of educational resources for teachers. As other studies have shown, the support of the principal, in terms of encouragement and involvement in participatory decision-making processes has positive repercussions on the well-being of teachers (Singh and Billingsley, 1996; Howard and Johnson, 2004). Finally, no evidence was provided in the data about the effectiveness of policies and practices that school colleagues can implement to support teachers' well-being in terms of exchange and coordination for teaching (RQ 4).

The identification of factors with a positive effect might enable schools to undertake actions to enhance teachers' well-being. The school working environment factor indeed plays an important role in teachers' well-being, and has to do with the school availability of resources, and/or local practices and policies. Results like this one provide evidence that there are indeed factors influencing teachers' well-being which can be optimized at the school level. These findings are in accordance with the theoretical framework described by Bakker and Demerouti (2014).

Findings also support the idea that principals play a pivotal role in steering the direction of their school which requires guiding the day-to-day business of the school including matters associated with both students and teachers. Furthermore, the growing autonomy of Italian schools compared to the ministerial and regional authorities is creating confused dynamics with respect to organizing school activities. This element became particularly evident with the latest educational reform (Law 107/2015), which increased the level of autonomy of the principal with respect to the evaluation and management processes inside the school. From a practical point of view, our results highlight the importance of the "school effect," in terms of teachers' perception of having both material and staff resources at their disposal in work context. At the same time, the possibility of continuous professional development, and involvement in the management processes in school environment result in influencing teachers' satisfaction and well-being. It is therefore essential that school leaders try to create a stimulating and supportive environment in order to promote the professional growth of their teachers.

#### LIMITATION AND FUTURE RESEARCH

This is an exploratory study which provides some initial evidence on factors affecting well-being among Italian teachers. However, the present study has several limitations. First, the high rate of missing values in some variables forced authors choices of variables to consider in the analysis and aggregation of the variables at school level instead that at teacher level. Second, the availability of data related to a single wave of the PISA survey and to a single large scale assessment survey did not allow us to generalize the results over time. Since we were interested in the assessment of the effect of organizational aspects, practices and policies in affecting teachers' well-being, we focused on influences of the organizational context controlling for teachers' heterogeneity with respect to professional and personal details using cross-national data. Moreover, information related to the investment in professional development actions has been summarized in general terms in the PROPDT20 variable, using responses related to different practices. A deeper analysis of the effect of different actions described in the items on wellbeing dimensions would be recommended in order to address teachers and principals in supporting policies which can provide the greatest value added. Considering the fact that our sample consisted of teachers in secondary schools, it would be interesting to test the consistency of the results in other educational frameworks and other countries. For this reason, we believe that

#### REFERENCES


futures researchers should include additional future waves of the PISA survey and should consider teachers of different school levels and countries.

### DATA AVAILABILITY

Publicly available datasets were analyzed in this study. This data can be found here: www.oecd.org/pisa/data/.

#### ETHICS STATEMENT

All participants gave their written informed consent before the administration of the questionnaire, in accordance with the Declaration of Helsinki. The study was carried out in accordance the rules of AIP (Associazione Italiana di Psicologia—Italian Association of Psychology), according to which there was no need for previous ethics approval, since it would not deal with animals or vulnerable groups, or would involve risk for the well-being of participants, or use biomedical devices, or invasive investigation tools. Our study did not need ethics approval, according to our national regulations as well as to the Ethics Committee of the University of Cagliari.

# AUTHOR CONTRIBUTIONS

BB developed the research project, with the contribution of IS, MP, and MT. MP prepared the data set. IS carried out the data analysis. BB reviewed the literature. MT reviewed and edited the paper.


Day, C., Kington, A., Stobart, G., and Sammons, P. (2006). The personal and professional selves of teachers: stable and unstable identities. Br. Educ. Res. J. 32, 601–616. doi: 10.1080/01411920600775316

Day, C. (2004). A passion for teaching. Routledge. London and New York


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The handling editor declared a past collaboration with one of the authors BB.

Copyright © 2019 Barbieri, Sulis, Porcu and Toland. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Organizational Climate and Teachers' Morale: Developing a Specific Tool for the School Context – A Research Project in Italy

Daniela Converso<sup>1</sup> , Michela Cortini<sup>2</sup> , Gloria Guidetti<sup>1</sup> \*, Giorgia Molinengo<sup>1</sup> , Ilaria Sottimano<sup>1</sup> , Sara Viotti<sup>1</sup> and Barbara Loera<sup>1</sup>

<sup>1</sup> Department of Psychology, University of Turin, Turin, Italy, <sup>2</sup> Department of Psychological, Humanistic and Territorial Sciences, D'Annunzio University of Chieti–Pescara, Chieti, Italy

The school context is exposed to several demanding factors relating to student and family needs and external evaluative processes of students' learning and process outcomes, such as abilities in planning training courses and a learning environment. However, there is a need to develop tools that adequately support schools in making self-assessment evaluations of the internal organizational climate and teacher morale (TM). The present study proposes an Italian version of the School Organizational Health Questionnaire (SOHQ), developed by Hart et al. (2000). An Italian version of the SOHQ was deployed to 9 public primary schools in the north of Italy, and 325 cases were eventually retained as being valid for the analysis. Using confirmatory factor analysis, results highlight that a 56-item version is model fit and presents satisfactory psychometric properties, demonstrating the suitability of a latent structure composed of 12 intercorrelated factors. The present study gives further insight into increasing the use of self-assessment tools in the development of good practices and the monitoring of teacher morale within the school context.

#### Edited by:

Paula Benevene, Libera Università Maria SS. Assunta, Italy

#### Reviewed by:

Amelia Manuti, University of Bari Aldo Moro, Italy Massimo Miglioretti, University of Milano-Bicocca, Italy

> \*Correspondence: Gloria Guidetti gloria.guidetti@unito.it

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 27 June 2019 Accepted: 03 September 2019 Published: 20 September 2019

#### Citation:

Converso D, Cortini M, Guidetti G, Molinengo G, Sottimano I, Viotti S and Loera B (2019) Organizational Climate and Teachers' Morale: Developing a Specific Tool for the School Context – A Research Project in Italy. Front. Psychol. 10:2132. doi: 10.3389/fpsyg.2019.02132 Keywords: school climate, teachers' morale, teachers' wellbeing, confirmatory factor analysis, adaptation, validation

# INTRODUCTION

In recent years, schools have been required to adapt and meet new demands coming from society, as well as from students and their parents. Coherently, teachers' activities and roles become more differentiated, frequently updated, and "sophisticated," especially in the management of their interaction with students.

Therefore, research on teachers' wellbeing has largely focused on the conditions of stress and burnout related to organizational conditions as well as to the relationships with colleagues and also those with students and their families (Kyriacou, 2001; Cordeiro Castro et al., 2002; Salanova et al., 2005; Chaplain, 2008; Favretto and Rappagliosi, 2009; Avanzi et al., 2012; Velasco et al., 2013; Capone and Petrillo, 2016). However, the excessive relevance paid to the negative aspects of the teaching work and the consequences in terms of psychological distress has mostly neglected the role of positive aspects that could influence wellbeing. Indeed, as outlined by Hart et al. (1995), and according to the job demands-resources model (Demerouti et al., 2001; Hakanen et al., 2006), positive working experiences may exert independent roles on wellbeing outcomes (Martini et al., 2015), emphasizing the link between

organizational climate and teachers' health (e.g., Wearing et al., 1990; Hakanen et al., 2006; Skaalvik and Skaalvik, 2011; Benevene and Fiorilli, 2015; De Stasio et al., 2017).

Research has also shown how teacher wellbeing could be an important determinant for student learning outcomes and wellbeing, highlighting the importance of establishing not only the determinants of teachers' negative outcomes, but also those of the positive ones associated with their work (Rowe, 1992; Denny et al., 2011; Maxwell et al., 2017; Benevene et al., 2018; Guidetti et al., 2018; Lei et al., 2018). Moreover, as highlighted by the health promoting schools perspective, an extension of the Ottawa Charter holistic definition of health promotion (World Health Organization [WHO], 1986), teachers' wellbeing is part of a wider process able to sustain, at the same time, the health of children, adolescents, and the entire school community (Moon, 1999; Mitchell and Sackney, 2000; Konu and Rimpelä, 2002).

Among the factors that could be identified as achieving positive occupational wellbeing outcomes, such as job satisfaction, engagement, or commitment, the concept of morale is one of the least studied. This concept refers to the presence of energy, persistence, cohesion, and cooperation, that reflects a positive psychological state of mind (Hart et al., 2000) and can represent an important indicator of individual and group wellbeing (Peterson et al., 2008). Concerning the school context, teacher morale (TM) should be regarded as the professional interest and enthusiasm displayed toward the achievement of individual and group goals within the school setting (Bentley and Rempel, 1980). Therefore, it represents a form of a positive mental and emotional state that has the power to establish the character of a school. In addition, TM is considered one of the leading factors in determining the best functionality of a school (Eboka, 2017). In spite of this, past studies developed a measure of TM that was referred to as "a number of heterogeneous items related to adjustment and mental health" (Doherty, 1988, p. 72) that are more likely to derive dispositional levels of negative effects, instead of positive ones. Hart and Conn (1992) stated, on the contrary, that morale is a construct associated with enthusiasm, energy, and group spirit that teachers feel as a unique experience at work. Moreover, as stated by Hart et al. (2000), measurement tools, like the staff morale questionnaire (Meaney and Smith, 1988), failed to differentiate the causes from the manifestations of morale. According to Evans (1992, p. 832), TM could be defined as "a state of mind determined by the individual's anticipation of the extent of satisfaction of those needs which s/he perceives as significantly affecting her/his total work situation." Based on this definition, TM is thus intended to evaluate a future-oriented perspective, and in line with this, it is of upmost importance to identify the conditions under which TM could be fed.

In order to understand factors that could affect TM, the School Organizational Health Questionnaire (SOHQ) was developed by Hart et al. (2000) in Australia, and mainly adopted in English speaking countries (Stewart et al., 2004; Burns and Machin, 2012) and China (Wang, 2009; Wang et al., 2013). Recognizing the elements that are typically addressed by organizational development processes, the questionnaire was intended to measure different aspects that characterize a healthy organizational school climate, moving on from the assumption that organizational factors are much more important than specific classroom factors in influencing TM. Organizational climate has been defined as a multidimensional construct that refers to a variety of individual evaluations of the work environment (James and James, 1989; Neal et al., 2000). Therefore, it is based on how individuals attribute meaning to their organizational environment, and, in the school context, may concern the interpersonal relationships, such as staff affiliation and student support and organizational behavior and aspects of human resource management, such as appraisal and recognition, professional growth, and role clarity. Identifying factors of a healthy school environment thus serves in taking into account the complexity of a school as a social system where different roles – administrative, teaching, and learning – constantly interact, and in monitoring the school's effectiveness in performing its various functions (Tsui and Cheng, 1999).

The SOHQ (Hart et al., 2000) has then been developed in order to evaluate a healthy school environment, tackling some limitations of past instruments [see, for example, the School Level Environment Questionnaire (SLEQ, Rentoul and Fraser, 1983); the Organizational Climate Description Questionnaire (OCDQ, Haplin and Croft, 1963); or the Organizational Health Inventory (OHI, Hoy and Feldman, 1987)]: the restricted focus on social interaction between teachers and the principal proposed by the OCDQ (Haplin and Croft, 1963) and the absence of human resource management characteristics within the OHI (Hoy and Feldman, 1987) and the SLEQ (Rentoul and Fraser, 1983), such as appraisal and recognition, opportunities for professional growth, and role clarity, which past studies have shown to be central elements for teachers' occupational wellbeing (Kelchtermans and Strittmatter, 1999; Kokkinos, 2007).

Therefore, the SOHQ (Hart et al., 2000) has been developed with the aim of providing a broader emphasis on aspects concerning organizational behavior and human resource management within schools and, differently from the other tools, it could represent a more tailored, context-specific questionnaire for assessing teachers' needs. The eleven organizational dimensions that authors highlighted, after their preliminary studies, for defining a healthy organizational school climate are the following: Appraisal & Recognition (having feedback, being encouraged), Curriculum Coordination (interprofessional collaboration and contacts), Effective Discipline Policy (agreement on discipline and rules between teachers), Excessive Work Demands (workload perception), Goal Congruence (commitment, clear and agreed values and aims), Participative Decision Making (perception of being included in school policies), Professional Growth (satisfaction of personal and professional development in the school), Professional Interaction (social support and positive interpersonal climate), Role Clarity (having clear role expectations), Student Orientation (positive climate toward students), and Support Leadership (quality of school leadership).

The present study aims to propose an adaptation and validation of the SOHQ (Hart et al., 2000) for the Italian context and a tool for assessing systematically different factors related to the school climate and teachers' wellbeing and for better understanding criticisms and resources that can reduce or strengthen students' wellbeing and learning outcomes.

# MATERIALS AND METHODS

fpsyg-10-02132 November 1, 2019 Time: 13:51 # 3

#### Data Collection and Ethics Statement

Data were collected through a survey, using a cross-sectional and non-randomized design. Teachers from nine public primary schools in a region of Northern Italy were involved in this survey. The study database originally included 407 participants (22% of the entire population), but 82 of them avoided responding to some items. The final dataset used in the analysis consisted of 325 valid cases. According to the target population specificity, most participants were female (93.8%), with a mean age of 44.7 (SD = 9.7) and an average length of experience of 20 years (SD = 11.2). Moreover, 70.6% of the sample was composed of married persons with children, and 17.8% had at least one adult relative to care for. Generally, participants declared that they were quite satisfied with their work and appreciated their life in general: the mean judgments were equal to 7.9 and 8.2, respectively, on a scale of ten points.

Data were gathered using self-report questionnaires distributed to all employed teachers during working hours. To ensure anonymity, teachers were instructed to enclose the completed questionnaire in an envelope and leave it in a box placed by the researchers in each school. The participants volunteered for the research without receiving any reward and agreed to anonymously complete the questionnaire, signing the informed consent forms, which were placed in a box before completing the questionnaire. The research conforms to the Declaration of Helsinki of 1995 (as revised in Edinburgh, 2000), and all ethical guidelines were followed, as required for conducting human research, including adherence to the legal requirements of the country under study. Additional ethical approval was not required because no treatment was involved, including medical, invasive diagnostics, or procedures causing psychological or social discomfort for the participants.

#### Measures

The self-reported questionnaire included a sociodemographic section and the SOHQ (Hart et al., 2000). The original version of the SOHQ consisted of 57 Likert-type items (from 1 = completely disagree to 4 = completely agree), grouped in 12 sub-dimensions: Morale (M, 5 items), Appraisal & Recognition (AR, 6 items), Curriculum Coordination (CC, 4 items), Effective Discipline Policy (EDP, 4 items), Excessive Work Demands (EWD, 4 items), Goal Congruence (GC, 5 items), Participative Decision Making (PDM, 4 items), Professional Growth (PG, 5 items), Professional Interaction (PI, 7 items), Role Clarity (RC, 4 items), Student Orientation (SO, 4 items), and Support Leadership (SL, 4 items). All items included in the version presented in the study by Hart et al. (2000) were translated from English into Italian, following the International Guideline on Test Adaptation (International Test Commission [ITC], 2015).

On the basis of previous literature (Dolbier et al., 2005; Kristensen et al., 2005), job satisfaction and general life satisfaction were measured by a single item each [i.e., "Taking everything into consideration, how satisfied do you feel with your job (or life) as a whole?"] Each item was rated on a 10-point scale that ranged between 1 (extremely dissatisfied) and 10 (extremely satisfied).

#### Statistical Analyses

The SOHQ items were preliminarily analyzed with descriptive statistics; Cronbach's alpha coefficient was calculated in order to assess scale reliability whereas the contribution of each item to internal consistency was examined by the item–total correlations.

SOHQ dimensionality was investigated by confirmatory factor analysis (CFA), which is recommended over exploratory factor analysis (EFA) because it allows testing of whether the data fit a structure when there is an a priori hypothesis regarding dimensionality (Floyd and Widaman, 1995); the SOHQ questionnaire was designed to measure TM and 11 dimensions of school organizational climate which were both theoretically and empirically distinct.

The model fit was considered acceptable if the following criteria were satisfied: root mean square error of approximation (RMSEA) <0.08; comparative fit index (CFI) >0.90; and standardized root mean square residual (SRMR) <0.08 (Hu and Bentler, 1995, 1999).

The analyses were performed using the IBM SPSS Statistics 25.0 and Mplus7 software programs.

# RESULTS

Descriptive statistics for the items are shown in **Table 1**. For all items, the corrected item–total correlation achieved values equal or greater than r = 0.40, except for item 45 that reported a value of 0.29. All skewness and kurtosis values are included in the range from –1.0 to +1.0.

The subscales homogeneity was adequate: Cronbach's alpha (α) coefficients ranged from 0.7 to 0.91, with the exception of Role Clarity (RC). The items in this sub-dimension resulted in being less coherent because of item 48, which refers to others' expectations and, in fact, after its deletion, the α coefficient was 0.70, and it became acceptable. In the RC subscale, this was the only item that referred to others, and so its content was quite different and specific (item–total correlation 0.288).

Given its conservative nature, at first this item was included in the CFA, but its performance was quite scarce: a loading of 0.36 and an R-squared of 0.13. Consequently, item 48 was omitted in order to improve the measurement of the "role clarity" factor.

A model based on the restricted set of 56 items resulted adequate, with all fit indexes being at least satisfactory (RMSEA = 0.07, SRMR = 0.05, CFI = 0.90), supporting the adequacy of the 12-factor structure for the SOHQ in the Italian sample. All parameters estimated in the confirmatory analyses were high (loadings from 0.40 to 0.89) and significant at the 0.001 level (**Figure 1**).

The estimated correlations showed that there was no significant relationship between Excessive Work Demands and



TABLE 1 | Continued


Curriculum Coordination (r = 0.09, p > 0.68) and RC (r = 0.02, p > 0.77). Excessive Work Demands was significantly correlated with the other nine factors (p < 0.05), but the correlations, ranging between 0.12 and 0.37, were relatively small compared to other correlations between sub-dimensions. The correlations among the other 11 factors were in fact all significant at the 0.001 level and ranged between 0.28 and 0.84 (**Table 2**). Focusing on the Morale sub-dimension, it was strongly correlated with Effective Discipline Policy (r = 0.81), Goal Congruence (r = 0.78), Participative Decision Making (r = 0.76), and Professional Interaction (r = 0.77), and, although statistically significant, the correlation was weaker with Professional Growth (r = 0.28) and RC (r = 0.28).

Finally, all the SOHQ sub-dimensions significantly correlated with job satisfaction at the 0.001 level, with r values ranging between 0.24 and 0.45. On the other hand, only four SOHQ sub-dimensions correlated with general life satisfaction at the 0.001 level (i.e., Morale, Appraisal and Recognition, Professional Interaction, and Role Conflict), whereas Curriculum Coordination, Effective Discipline Policy, Student Orientation, and Excessive Work Demands correlated at the 0.05 level with general life satisfaction; the remaining sub-dimensions reported non-significant associations.

#### DISCUSSION

The aim of the present study was to adapt the SOHQ to an Italian school context. CFA provided evidence for the adequacy of psychometric properties, highlighting that the SOHQ represents a valid tool for monitoring staff perceptions of the school climate and their morale. Therefore, these findings confirmed those of previous healthy promoting school (HPS) surveys, which have adopted the SOHQ with efficacy (Stewart et al., 2004).

Nowadays, investing in positive school climates is a key issue in schools, for the reciprocity of teacher and student wellbeing. The HPS "ethos" (Penney et al., 2017) highlights the positive gain circle where healthy school environments improve students' growth and resilience, and, therefore, resilient students may strengthen future communities.

In the Italian context, where the present study has been conducted, schools are subject to different standardized processes of external evaluation [such as the PISA (Program for International Student Assessment) and the systematic assessment

performed by the INVALSI (Istituto Nazionale per la Valutazione del Sistema di Istruzione e Formazione/National Institute for the Evaluation of the Education and Training Education System)] that are intended to assess, on one hand, students' learning outcomes, and, on the other one, the quality of process outcomes, such as ability in planning training courses and a learning environment with inclusion and differentiation criteria of students' needs. Despite these aims, the evaluation is only focused on students' learning outcomes, but it could be helpful to analyze, along with these, the systemic conditions that represent the general frame where those outcomes are developed, like organizational and managing practices, leading processes, strategic orientation and the organization of the school in the identification of the objectives and their monitoring, human resource development and valorization (Aldridge and Fraser, 2018), and the integration with the social context, with the same perspective as HPS projects. In this vein, the SOHQ represents a valid tool for schools looking to evaluate several aspects of the organizational climate and consider teachers' wellbeing at the same time. Results from this study have shown that all the organizational climate factors significantly correlate with morale, evidencing moreover a strong correlation with the aspects of organizational behavior and human resource management (e.g., professional growth, goal congruence, and participative decision making) which, differently from other questionnaires, have been specifically assessed by the SOHQ. Moreover, it has been evidenced that significant correlations emerged between SOHQ sub-dimensions and job satisfaction. Therefore, it can be stated that all the aspects of the school environment evaluated using this questionnaire represent relevant factors that are able to intercept the quality of working life and the wellbeing of teachers, thus representing a valid tool for school leaders in managing teachers' work and planning school courses.

Although this study also gives an important insight into the usefulness of this measurement tool for the Italian context, the sample is not representative of the entire Italian teaching population. Future studies could overcome this limit by using this tool on a larger sample that is representative of the larger Italian school system, widening its scope, for example, to the other two teaching levels – middle and high school – and comparing public and private school teacher perceptions of organizational climate.

Moreover, the SOHQ should be adapted to other non-English speaking countries in order to evaluate its psychometric characteristics and to improve cross-cultural comparisons.


All correlation significant, p < 0.01, with the exceptions of those indicated n.s. – not significant. M – morale; AR – appraisal and recognition; CC – curriculum coordination; EDP – effective discipline policy; EWD – excessive work demands; GC – goal congruence; PDM – participative decision making; PI – professional interaction; SO – student orientation; SL – supportive leadership; RC – role clarity; JS – job satisfaction; LS – life satisfaction.

# DATA AVAILABILITY

fpsyg-10-02132 November 1, 2019 Time: 13:51 # 6

The datasets generated for this study are available on request to the corresponding author. The School Organizational Health Questionnaire is subject to copyright and cannot be used without the permission of Insight SRC Pty Ltd (info@insightsrc. com.au).

#### REFERENCES


#### AUTHOR CONTRIBUTIONS

DC, BL, GG, GM, and SV contributed to research conception and design, and acquired the data. DC, MC, and BL involved in critical revision of the article content. BL, GM, and SV performed quant data analysis. All authors interpreted the data and drafted the manuscript.


Meaney, D., and Smith, K. R. (1988). The assessment and development of staff morale. Pract. Adm. 10, 14–17.


from a comprehensive 'health promoting school'. Approach. IJMHP 6, 26–33. doi: 10.1080/14623730.2004.9721936


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The handling Editor declared a past collaboration with one of the authors MC.

Copyright © 2019 Converso, Cortini, Guidetti, Molinengo, Sottimano, Viotti and Loera. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# The Effect of Positive Working Conditions on Work Engagement and Teaching Classroom Practices: A Large Cross-Sectional Study in Switzerland

#### Loredana Addimando\*

Center for Innovation and Research in Education, Department of Teaching and Learning, University of Applied Science and Arts of Southern Switzerland, Locarno, Switzerland

Introduction: Teaching methods and techniques represent important classroom practices that promote both academic achievement and the efficacy of learning processes: the use of a complex array of teaching practices by educators in the classrooms has been frequently associated with better student academic outcomes (Hattie, 2012, 2009). So far, less is known about the psychological aspects linked to the working endeavors able to promote the adoption of different teaching techniques in educational contexts.

#### Edited by:

Wong Yau Ho Paul, Tung Wah College, Hong Kong

#### Reviewed by:

Dan Yu, Hong Kong Institute of Technology, Hong Kong Edith Leung, The Education University of Hong Kong, Hong Kong

> \*Correspondence: Loredana Addimando loredana.addimando@supsi.ch

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 06 June 2019 Accepted: 02 September 2019 Published: 20 September 2019

#### Citation:

Addimando L (2019) The Effect of Positive Working Conditions on Work Engagement and Teaching Classroom Practices: A Large Cross-Sectional Study in Switzerland. Front. Psychol. 10:2129. doi: 10.3389/fpsyg.2019.02129 Aim: The present cross-sectional quantitative research aimed at estimating the cumulative network of associations between psychological resources at work (e.g., social support, sense of autonomy, and job satisfaction) and both work engagement and the use of teaching classroom practices.

Sample and Method: The sample was composed of in-service teachers (N = 1,370) recruited in primary and lower secondary public schools in urban and suburban areas of Canton Ticino (Switzerland). Self-administered quantitative measures were used to evaluate the following constructs: social support, job satisfaction, teachers' autonomy, work engagement, and teaching classroom practices. The data were analyzed using structural equation modeling with decomposition analysis of total effects in direct and indirect effects.

Results: The hypothesized structural model reported a good fit with the empirical data (normed χ 2 (NC) = 2.7; root mean square error of approximation = 0.051, normed fit index (NNFI) = 0.951, non normed fit index = 0.950, comparative fit index = 0.968) and support the interpretation of the structural model. The main results revealed a large positive association between psychological resources at work and teaching classroom practices (β = 0.37). The level of work engagement seemed to be a minor element in promoting the use of teaching classroom practices (β = 0.10).

Discussion: Overall, the results provide a fresh piece of evidence for the association between working conditions and teaching practices. In fact, teachers who perceive a

**95**

supportive and satisfying working environment (both in terms of internal and external resources) are more likely to be engaged in their activities, and this in turn leads to a more heterogeneous array of teaching practices with students. The practical implications in terms of professional training and on-field intervention are discussed.

Keywords: teaching classroom practices, teacher job satisfaction, social support, work engagement, selfdetermination theory

#### INTRODUCTION

fpsyg-10-02129 September 19, 2019 Time: 10:16 # 2

The teaching and learning processes in the school setting are a typical object of study in many disciplines of the human and social sciences. Psychology, sociology, and pedagogy agree that teacher resources, competencies, skills, and behavioral repertories, along with students' needs, are considered critical points of paramount importance for research in the education field. Being an effective teacher today requires the implementation of different teaching strategies and techniques (e.g., classic lecture, case-study, simulation, and discussion) to achieve academic goals and support students' learning (Wray et al., 2000). From a formative point of view, teachers' teaching skills are conceptualized as the ability to implement different practices depending on the specific needs of the students (Stronge, 2018). Coherently with the idea of studying factors affecting the experience of being a professional teacher, the scientific literature has mainly focused on understanding their professional characteristics as being indicative of their conceptions and professional competencies (Park and Oliver, 2008; Fiorilli, 2009). Other research has been conducted on linking practices to skills learned by students as an indication of students' progress (e.g., OECD programs for international student assessment, such as the PISA project). That said, fewer studies have focused on exploring the psychological, contextual, and organizational features of teachers' work that might promote or inhibit the adoption of a variety of teaching practices and techniques. Although the scientific literature extensively examines the connection between teacher practices and classroom characteristics (Allen, 2010; Dudek et al., 2019), teacher beliefs (Farrell and Guz, 2019; Mills et al., 2019), or the presence of technological devices (Uerz et al., 2018; Nelson et al., 2019), much remains to be learned about psychological resources that might promote or inhibit the use of practices in classroom teaching. Studying how teaching strategies are linked to psychological and environmental resources is critical to determining which strategies to favor over others and expand the implementation of flexible teaching practices in empirical contexts.

According to the paradigms of positive psychology and by adopting the theoretical framework of self-determination theory (SDT) (Ryan and Deci, 2000), the present research aims at exploring the patterns of association among the variety of practices adopted in the classroom, psychological resources related to contextual working conditions (i.e., social support, sense of autonomy, job satisfaction), and levels of work engagement. This is the focus because a satisfied and engaged person not only performs better in teaching (Bakker et al., 2008) but also have a subsequent impact on students' learning (Mercer, 2010). The expected results are aimed at identifying a model capable of evaluating psychological resources in relation to the use of a wide array of teaching practices in the classroom.

#### Literature Review

The study of teaching practices is the cornerstone and antecedent of studies on the evaluation of the effectiveness of teaching and learning processes. Longitudinal research at the international level has reported that improvement in education and the academic success of students is closely related to the daily processes that take place in the classroom (Willis, 2017; Zierer and Wisniewski, 2019). In other words, a connection between the daily training practice and the scholastic and learning results of the students has steadily emerged from the field.

In recent years, Hattie (2012, 2008) and later Masters et al. (2015) conducted over 800 meta-analyses on a set of 52,637 international scientific studies (i.e., mainly in the Anglo-Saxon context) to identify the aspects that, more so than others, have a positive impact on learning. Specifically, Masters et al. (2015) studied the teacher–student relationship, identifying the following 12 elements (hierarchically ordered on the basis of their impact-effect size) able to influence learning in a direct way: first, the expectations of the students (+1.44), followed by class discussions (+0.80), the clarity of the teacher (+0.75), the ability to give feedback (+0.75), skills (+0.62), classroom management (+0.52), cooperative learning (+0.40), homework (+0.29), exercise/rest ratio (+0.28), group skills (+0.12), rejection (−0.13), and mobility (school change, −0.34). In addition, the study identified the following five emerging characteristics of "excellent teachers" (see Masters et al., 2015): the ability to identify the essential representations of their students; the ability to guide learning through the interactions that take place in the classroom; the ability to provide continual feedback, and the ability to demonstrate empathy and express positive emotions toward students and families; and finally, the ability to influence students' results and to know how to "make a difference." All these aspects can be considered complex constructs resulting from interactions among cognitive aspects, emotions, empirical experiences, and attitudes of teachers.

Another interesting line of research has focused on the relationship between students and their teachers as a core critical determinant in the development of teachers' behavior and performance (Douglas et al., 2008; Pianta et al., 2008). In this sense, the goodness of fit theory (Thomas and Chess, 1986; Lerner and Lerner, 2018) has long described how the

teaching/learning process is, to a large extent, influenced by the student–teacher degree of fit and relationship. The participation of the students and the flexibility of the training devices can be the focus of both educational objectives (Saravani and Haddow, 2011) and interactive orientation to competence, which refers to the way in which teachers and students work together on ideas and knowledge or mis-knowledge (Molinari et al., 2013). In particular, innovative teaching practices must be studied in their developmental contexts (Honigsfeld and Dove, 2019) since adopting innovations in teaching depends on the sustainability of contextual conditions as well as the attitude of the teachers, which is also the result of their current teaching experience (Ghavifekr and Rosdy, 2015). Teaching practices are therefore also influenced by a teacher's expectations, and the concepts become powerful mediators of the teacher–student relationship (Roorda et al., 2011). In fact, the extent to which teachers feel free to use new and innovative teaching practices is determined not only by their professional skills and knowledge but also by other psychological drivers (e.g., motivation and feelings of efficacy; Pyhältö et al., 2012).

#### Self-Determination Theory: Social Support, Job Satisfaction, Teachers' Autonomy, and Work Engagement as Shaping Elements of Teacher Practices

In recent years, positive psychology (i.e., the scientific study of human resources and optimal functioning; Seligman and Csikszentmihalyi, 2014) has gained increased attention in educational contexts. This approach has tried to counterbalance the traditional focus of psychology on disease, damage, or disorder by paying particular attention to the functional facets of the human experience. These recent shifts have also been adopted by applied psychology to "study positively oriented human resource strengths and psychological capacities that can be measured, developed, and effectively managed for performance improvement in today's workplace" (Luthans, 2002, p. 698). One important contribution that emerged from this new conceptual wave was SDT (Ryan and Deci, 2000) as a general functioningfocused framework for the comprehension of people's behavioral tendencies and innate psychological needs. In addition, the SDT "is concerned primarily with explicating the psychological processes that promote optimal functioning and health" (Deci and Ryan, 2002, p. 262). In other words, SDT represented a general approach to the comprehension of human behaviors that was based on organismic metatheory and that emphasized the centrality of evolved inner psychological resources for individual development in relation to behavioral self-regulation (Ryan et al., 1997). The core idea of STD is that human beings "are assumed to be active, growth-oriented organisms who are naturally inclined toward the development of an organized coherence among the elements of their psychological makeup and between themselves and the social world" (Deci and Ryan, 2002, p. 262). From this perspective, psychological needs represent the basic elements for self-motivation and goal-oriented behaviors in a given environment and provide a possible explanation for context-specific behavioral decisions. Some examples of empirical application of SDT in educational settings include studies on student motivation (Reeve et al., 2018), teachers' best practices (Stupnisky et al., 2018) and innovative behaviors (Klaeijsen et al., 2018), leadership (Eyal and Roth, 2011), classroom processes (Pianta and Hamre, 2009) and teaching style (Bartholomew et al., 2018), and student violent behaviors (Assor et al., 2018). According to SDT, human beings can experience different types of motivation with respect to their work. The presence of the different types of motivation is important given that, compared to controlled regulation (introjected and extrinsic motivation), autonomous regulation (intrinsic and identified motivation) leads to more positive individual and organizational outcomes (Slemp et al., 2018). From this point of view, teachers' experiences seem to be deeply affected by perceived working conditions and the presence of specific psychological resources. In addition, SDT provided an integrated theoretical framework able to conceptualize what stances stand behind a behavior and how people make sense of their own and others' behavior (Deci et al., 1989). As a result of such important conceptual innovation, a full spectrum of new constructs emerged, supporting both the comprehension of the educational field and teachers' behavior in the workplaces.

More specifically, SDT identified three general psychological domains able to influence external behaviors: need for competence (Elliot et al., 2002), social relatedness (Baumeister and Leary, 1995), and autonomy (Deci and Ryan, 1987). The need for autonomy refers to the need to have a sense that one can exercise free will and that any activity one undertakes is freely chosen rather than imposed. The need for relatedness is the need to feel a sense of belonging, simply to be loved and cared for (Van den Broeck et al., 2010). Finally, the need for competence is the need to develop mastery over tasks that are important to one (Deci and Ryan, 2002). These areas were saturated by a constellation of psychological constructs that appear to be of paramount importance for "facilitating optimal functioning of the natural propensities for growth and integration, as well as for constructive social development and personal well-being" (Ryan and Deci, 2000, p. 68). The present study selected two of three domains (i.e., social relatedness and perceived autonomy) presented by the SDT as specific areas linked to working conditions in the educational setting. In fact, the third domain (e.g., need for competence) seems to be more focused on internal aspects of individuals and less related to mechanisms shaping professional teaching practices. From this point of view, other constructs (e.g., job satisfaction, social support, and work engagement) appear more informative in relation to teacher work and the use of different practices. In fact, teaching work is entirely based on relationships, given that teachers engage in continuous interaction with students, families, and colleagues. For instance, Klaeijsen et al. (2018) report that job satisfaction affects both intrinsic motivation and occupational self-efficacy and that the latter strongly supports innovative behavior in classrooms. In this sense, Pindek et al. (2019) affirm that: "It is certainly not surprising that inadequate resources would adversely affect job performance and that [. . .] perhaps the most important contribution is to show that the most common harmful elements in terms of motivation do arise from colleagues and managers" (p. 91).

The term job satisfaction usually refers to the degree to which employees like the components of their job (Spector, 1997). Over time and across different paradigms, many definitions of the construct have also been adopted (Van Saane et al., 2003). Even with the variations in this broad range of available definitions, the majority of them share the knowledge that job satisfaction is essentially an effective and positive job-related reaction to the workplace (Addimando, 2013; Judge et al., 2017) that explains how people feel about their work (Kianto et al., 2016). Job satisfaction in teaching is a crucial focus for educational research, mainly because of the benefits, for both teachers and students, that "satisfied" teachers are known to contribute to organizational outcomes (Zee and Koomen, 2016).

Other important constructs were represented by both the internal support provided by professional figures such as peers, senior teachers, tutors, head teachers, or psychologists and the external support provided by private relational networks such as partners, relatives, and friends. As confirmed by Betoret (2006), the feeling of being exhausted or oppressed by the demands of the job is markedly lower when a social support network of colleagues, superiors, and tutors is available (Fiorilli et al., 2015). Social support is also associated with the sense of being related to others (Wentzel, 1998) and appears to be an important promoting factor of well-being and efficacy in schools (Collie et al., 2012).

Similarly, job autonomy was found to be significantly and positively correlated with organizational commitment and work engagement (Gillet et al., 2018; Van Wingerden et al., 2018). Teachers' job autonomy is the freedom given to teachers to make their own decisions while doing their job, whereby they elect their practices, design their tasks and/or materials, evaluate outcomes, cooperate with others to solve problems, take responsibilities for their own decisions (Balkar, 2015), get involved in organizational decision making (Tims et al., 2015), and improve their professional skills (Guglielmi et al., 2016). When autonomy is enhanced, teachers are more involved in achieving new skills, and they are more responsible for difficulties at work (Swartz and Perkins, 2016). Job autonomy has been consistently linked to employee satisfaction as a positive factor in job satisfaction (Dou et al., 2017).

Finally, throughout the decades, organizational studies have found that work engagement is defined as a positive, fulfilling, work-related state of mind that is characterized by vigor, dedication, and absorption (Lisbona et al., 2018; Carmona-Halty et al., 2019). Rather than a transitory and specific state, engagement refers to a more persistent and pervasive affective-cognitive status that is not specifically focused on any object, event, individual, or behavior. Work engagement was operationalized using three different work-related domains: vigor, dedication, and absorption. Vigor is characterized by high levels of energy and mental resilience while working, the willingness to invest effort in one's work, and persistence even in the face of difficulties. Dedication refers to being actively involved in one's work and experiencing a sense of significance, enthusiasm, inspiration, pride, and challenge. Finally, absorption is characterized by being fully concentrated and happily engrossed in one's work, whereby time passes quickly, and one has difficulties with detaching oneself from work. Accordingly, vigor and dedication are considered direct opposites of the core burnout dimensions of exhaustion and cynicism, respectively (Maslach et al., 2001). The scientific literature has demonstrated the effect of job satisfaction and teachers' autonomy on work engagement (Alonderiene and Majauskaite, 2016; Van den Broeck et al., 2016; Cassidy et al., 2017; Vangrieken et al., 2017) and the protective role of satisfaction in relation to teacher burnout (Fiorilli et al., 2015, 2017a).

The intrinsic features of teachers' job autonomy, satisfaction, and social support (as framed within the SDT) as antecedents of work engagement support the idea that such constructs could be conceived as important structural antecedents of teachers' behaviors, especially in relation to their roles in favoring positive organizational outcomes and student achievement. In addition, teaching strategies and techniques are highly influenced by the personal evaluation of psychological inner states (Stronge, 2018), and there is a lack of empirical studies focusing on how teachers' appraisals of their psychological resources linked to working environments (expressed in terms of autonomy, satisfaction, and social support) affect the engagement and teaching practices adopted by the teachers. Hence, it is relevant to study whether and to what extent psychological intrinsic inner states are related to engagement and teaching processes in a fully comprehensive model tested on empirical data gathered in real educational settings.

# Research Objectives and Hypotheses

In line with the aforementioned reviewed literature, the present study aimed to focus on the following three objectives: (1) to examine the relationship between psychological intrinsic basic factors (perceived job autonomy, satisfaction, and social support) and work engagement; (2) to examine the relationship between work engagement and teaching practices; and (3) to examine an integrated cumulative model assessing the total, direct, and indirect effects on the appraisal of "psychological resources" linked to the working environment, engagement, and teachers' direct practices in the classroom. Following that, three directional hypotheses were formulated:

H1: Perceived job autonomy, social support, and job satisfaction are directly and positively associated with work engagement.

H2: Work engagement is positively associated with the heterogeneity of educational practices adopted by teachers in the classroom.

H3: Psychological resources linked to working environment and engagement favor the use of different didactic practices in the classroom.

The full conceptual model (**Figure 1**) was designed with three latent endogenous variables and their corresponding empirical indicators. In line with social determination theory, the first latent variable was the appraisal of "psychological resources" linked to working environments, as estimated by the following three empirical measures: teachers' perceived job autonomy,

job satisfaction, and social support. Work engagement was estimated using the standard model of measurement (i.e., physical, cognitive, and emotional energy) expressed by teachers in performing their jobs (Lisbona et al., 2018). The last latent target variable (criterion) of the study included teaching practice techniques and strategies. The variable years of teaching were modeled as a potential source of covariation. This procedure allowed us to analyze segregated data, given that it is a more informative technique (Pepe and Addimando, 2014; Pearl et al., 2016; Fiorilli et al., 2017b; Pepe et al., 2018) and a means of controlling for the Yule–Simpson effect (i.e., a statistical association that holds for the full sample but is reversed in all subpopulations, Simpson, 1951).

# MATERIALS AND METHODS

#### Participants and Procedures

Participants were in-service teachers from mandatory public schools in Canton Ticino, Switzerland. The number of teachers enrolled in the study was 1,370 (37.7% of the whole population). The respondents selected in this study were teachers in preprimary (18%), primary (43%), and lower secondary schools (39%). The sample was composed of 25% of men (n = 342) and 75% of women (n = 1,028), with an average age of 42.37 (ds = 10.59) years and a seniority (mean of teaching years) of 15.19 (ds = 10.72) years. The study conducted was a quantitative, cross-sectional study, meaning that data from the participants were collected at one specific point of time. All self-reported questionnaires were administered on-site during school hours. Two trained researchers distributed the self-reported measures to the participants in approximately 50 different school settings. The mean time for completing the research protocol was close to 45 min. Inclusion criteria for involvement in the study were: (1) being an in-service teacher, (2) having worked for at least 1 year, and (3) being in charge of a primary or lower secondary classroom.

#### Quantitative Measures

#### Inventory of Teaching Practices

Teaching practices were evaluated with different tools focused on traditional teaching practices (e.g., instructional teaching, oneway lessons, and traditional class), collaborative practices (e.g., co-teaching and team teaching, and projects), and innovative teaching (e.g., ICT in teaching, differentiation, and personalized learning). Due to the unavailability of quantitative tools to observe practices in Ticino, the study used the CWSEI Teaching Practices Inventory (Wieman and Gilbert, 2014), a tool specifically designed to evaluate teaching practices used in undergraduate science and mathematics courses. Since practices refer to observable behaviors, and we did not have a set of practices defined beforehand, we preferred to integrate the CWSEI with other tools. In particular, teachers' strategies and activities were evaluated with items taken from a validated Italian questionnaire specifically intended to assess motivations, strategies, and actions in teaching/learning processes. The MESI (STRAT) of Moè et al. (2010) consists of items assessing classroom activities and personalized learning. The inventory

provides a detailed inventory of practices used in all aspects of a "lecture" course (this means that it is not suitable for use in courses that are primarily laboratory, seminar, or project courses). Before administering the tools, focus and discussion groups were conducted with teachers and principals to adapt the survey to the target context. Finally, principal component analysis was conducted to reduce the number of practices to a set of more manageable size. The results of the analysis suggested that the items can be fruitfully grouped into three separate domains: co-teaching (examples of such items were "Participate in the didactic-subject group meetings," "Teach together with another colleague," "Observe the lessons of the other teachers," and "Organize activities for students in collaboration with colleagues"), differentiation (e.g., "Use observation to identify students' needs," "Define different skills to be achieved depending on the needs of the students," "Prepare different materials to allow each student to work and learn at his/her personal pace," and "Positively reinforce students' commitment"), and classroom management (i.e., "Before explaining a new topic, refer to and discuss topics previously discussed with students," "Before starting a new topic, understand with the students what the objectives I want to achieve are," "Organize group work in classroom," "Provide summary or in-depth materials, such as summaries, work plans, etc.").

#### Work Engagement, Job Satisfaction and Autonomy

To measure teachers' psychological working engagement, the present study adopted the Utrecht Work Engagement Scale (UWES-9, Simbula et al., 2011). The nine items of the UWES are grouped into the following three subscales composed of three items in each dimension that reflect the underlying dimensions of engagement: vigor, dedication, and absorption. All items are scored on a 5-point frequency rating scale ranging from 1 (never) to 5 (always). The reliability of the score was higher than standard recommendations (α = 0.84).

Job satisfaction was evaluated with the Teacher Job Satisfaction Scale (TJSS-9, Pepe et al., 2017). The TJSS is a questionnaire aimed at measuring job satisfaction that has been specially developed for use in educational contexts. The TJSS-9 is composed of nine items grouped in three dimensions and is already translated into the target language (e.g., Italian), with three items in each dimension: satisfaction with coworkers, satisfaction with parents, and satisfaction with students' behaviors. The reliability of the scale was over the recommended threshold (α = 0.75).

The perceptions about the levels of autonomy in teaching were investigated with the Teacher Autonomy Scale (Ulas and Aksu, 2015), which identifies three areas of autonomy in teaching: autonomy in instructional planning and implementation (two items), autonomy in professional development (one item), and autonomy in determining the framework of the curriculum (two items). All items are scored on a 5-point rating scale ranging from 1 (I totally disagree) to 5 (I totally agree). The reliability of the scale was appropriate (α = 0.84).

Finally, the perceptions about social support were investigated by five ad hoc items of social support in three different domains: support from colleagues (one item), support from school management (three items), and support from parents (one item). All items were scored on a 5-point rating scale ranging from 1 (I totally disagree) to 5 (I totally agree). The reliability of the scale was in line with recommendations (α = 0.81).

# Structural Equation Modeling

Structural equation modeling (SEM) is a multivariate statistical technique that allows the simultaneous modeling of the relationship between multiple latent and empirical constructs (Byrne, 2016). Thus, SEM is usually selected for simultaneously estimating patterns of relationships between variables under study. SEM is a particular type of path analysis that produces measurement estimation with a better validity and reliability (Hair et al., 2017). In addition, it can also be used to decompose total effects in both direct and indirect effects (Kline, 2015). SEM is fundamentally a hypothesis testing method (i.e., a confirmatory approach) rather than an exploratory approach (e.g., regression analyses). In addition, the SEM technique can explicitly estimate measurement error rather than ignore this issue, as is done with other traditional techniques (Byrne, 2016; Hair et al., 2016).

The statistical significance of the model (i.e., the degree of fit between the conceptual and the empirical model) was evaluated using the following goodness-of-fit indices: root mean square error of approximation (RMSEA, RMSEA < 0.05; Hu and Bentler, 1999); standardized root mean square residual (SRMR, SRMR < 0.05) (Marsh et al., 2004); normed fit index (NFI, NFI > 0.95) (Morin et al., 2013); Tucker–Lewis index (TLI, TLI > 0.95) (Morin et al., 2013); and the comparative fit index (CFI, CFI > 0.95) (Morin et al., 2013). Mahalanobis's distance (p < 0.001) was computed to identify multivariate outliers. No extreme values were removed from the dataset. Next, the distribution of the scores was assessed to evaluate whether they were normally distributed. Given that none of the variables under study displayed kurtosis or skewness values exceeding the recommended limits [−1,+1], the maximum-likelihood method (Gath and Hayes, 2006) was adopted to estimate the parameters for the SEM analysis. In line with indications from the literature (e.g., MacKinnon et al., 2004), 95% confidence limit intervals were computed using both Monte Carlo simulation and bootstrapping methods with a set of random samples (k = 500), meaning that the indirect effects for each of the k samples and the mean value for the selected pool of samples were analyzed. The software used for all analyses was Amos 23.0 (Arbuckle, 2014).

# RESULTS

Descriptive statistics and zero-order correlations are described in **Table 1**.

From the correlational point of view, zero-order correlations were robust and in line with the expected theoretical directions. The analysis revealed interesting and high associations among work engagement subscales, autonomy, job satisfaction, and perceived social support. Specifically, higher correlations were found between job satisfaction and vigor (r = 0.370) and job satisfaction and dedication (r = 0.377). We also observed an expected positive correlation between teachers' autonomy


and job satisfaction (r = 0.176), as well as between teachers' autonomy and social support (r = 0.195). Finally, the high correlations among the subscales of teaching practices (i.e., between differentiation and classroom management practices, r = 0.621) seem to indicate the intrinsic consistency of the teaching practices dimension.

The tested structural equation model is shown in **Figure 2**.

From the statistical point of view, regarding the evaluation of the goodness-of-fit indices (relative and absolute), the model was fully supported. The model shows an excellent overlap with all the indicators (relative and absolute): χ 2 (29) = 24.8978.2, p < 0.05; NC = 2.7; RMSEA = 0.051, 95% CI = 0.037–0.044, NFI = 0.951, NNFI = 0.950, CFI = 0.968.

Starting from left to right, the total effect of the covariate variable (e.g., teaching years) revealed a small negative total effect on work engagement [β = −0.13, p = 0.010; 95% CI (−0.31 to −0.008)] and a positive small total effect on teaching practices [β = 0.11, p = 0.014; 95% CI (0.009–0.011)]. This means that the more experienced teachers tend to be less engaged in their work and to use more teaching practices. In contrast, a not statistically significant total effect was found in relation to the appraisal of psychological resources linked to the working environment [β = −0.02, p = 0.830; 95% CI (−0.008–0.006)].

With regard to the first hypothesis (H1), the model supported the idea that the appraisal of psychological resources linked to the working environment was directly associated with work engagement. In fact, the total direct effect was positive, large in size, and statistically significant [β = 0.73, p = 0.009; 95% CI (1.01–3.25)]. In particular, the analysis suggests that the higher the teachers' satisfaction, support, and autonomy are, the higher their scores on work engagement. With regard to the second hypothesis (H2), the direct effect of work engagement on teachers' practices was positive and small but not statistically significant [β = 0.10, p = 0.580; 95% CI (−0.444–0.375)], thus not providing support for H2. Finally, the analysis of the total effect of the appraisal of "psychological resources linked to the working environment" on teacher practice was positive, medium in size, and statistically significant [β = 0.45, p = 0.005; 95% CI (0.776–3.94)]. The decomposition of the total effect revealed that the working conditions were both directly [β = 0.37, p = 0.005; 95% CI (0.418–4.39)] and indirectly [β = 0.08, p = 0.032; 95% CI (0.575–0.925)] associated with the repertoire of practices. From this point of view, the analysis suggests that when teachers perceive their work environment as being supportive and satisfying (in the sense of the fulfillment of basic intrinsic psychological needs), they tend to be more engaged and use a larger array of teacher practices, in particular, differentiation [indirect effect, β = 0.39, p = 0.014; 95% CI (1.48–7.22)] and classroom management [indirect effect, β = 0.31, p = 0.011; 95% CI (0.837–4.14)]. The results provided full support for H3.

#### DISCUSSION

The present cross-sectional research investigated the relationships among psychological resources linked to the working environment, work engagement, and classroom

practices in a large sample of in-service teachers. In particular, a conceptual model was empirically tested to understand whether and to what extent the way in which teachers perceive psychological resources in their work environment (in terms of job satisfaction, social support, and autonomy) were associated with their work engagement and in turn with the variety of their teaching practices adopted in classrooms. Within the framework of self-determination theories, the rationale for the study was to understand which factors represent promoting conditions for the adoption of heterogeneous repertories in teaching, since the use of different teaching practices and techniques favors the academic outcomes of children (Rowley et al., 2018; Fix et al., 2019). All in all, our findings demonstrate that the presence of psychological resource favors the adoption of heterogeneous and rich repertories of teaching practices, meaning that when teachers perceive their work environment as supportive, satisfying and having a high degree of autonomy (i.e., from another perspective, when their basic psychological needs were fulfilled), they tend to use a larger array of teaching practices.

Concerning the relationship between the psychological resources and work engagement (H1), our results support the idea that a good appraisal of "psychological resources linked to the working environment" was a promoting factor in work engagement in terms of dedication, absorption, and vigor. These results are in line with current literature demonstrating that job resources (Bakker et al., 2007), social support (Saks and Gruman, 2018), and autonomy (Heyns and Rothmann, 2018), might positively affect levels of engagement. Contrary to our expectation (H2), the direct effect of work engagement on teaching practices was small and not statistically significant. Finally, our results support H3, demonstrating that in our sample, intrinsic basic psychological resources (e.g., satisfaction, autonomy, and social support) were more associated to the promotion of teacher practices than were more other psychological aspects (e.g., work engagement). In particular, the analysis suggested that when teachers perceive their work environment as being supportive and satisfying, they tend to be more engaged and in turn use a larger array of teacher practices with students, in particular, differentiation techniques and classroom management. These results are in line with current self-determination theories (Ryan and Deci, 2000) and highlight the role of intrinsic factors in promoting innovative teacher practices (Klaeijsen et al., 2018). Even if the aims of our study were not focused on the role of demographic variables, the structural model also suggests that the length of teaching experience years is associated with work engagement and teacher practices, with more experienced teachers showing less engagement and more use of teaching practices.

# CONCLUSION

The results of the study show how the evaluation of the psychological working conditions environment (in terms of satisfaction, social support, and professional autonomy) is crucial in promoting teachers' decision to adopt different or innovative practices. From the point of view of ecological validity, the importance of the study is twofold: one theoretical and one

practical. First, from a conceptual point of view, the data from the present study support the indications posited by the SDT that low-ordered basic intrinsic psychological needs are powerful drivers of human behaviors rather than high-ordered needs (e.g., work engagement). In fact, by linking the appraisal of "psychological working conditions" to directly observable tasks (i.e., frequency of teaching practices) of teachers' jobs, our analysis supports the direct association between the fulfillment of internal psychological states and teacher occupational behaviors in the educational context. Other attempts have been made (see, for instance, Reeve, 2002) to apply SDT to educational settings; however, such studies have focused on students' behaviors rather than on teacher practices. From this point of view, the present study represents one of the first attempts to apply SDT to the comprehension of teaching behaviors in terms of teaching techniques. Second, and perhaps more germane, our results are an important indicator in terms of planning both preservice and in-service teacher training and promoting the use of innovative techniques. From this perspective, the study shows that to increase the adoption of innovative practices, it is first necessary to work on the perception of "psychological working conditions," especially in terms of general satisfaction and perception of social support and autonomy. This means that intervention programs aimed at improving the adoption of large and flexible repertories of teaching practices should include work on teachers' satisfaction, autonomy, and social support. In organizational terms, this would imply a shift from a microsystemic level (e.g., a teacher in interaction with his/her student in a classroom setting) to a macrosystem level (e.g., a teacher in interaction with the organizational system that interacts with students in a classroom setting), putting emphasis on the need for the individual to be supported, satisfied, and more independent and self-directed.

# Limitations, Suggestion, and Future Research

As in the case of other studies, the current research has some limitations that should be addressed. First, all the data were collected by self-reported measures; thus, the presented results may suffer from the common method bias (MacKenzie and Podsakoff, 2012), meaning that some of the statistical variance can be accounted for by the use of self-reported measures rather than by the structural model. Second, it should be noted that all the participants were from Canton Ticino (Switzerland), and generalization to other populations of teachers should be made with caution. In this regard, the finding of the present study should be read more in terms of transferability to similar contexts rather than generalizability to different educational systems or cultural settings. Third, the study was limited to compulsory schools, and higher level education was not taken into consideration. Further studies are welcome to expand current knowledge about teachers' practices by replicating the present research in other educational grades. Finally, the research design was cross-sectional; consequently, assumptions about cause–effect relationships should not be made. In this regard, the author agrees with the position of Pearl et al. (2016), who state that in cross-sectional design, the relationships between variables should be considered probabilistic rather than deterministic. Beyond the limitations, the present study supported the idea that practice-oriented research and enquiry should be framed within local contexts. In fact, psychological resources linked to working environments represent a powerful "tool" for protecting teachers from the adverse effects of their profession, especially from social-related aspects of their work environments. Teachers should be trusted, supported, and empowered as professionals who can be agents of change contributing to school development. Teachers should be expected, enabled, and encouraged to collaborate; their competences and capacities, as well as their autonomy and accountability, should be considered not only individually but also collectively as part of professional teams. These aspects also affected the variety of practices used by the teachers in classrooms. Future research on teaching practices should consider exploring other psychological resources (e.g., quality of life, principal leadership, collaborative working) linked to the working environment as perceived by workers.

# DATA AVAILABILITY

The datasets generated for this study are available on request to the corresponding author.

# ETHICS STATEMENT

The research was conducted by following American Psychological Association, (2010) ethical guidelines and code of conduct. In accordance to the declaration of Helsinki (1964) ethical guidelines, a written informed consent was obtained from all participants. Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements (Ethics Committee/IRB).

# AUTHOR CONTRIBUTIONS

The author confirms being the sole contributor of this work and has approved it for publication.

# ACKNOWLEDGMENTS

The author is grateful to the Center for Innovation and Research in Education (CIRSE) for their collaboration and to the Department of Teaching and Learning, University of Applied Sciences and Arts of Southern Switzerland, for their dissemination support. This work was sustained by the Department of Education, Culture, and Sports – School Division of the Canton Ticino.

# REFERENCES

fpsyg-10-02129 September 19, 2019 Time: 10:16 # 10


and coaching. Assess. Effec. Int. 44, 81–94. doi: 10.1177/153450841877 2919




and science. CBE Life Sci. Educ. 13, 552–569. doi: 10.1187/cbe.14- 02-0023


**Conflict of Interest Statement:** The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Addimando. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Validation of a Physical Education Teachers' Self-Efficacy Instrument Toward Inclusion of Students With Disabilities

#### Raúl Reina<sup>1</sup> , Roberto Ferriz<sup>2</sup> \* and Alba Roldan<sup>1</sup>

<sup>1</sup> Departamento de Ciencias del Deporte, Centro de Investigación del Deporte, Universidad Miguel Hernández, Elche, Spain, <sup>2</sup> Department of Didactics of the Musical, Plastic and Corporal Expression, Faculty of Social Sciences and Humanities, University of Zaragoza, Teruel, Spain

Acquiring specific training in disability seems to be a key aspect for achieving school inclusion. Teachers who receive such prior training would be more prepared to address diversity in the classroom, which could be related to their perception of self-efficacy. The aim of this study was to validate the Spanish version of the Self-Efficacy Scale for Physical Education Teacher Education Majors toward Children with Disabilities (SE-PETE-D). Two hundred and eighteen in-service physical education teachers participated in this study, with a Mage = 38.06 years and Mteaching experience = 11.72 years. To obtain the three subscales resulting from intellectual, physical, and visual disabilities, several exploratory and confirmatory factor analyses were conducted. The results supported three independent models made up of three factors (intellectual, physical, and visual disabilities). The structure of the models was invariant with respect to gender, the educational stage in which the teaching was taking place, previous teaching experience, previous training, and previous experience in adapted or inclusive physical activity and sports. The subscales presented high reliability values for Cronbach's alpha, and Omega's index ≥0.81. This study provides evidence of the validity and reliability of an instrument to measure the perceived self-efficacy of physical education teachers to include students with disabilities in their classes and is the first study to be applied with in-service teachers. In addition, some methodological and conceptual limitations of the original scale are identified, opening new lines of work in relation to training situations to assess the perception of self-efficacy or the type of disability.

Keywords: professional development, special educational needs, education, scale, diversity

# INTRODUCTION

Spain ratified the United Nations (2006) on the Rights of Persons with Disabilities in 2007, whose article 24 states that the inclusion of students with disabilities in the educational system must be the rule and not the exception. Ministerio de Educación y Cultura (2013) postulates that the acquisition of key competencies must be based on an educational model that stimulates students' autonomy, with an emphasis on active and social methodologies, including the promotion of values. However, many of the training programs for Physical Education (PE) teachers in Spain lack specific training in inclusion (Reina et al., 2018), which implies a lack of didactic tools to adequately address

#### Edited by:

Florencio Vicente Castro, University of Extremadura, Spain

#### Reviewed by:

Antonio Baena Extremera, University of Granada, Spain Leonidas A. Zampetakis, University of Crete, Greece

> \*Correspondence: Roberto Ferriz roberto.ferriz@unizar.es

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 04 June 2019 Accepted: 09 September 2019 Published: 01 October 2019

#### Citation:

Reina R, Ferriz R and Roldan A (2019) Validation of a Physical Education Teachers' Self-Efficacy Instrument Toward Inclusion of Students With Disabilities. Front. Psychol. 10:2169. doi: 10.3389/fpsyg.2019.02169

diversity in the classroom (Rust and Sinelnikov, 2010). In addition, this need for training seems to be related to teachers' perception of self-efficacy, and this is the main element underlying the motivation for effective and efficient professional performance (Taliaferro et al., 2015; Tindall et al., 2016).

PE teachers' attitude toward the inclusion of students with disabilities in their classes is a widely studied construct in the specific literature (see Wilhelmsen and Sørensen, 2017) and is determined by the interaction of personal, environmental, and behavioral factors (An and Meaney, 2015). However, it has also been suggested that teachers' degree of self-efficacy (Bandura, 1997), perceived competence (Harter, 1985), or behavioral control are moderator mechanisms of the inclusive process (Ajzen, 1991). Of all of them, self-efficacy, a pillar of the socialcognitive theory (Bandura, 1997), has been considered one of the mechanisms that most determines a positive attitude and intent toward appropriate and responsible behavior in the classroom. The self-confidence or self-assurance shown by teachers in specific environments (i.e., the inclusion of students with disabilities in PE) is considered to be self-efficacy, and teachers thereby display adequate levels of professional performance, relying on their knowledge and skills (Bandura, 1997). In the context of PE, self-efficacy represents the way in which teachers adapt learning situations, adjust objectives, manage the classroom, apply the methodology, or resolve conflicts to attend to diversity in the classroom.

Given the moderator role of self-efficacy in the inclusion of students with disabilities, it is necessary to develop valid and reliable instruments for its measurement in relevant facets such as teacher training, especially regarding their continuous education. In the context of PE, one the first instruments created to evaluate teachers' self-efficacy was the Self−efficacy in Teaching Physical Education under Inclusive Conditions Instrument (SEIPE) of Hutzler et al. (2005), consisting of 15 items with vignettes [eight items for physical disability (PD), three for general development and coordination disorders, two for attention deficit and hyperactivity, and two for visual impairment (VI)], in which questions are asked about teachers' degree of confidence in their skills to create an appropriate learning environment. The Physical Education Teaching Efficacy Scale (PETES) of Humphries et al. (2012) was validated with 592 in-training PE teachers, and includes a total of 35 items organized in 7 factors, one of them called "teaching students with special needs." However, the PETES is considered a generic instrument for the evaluation of teachers' self-efficacy because the inclusion of the students with disabilities is just a part of it.

To date, one of the instruments of reference, due to the specificity of the types of disability and habitual situations it analyzes for PE, is the Self-Efficacy Instrument for Physical Education Teacher Scale (SE-PETE-D; Block et al., 2013). It is the object of discussion of this work as it is validated to the Spanish context with in-service teachers. The SE-PETE-D evaluates teachers' self-efficacy for the inclusion of students with intellectual disabilities (ID), PD, and VI. The three subscales are made up of factors relating to the teacher's selfefficacy; teaching students to help their peers with disabilities in PE [Instruction to Peers (IP)], modifying the design of a task for students with disabilities [Specific Adaptations (SA)], staying focused and helping the student with disabilities to understand what to do in the task [Staying on Task (ST)], and creating a safe environment during a PE session [Safety (S)]. The SE-PETE-D has been administered in countries like the United States (Taliaferro et al., 2015), Greece (Tekidou et al., 2015), Ireland (Tindall et al., 2016), the Czech Republic (Baloun et al., 2016; Kudlacek et al., 2018 ˇ ), or Serbia (Jovanovic et al., ´ 2014). However, many of these works were applied with nonactive in-training teachers (e.g., Taliaferro et al., 2015; Kudlacek ˇ et al., 2018; Abellán et al., 2019), so little is known about its reliability in day-to-day PE. In addition, previous studies have not provided evidence of its psychometric properties or reliability (e.g., Jovanovic et al., 2014 ´ ; Tekidou et al., 2015; Reina et al., 2016; Tindall et al., 2016) despite its application in countries with disparate demographic, cultural, and linguistic characteristics. In terms of the Spanish context, approximations have been made. On the one hand, Reina et al. (2016) reported excellent reliability values for the SE-PETE-D in a sample of 102 in-service PE Teachers. On the other hand, Abellán et al. (2019) using the version of Reina et al. (2016) in a sample of 228 university students of Childhood and Primary Education Degrees who studied the PE specialty, obtained values below the recommended ones for three fit indexes when testing the factorial and invariance structure of the scale, although it obtained acceptable reliability.

The study of PE teacher's self-efficacy and perceived preparedness to teach has usually focused on intellectual, physical or visual disabilities (Jovanovic et al., 2014 ´ ), and these are precisely the ones included in the different subscales of the SE-PETE-D. However, although the study of attitudes toward inclusion has been widely studied in these types of disability (Wilhelmsen and Sørensen, 2017), there is no evidence in PE about the perceived self-efficacy of teachers toward the inclusion of these students (Hutzler et al., 2019). In addition, it is not known whether this instrument could be applied to the moderator variables inherent to PE teachers and their performance setting when evaluating their competencies toward inclusion in PE. As to (a) teachers' sex, there is no evidence of a different level of perception of self-efficacy between men and women (Hutzler et al., 2005; Block et al., 2013; Jovanovic et al., 2014 ´ ; Reina et al., 2016; Abellán et al., 2019). Another variable of interest, analyzed at the level of attitudes but not of self-efficacy, is (b) the professional context or educational stage where the teaching is performed. The authors of this paper have not found any studies that address the potential mediating effect of this demographic variable, although they did find works applied to professionals in training (e.g., Taliaferro et al., 2015) or in service (Tekidou et al., 2015). This study involves in-service PET at primary schools (6– 12 years), secondary schools (13–16 years) and a group that teach PE and sports professional training (usually 17–18 years) but have the same qualifications as those who teaching at secondary schools. A third variable of interest is (c) the years of teaching experience, where younger generations of teachers who have been able to benefit from pro-inclusive social and educational policies may feel more prepared to attend to students with disability in PE. However, it is not known whether the same instrument

can be used for all PE teachers in general to evaluate their selfefficacy and, therefore, the future mediating effect of this variable when intervening in PE practice. There is evidence of the variable (d) previous training acquired in the attention to/inclusion of students with disability, which is sensitive to the evolution or progress of the training programs implemented (Hutzler et al., 2005; Taliaferro et al., 2015; Tindall et al., 2016). Although when using the SE-PETE-D (e.g., Tindall et al., 2016), it has been suggested that training formation (i.e., pre-service training) would be determinant to increase the level of self-efficacy, the evaluation of this construct is not the same in controlled training settings versus disparate or changing environments (i.e., inservice teachers), where the characteristics of the educational ecosystem may condition the degree of teachers' perceived selfefficacy. Fifthly, (e) real previous experience or contact with disabled students is worth mentioning, as it is determinant for an educational culture based on equal opportunities and equity (McKay, 2018). Following the theoretical postulates of selfefficacy (Bandura, 1997), it is plausible that teachers with prior and positive experience of contact with the inclusion of students with disabilities would show higher levels of competence, perceived efficacy, and better attitudes toward inclusion, whereas those with negative experiences would show frustration, low perceived competence, and reluctance to include these students. The work of Reina et al. (2016) with a group of in-service PE teachers indicates the mediating effect of this variable. Summing up, at least five key variables are identified when analyzing the self-efficacy of a PE teacher to include students with disabilities. However, the instruments used to date do not provide evidence that such comparisons can be carried out with guarantees.

Finally, the absence of previous evidence to compare the results of the SE-PETE-D with respect to demographic variables that may be the subject of future studies is added to the lack of internal consistency in terms of the four factors that make up the three subscales (i.e., ID, PD, and VI) and that would be considered dimensions of teachers' self-efficacy: IP, SA, ST, and S. Having presented the limitations of previous works with the SE-PETE-D (Block et al., 2013) and drawing on the Spanish version of Reina et al. (2016), the aim of this work was to test the validity and reliability of the SE-PETE-D with in-service PE teachers. In line with previous studies that explored psychometric properties of the SE-PETE-D (Block et al., 2013; Reina et al., 2016), first, we hypothesize that this scale will have a factorial structure with three types of disability (ID, PD, and VI) and maintain the original dimensions of teachers' self-efficacy. Second, we hypothesize that three subscales will be invariant to the following sociodemographic variables: (1) gender, (2) educational stage, (3) teaching experience, (4) previous training, and (5) experience in adapted physical activities or adapted sports. Third, it is expected that the scale will exhibit adequate levels of internal consistency.

# MATERIALS AND METHODS

#### Participants

Participants were 137 men and 81 women (Mage = 38.06, SD = 8.17) from 103 different locations and 177 education centers. They taught subjects related to PE in schools in Spain in the stages of Compulsory Primary Education (N = 106), Compulsory Secondary Education (n = 81), or the higher degree cycle of Animation of Physical and Sporting Activities (n = 31). Participants reported their years of teaching experience (Myears = 11.72, SD = 7.83), training in activity/adapted/inclusive PE (Yes = 80; No = 138), and experience in physical-sport activity/adapted/inclusive PE (Yes = 142; No = 76). All the teachers signed a prior consent to data collection, endorsed by the Project Evaluation Agent of the authors ' University (DPS Reference RRS. 01.15), participating voluntarily in the study.

#### Measure

"Escala de Autoeficacia del Profesorado de Educación Física hacia el Alumnado con Discapacidad" (EA-PEF-AD-2, in English: Scale of Physical Education Teachers' Self-efficacy toward Students with Disabilities). We used Reina et al. (2016) Spanish translation of the Self-Efficacy Scale for Physical Education Teacher Education Majors toward Children with Disabilities (SE-PETE-D; Block et al., 2013). This scale (see **Supplementary Material**) begins with general instructions, the objective of the study, an explanation of the contact theory of Bandura (1997), and how to register the responses. The instrument consists of four parts: the first three parts for each of the subscales associated with ID, PD, and VI, while the last part collects demographic variables.

Each subscale is preceded by a narration (i.e., vignette) which describes situations that a student with ID, PD, or VI, respectively, would have during PE classes (e.g., skill level or way of interacting with peers). The first subscale (ID) consists of 11 questions and covers the factors of Self-efficacy regarding: IP (3 items), SA (4 items), and ST (4 items). The second subscale (PD) presents 12 items that include: IP (3 items), SA (6 items) and S (3 items). The third subscale (VI) presents 10 items with the factors: IP (3 items), SA (4 items), and S (3 items). All responses are rated on a Likert scale with a range of 1 (no confidence) to 5 (complete confidence). Higher scores indicate a higher perception of the teacher's self-efficacy to include students with ID, PD, or VI in PE classes. Each of the three subscales is organized in blocks, from 3 to 5 items, according to the teaching situations to which the scale is being applied: (a) a physical condition test, (b) the teaching of specific skills of a collective sport, and (c) the teaching of the playing dynamics of the collective sport itself.

The fourth part consists of a series of demographic questions about age, gender (male/female), years of experience as a PE teacher (number of years), whether they had received previous training in adapted/inclusive PE (yes/no), and whether they had any teaching experience in which they had to include a student with a disability in their PE (yes/no).

#### Procedure

To test the validity and reliability of the scale, the translation (Reina et al., 2016) of the instrument was used, considering all the items that Block et al. (2013) had originally proposed to capture the essence of the constructs associated with the types of disability. The objective was to obtain an instrument with theoretical and statistical support, in order to eliminate

the limitations reported in previous studies (i.e., Abellán et al., 2019). The PE teachers were contacted and informed of the objective of the investigation, and their participation was requested. The sampling was not random, because the teachers who participated were selected attending to their geographical proximity and willingness to participate. After the consent was signed, the teachers received instructions from the principal investigator about the structure of the questionnaire. Any doubts about the process of completing the scale were resolved. The teachers needed approximately 20 min to complete it. All the measurements were conducted in the second semester of the academic year, that is, between February and May.

#### Data Analysis

To determine the validity and reliability of each subscale of the EA-PEF-AD-2, exploratory factorial analyses (EFA) and confirmatory (CFA) were performed. To verify the suitability of applying EFA, the Kaiser–Meyer–Olkin (KMO) statistic was used as well as Bartlett's sphericity test, considering values of 0.70 for the KMO index (Hair et al., 1999), and a significance of p < 0.05 for Bartlett's sphericity test (Everitt and Wykes, 2001). To replicate the EFA of the study of Block et al. (2013), we used the principal components extraction method (specifying and not specifying the number of factors to be extracted) and varimax rotation. According to Stevens (1992), loading values ≥0.40 are acceptable for items in an EFA.

On another hand, the Mardia coefficient revealed that the normality distribution was not met (normalized mean = 32.15 for ID; 37.75 for PD; and 14.65 for VI) for the CFA, so the maximum likelihood method was used along with the bootstrapping procedure. The estimators were not affected by the lack of normality, so they were considered sufficiently robust (Byrne, 2001). The goodness of fit of the models was analyzed through a set of several indexes: the ratio between Chi Square and degrees of freedom (χ 2 /df), Comparative Fit Index (CFI), Incremental Fit Index (IFI), Tucker-Lewis Index (TLI), Root Mean Square Error of Approximation (RMSEA) with its 90% confidence interval, and Standardized Root Mean Square Residual (SRMR). As chi square is very sensitive to sample size (Jöreskog and Sörbom, 1993), the χ 2 /df was used, considering values <3 acceptable (Schermelleh-Engel et al., 2003). In addition, the incremental indexes (i.e., CFI, IFI, and TLI) would reveal an acceptable fit with values ≥0.95, whereas for error rates, values ≤0.08 are considered acceptable for the RMSEA and SRMR (Hu and Bentler, 1999).

The invariance of the factorial structure of the three subscales with respect to the five demographic variables of interest was verified: to accept invariance, there must be no significant differences between the model without restrictions (Model 1) and the model with invariant measuring weights (Model 2) (Byrne et al., 1989). If this criterion is not met, invariance is also accepted when the 1CFI ≤ 0.01 (Cheung and Rensvold, 2002).

In addition, descriptive statistics, bivariate correlations between items and internal consistency of the scale (Cronbach alpha index = α) and the construct (Omega index = ω) were calculated. For alpha (Nunnally, 1978) and omega (McDonald, 1981) values ≥ 0.70 are considered acceptable. In this study, we used the statistical package SPSS v. 21.0 and the SPSS Amos 21.0 (IBM Corp. Released, 2011).

When obtaining the version that reveals the best evidence of validity and reliability of the SE-PETE-D in Spanish, the following criteria were considered. Firstly, we attempted to maintain the largest number of items for each factor of selfefficacy (IP, SA, ST, and S) of the different subscales of disability groups (ID, PD, and VI), in order to obtain an instrument that adheres to the original factors of the subscales. Secondly, in the validation process, CFA was used to test the original version of the Spanish SE-PETE-D, and if it did not obtain appropriate fit indexes, EFA was used. Thirdly, to consider the results of EFA or CFA as valid, they should provide statistical and theoretical support (e.g., not fusing items of different factors in the same factor to obtain statistical support).

# RESULTS

### Factorial Analysis for the Intellectual Disability Subscale

For the ID subscale (**Figure 1**), CFA was carried out to test the factorial structure of 11 items and three factors (IP, ST, and SA). The results did not show good fit indexes: χ 2 (41, N = 218) = 139.36, p < 0.001; χ 2 /df = 3.40, CFI = 0.93, TLI = 0.91, IFI = 0.93, RMSEA = 0.105, 90% CI [0.086, 0.125], SRMR = 0.0425. The modification indexes showed that the fit indexes improved when correlating the errors of items C (IP) and B (SA), as well as the errors of the items K (IP) and I (SA), producing the best solution (see **Figure 1**): χ 2 (39, N = 218) = 88.69, p < 0.001; χ 2 /df = 2.27; CFI = 0.97, TLI = 0.95, IFI = 0.97, RMSEA = 0.077, 90% CI [0.056, 0.098], SRMR = 0.0356. The standardized regression weights ranged between 0.61 and 0.90. The correlation between the SA and IP factors was 0.76; between the SA and ST factors, it was 0.85; between the IP and ST factors, it was 0.71.

# Factorial Analysis for the Physical Disability Subscale

For the PD subscale (**Figure 2**), CFA was carried out to test the factorial structure of 12 items and three factors (IP, SA, and S). The results of the first CFA did not yield acceptable fit indexes: χ 2 (51, N = 218) = 419.58, p < 0.001; χ 2 /df = 8.23; CFI = 0.85, TLI = 0.81, IFI = 0.85; RMSEA = 0.182, 90% CI [0.167, 0.199], SRMR = 0.0603. The correlation between the SA and IP factors was 0.80; between SA and S, it was 0.82; and between IP and S, it was 0.78. No acceptable solution was obtained in a second CFA by using the modification indexes and correlating the errors of two pairs of items (A-B and C-D): χ 2 (49, N = 218) = 255.77, p < 0.001; χ 2 /df = 5.22; CFI = 0.92, TLI = 0.89, IFI = 0.92, RMSEA = 0.139, 90% CI [0.123, 0.157], SRMR = 0.0498.

Before eliminating some of the items from the PD subscale that could be identified as problematic, two EFAs were carried out to analyze the grouping of the items belonging to the three factors. The KMO sample adequacy measure (0.90) and Bartlett's

sphericity test [χ 2 (66) = 2489.58, p < 0.001] allowed this analysis. A first EFA was performed, freeing the number of factors to be extracted. When the items loaded on two factors, the distribution had no theoretical and statistical support (e.g., items of factor S and IP were grouped in the same factor, and Items C, D, G, and I presented loading values ≥0.40 on the two factors). Therefore, a second EFA was performed, specifying three factors to be extracted. The second EFA presented results similar to the previous one (items C, D, G, H, I, K, and L showed loading values ≥0.40 on two factors), ruling out the possibility of testing the results obtained from EFA with CFA.

To achieve the best possible solution, anther CFA was performed, for which we had to eliminate items A, B, and E from the SA factor, as they were the ones that contributed the least to the construct, and it was necessary correlate the errors of items G (SA) and H (IP), and C (IP) and D (S). Finally, a nine-item, threefactor model was assessed, which obtained acceptable fit indexes (see **Figure 2**): χ 2 (22, N = 218) = 49.16, p < 0.001; χ 2 /df = 2.23;

CFI = 0.99, TLI = 0.98, IFI = 0.99, RMSEA = 0.075, 90% CI [0.047, 0.104], SRMR = 0.0284. The standardized regression weights ranged between 0.80 and 0.95. After eliminating the three items from the SA factor, the correlation between the factors was lower with regard to the original 12-item structure. Specifically, the correlation between factors SA and IP was 0.78; between SA and S, it was 0.79; and between IP and S, it was 0.76.

# Factorial Analysis for the Visual Impairment Subscale

For the VI subscale (**Figure 3**), CFA was performed to test the factorial structure of 10 items and three factors (IP, SA, and S). The results of the first CFA did not yield appropriate fit indexes: χ 2 (32, N = 218) = 205.63, p < 0.001; χ 2 /df = 6.43, CFI = 0.91; TLI = 0.87, IFI = 0.91, RMSEA = 0.158, 90% CI [0.138, 0.179], SRMR = 0.0571. The correlation between the SA and IP factors was 0.91, between SA and S, it was 0.91; and between IP and S, it was 0.80. A second CFA, which required correlating the errors of items A (S) and B (PI), and E (PI) and D (EA) to obtain the best possible model, was not acceptable: χ 2 (30, N = 218) = 107.05, p < 0.001; χ 2 /df = 3.57; CFI = 0.96, TLI = 0.94, IFI = 0.96, RMSEA = 0.109, 90% CI [0.087, 0.131], SRMR = 0.0371.

Considering the high correlation between the factors (between 0.72 and 0.93) and the fact that the fit indexes were unacceptable, two EFAs were carried out to analyze the grouping of the items. The KMO sample adequacy measure (0.91) and Bartlett's sphericity test [χ 2 (45) = 1878.83, p < 0.001] allowed this analysis. The first EFA was carried out, freeing the number of factors to be extracted, which distributed the items in a single factor that explained 67.45% of the variance. For the second EFA, specifying three factors to be extracted, some of the items (A, B, D, F, and J) showed loading values ≥0.40 on two factors, so, based on the results of the first EFA and the existence of prior evidence (see Taliaferro et al., 2015), a unifactorial CFA was performed.

To obtain the best possible fit for the CFA consisting of one factor and 10 items, it was necessary to remove items with low factor loadings (items A, B, C, and D), taking as a criterion to leave at least two representative items per factor,

and to correlate the errors of the items I (IP) and E (IP), and H (S) and G (S). However, the fit indexes were not acceptable: χ 2 (7, N = 218) = 23.11, p < 0.001; χ 2 /df = 3.30; CFI = 0.99, TLI = 0.97, IFI = 0.99, RMSEA = 0.103, 90% CI [0.058, 0.151], SRMR = 0.0148.

According to the statistical results, in which the unifactorial model required the elimination of four items corresponding to all the questions associated with performing a physical test in PE, despite which this did not guarantee a model with acceptable fit indexes, we decided to consider the best solution to be a three-factor model that ensured acceptable fit indexes and presented high correlations between some of the factors. To achieve the best possible model, the criterion was to retain three items per factor, eliminating item C for being the one that contributed the lowest factor loading to SA, and correlating the errors of items D (SA) and E (IP), and B (IP) and A (S). The fit indexes of the final model were (see **Figure 3**): χ 2 (22, N = 218) = 38.77, p < 0.001; χ 2 /df = 1.76; CFI = 0.99, TLI = 0.98, IFI = 0.99, RMSEA = 0.059, 90% CI [0.026, 0.089], SRMR = 0.0175. The standardized regression weights ranged between 0.73 and 0.93. The correlation between the SA and IP factors was 0.88; between SA and S, it was 0.89; and between IP and S, it was 0.78. Despite obtaining correlations between 0.78 and 0.89, this model was the one that obtained the best fit indexes, and was also the most coherent with the theoretical postulates of Block et al. (2013) by allowing a differentiation between three factors.

#### Confirmatory Factorial Analysis for Alternative Models of the Subscales

With the aim of analyzing the possibility of obtaining better factorial models of the three subscales, we assessed the sustainability of: (a) two alternative models of one and two factors for ID; (b) four models of one and two factors for PD; and (c) four models of one and two factors for VI. The 10 alternative models showed worse fit indexes compared to the final models presented for the subscales (see **Table 1**).

TABLE 1 | Alternative factor models of the three sub-scales.

# Descriptive Statistics, Correlation Between Items, Internal Consistency of the Instrument and the Constructs

**Table 2** shows that a score above the mid-range of the three subscales was obtained, except for items H and K of ID. Mean scores ranged from 2.27 to 3.44 (i.e., perception of mean selfefficacy). For each subscale, the correlation between its items revealed a positive and moderate correlation, ranging between 0.39 and 0.82. The internal consistency values of the instrument ranged between α = 0.81 and 0.92, whereas the construct reliability values ranged between ω = 0.92 and 0.96.

### Multigroup Analysis of Invariance

The fit indexes for all the models compared are shown in **Supplementary Table 1**. In general, the analysis of invariance by: (1) gender, (2) educational stage, (3) teaching experience, (4) training, and (5) experience in physical-sport activities adapted for the three subscales did not reveal significant differences between Model 1 and Model 2 in the chi square statistic. However, differences were found between Model 1 and Models 2, 3, and 4 for the analysis of invariance by gender of the PD subscale, showing a difference lower than 0.01 in the 1CFI between Models 1 and 2.

# DISCUSSION

Drawing on the study of Block et al. (2013), the aim of this work was to offer evidence of the factorial validity, invariance, and reliability of the SE-PETE-D-2 in the Spanish context, although now with a total of 29 items taken from the original 33 items. With the new version of the scale, which is based on of the translation of the items of Reina et al. (2016), we address the limitations identified in their work, in which these psychometric properties were not analyzed, and in the work of Abellán et al. (2019), who applied the scale to university degree students and


PI, Peers' Instruction; ST, Staying on Task; SA, Specific Adaptations; S, Safety.

TABLE 2 | Range, descriptive statistics, cronbach alpha, omega index, and Pearson's correlations of all the items of the subescales.


All the correlations were significant (p < 0.001). PI, Peers' Instruction; ST, Staying on Task; SA, Specific Adaptations; S, Safety; α, Cronbach alpha; ω, Omega Index.

students of Early Childhood and Primary Education Degrees. The results show that the Spanish EA-PEF-AD-2 is a valid and reliable instrument to measure the perceived efficacy of PE teachers toward the inclusion of students with ID, PD, and VI. However, the results also suggest that future studies should address several limitations, shared in most works of the international literature.

According to our first hypothesis, the three subscales of the EA-PEF-AD-2 maintained the original factorial structure of the three types of disability that Block et al. (2013) hypothesized in the American context, and that other authors (Jovanovic et al., 2014 ´ ; Tekidou et al., 2015; Baloun et al., 2016; Tindall et al., 2016) adapted for other countries (Greece, Ireland, the Czech Republic, and Serbia). The findings of this study are in line with other versions of the scale in which CFA was performed (Baloun et al., 2016; Kudlacek et al., 2018 ˇ ), and which revealed a multidimensional nature of the instrument with a reduction of the number of items (ID = 6, PD = 10, VI = 9) as the best solution. The results are discussed by type of disability.

First, to obtain good fit indexes for the 11-item ID subscale, it was necessary to correlate the errors of two pairs of items, sharing

Reina et al. Validation of a SE-PETE-D

with the original version of the subscale the need to correlate item K of factor IP with the errors of items of other constructs. To achieve acceptable fit indexes, Block et al. (2013) obtained a final version formed by factors IP and ST (which includes items from the constructs ST and SA), whereas for the EA-PEF-AD-2, we could maintain the more complete hypothesized version that allows discriminating between the items of ST and SA. Second, the PD subscale of nine items (of the original 12) presented good fit indexes. For this purpose, we had to perform a CFA eliminating three items from the SA factor and establishing two pairs of correlations between the errors associated with the three factors. As an advance over the original version (Block et al., 2013) and other versions of the scale (Baloun et al., 2016), we note that, in the Spanish version, the subscales of PD and VI did not present limitations in the RMSEA. Third, in order to obtain the most satisfactory version of the VI subscale, we had to remove item C (SA), associated with the performance of a physical test, and the correlation of the errors of two pairs of items. Although in the original version of the SE-PETE-D, item C did form part of the SA factor, it was the item that presented the lowest regression weight (0.56) with regard to the entire original subscale. On another hand, item F (SA), related to the learning of sports skills, eliminated in the version of Block et al. (2013), was the item with the best regression weight in the Spanish version. The deletion of item C (VI) and two of the three items (A and B) deleted from the PD subscale for the Spanish version, all associated with the performance of a physical test, could be related to the difficulties that teachers may perceive when imagining an analytical situation in which they must adapt a standardized test, in which reference standards or performance values would normally be applied, to a student with a disability. Accordingly, the proposed version of the scale mainly includes items that refer to the teaching of sportspecific skills in team sports and game dynamics, that is, items that reflect a competence-based learning model.

The EA-PEF-AD-2 presented better regression weights for the three subscales, at least with regard to the subscales of ID and PD (ID between 0.61 and 0.90; PD entre 0.80 and 0.95; VI between 0.73 and 0.93) with regard to the original SE-PETE-D (ID between 0.53 and 0.87; PD entre 0.58 and 0.91; VI between 0.73 and 0.93). On another hand, in the CFA of the EA-PEF-AD-2, high correlations were obtained between some factors of the subscales, highlighting a correlation of 0.89 between the SA and S constructs for the VI subscale. However, alternative CFAs did not improve the resulting fit indexes for the SE-PETE-D models, which followed to the postulates of Block et al. (2013). In addition, the correlations between the new fused factors did not show a substantial improvement over those presented by the factorial structures that distinguished between all the constructs that made up the three subscales, which would support the discriminant validity of this Spanish version. On another hand, as the statistical results indicate, maintaining the association of each item with its initial factor would allow respecting the essence of the theoretical constructs. A possible explanation of the high correlations between some constructs, such as SA and ST, could be related to their theoretical affinity and the difficulty in differentiating between the design of an adaptation for an activity in PE (SA) and how to put it into practice (ST). That is, some teachers may not differentiate their perception of competence between design and implementation because, probably, teachers who perceive themselves as effective in the design of adaptations will apply them appropriately during their PE classes.

In line with our second hypothesis, the multigroup analyses supported the invariance of the factorial structures of the subscales of ID, PD, and VI by gender, educational stage, teaching experience, previous training, and contact experiences in adapted or inclusive physical-sporting activities. These results make a new contribution to the previous versions of the scale, making it possible in the future to compare the perception of self-efficacy of PE teachers in the five demographic variables, which the current literature has not yet addressed in depth.

Finally, in relation to the third hypothesis, we obtained a moderate and positive correlation between all the items of the constructs of each subscale and excellent reliability values (α ≥ 0.81 and ω ≥ 0.92), which support the affinity of the items belonging to the same construct and which contribute to improve the reliability of the subscales. In addition, the reliability values reported by SE-PETE-D in other countries are in the same line: United States (Taliaferro et al., 2015; α ≥ 0.96), Greece (Tekidou et al., 2015; α ≥ 0.90), and the Czech Republic (Baloun et al., 2016; α ≥ 0.76; Kudlacek et al., 2018 ˇ ; α ≥ 0.82), although these statistics have not been reported for applications in Ireland (Tindall et al., 2016) and Serbia (Jovanovic et al., 2014 ´ ). Therefore, eliminating items A, B, and E for the PD subscale and item C for the VI subscale, all of which are related to the SA factor, does not affect the reliability of this Spanish version. Indeed, it presents similar scores than the original scale (Block et al., 2013; α scores from 0.73 to 0.89).

# CONCLUSION

This study supports the EA-PEF-AD-2 as a valid and reliable instrument to measure the self-efficacy of PE teachers when including students with ID, PD, and VI in the classroom. This Spanish version is presented as the first multidimensional version that allows maintaining for each subscale the items associated with the constructs originally hypothesized by Block et al. (2013). This is the first version validated with in-service teachers, which opens new possibilities of research, given the importance of this construct in the professional development of teachers. In view of the lack of consistency in the international versions on the specific situations that the scales evaluate, it seems appropriate to reflect on the appropriateness of the specific situations to which the items refer. Thus, while a good consistency for the situations of teaching of sports skills and the sport game itself was obtained, this cannot be said for the situations of physical condition or their evaluation. This suggests not only intercultural differences, but also the different educational curricula available in Spain, which vary according to primary and secondary education levels, and include other blocks of content such as activities in the natural environment, body expression, or physical activity and health. Given this educational diversity, it would be necessary to optimize the instrument in relation to the dimensions of self-efficacy for each disability subscale. Although

the IP (with the consequent importance that they could have for the inclusion and cooperation of students with disabilities) and SA factors (essential element to compensate different capacities of students with disabilities) are present in the three subscales (ID, PD, and VI), this was not the case for factors S and ST. The lack of consensus in this regard in the works of the literature suggest that the structure of the SE-PETE-D is improvable, and it could include typical dimensions of inclusive models in PE such as the STEP (Space, Task, Equipment and People; Black and Williamson, 2011) or the TREE (Teaching style, Rules, Equipment, and Environment modifications; Tripp et al., 2007), among others.

Another line of work is the inclusion of new disability groups for evaluating teachers' perception of self-efficacy. At the international level, there have already been approaches to autism (Taliaferro et al., 2015) and cerebral palsy with different levels of mobility or autonomy (Hutzler and Barak, 2017) but they have not yet been applied in non-English speaking countries. This proposal could be extended to groups of interest such as students with attention deficit and hyperactivity, mental health problems, or hearing impairments. Finally, considering the prolific research on the attitudes of teachers toward the inclusion of students with disabilities in their classrooms (see the review of Wilhelmsen and Sørensen, 2017), it would be of interest to study the relationship between such attitudes and teachers' degree of competence and perceived efficacy. Future studies could address the relationship of both constructs, together with others of greater emotional depth such as their values or moral commitment to teaching, examining in depth the psychological mechanisms underlying the teaching process concerning inclusion.

#### DATA AVAILABILITY STATEMENT

All datasets generated for this study are included in the manuscript/**Supplementary Files**.

#### REFERENCES


# ETHICS STATEMENT

The studies involving human participants were reviewed and approved by Órgano Evaluador de Proyectos (University Miguel Hernández). The patients/participants provided their written informed consent to participate in this study.

#### AUTHOR CONTRIBUTIONS

RR and AR performed the data collection and developed the theoretical framework. RF performed the analyses and drafted the results, figures, and tables. RR, RF, and AR interpreted the results. All authors listed have made a substantial intellectual contribution to the research and contributed to drafting the manuscript.

# FUNDING

This study was funded by the National Paralympic Committee (Spain) and the Trinidad Alfonso Foundation (Refs. CPE1.16D and CPE1.17F).

# ACKNOWLEDGMENTS

Thanks for the collaboration of the participating teachers and the Ministry of Education, Culture and Sports of the Generalitat Valenciana (Spain) for facilitating access to them.

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyg. 2019.02169/full#supplementary-material



**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Reina, Ferriz and Roldan. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Work Burnout and Engagement Profiles Among Teachers

Katariina Salmela-Aro1,2 \*, Lauri Hietajärvi<sup>1</sup> and Kirsti Lonka1,3

<sup>1</sup> Faculty of Educational Sciences, University of Helsinki, Helsinki, Finland, <sup>2</sup> ETH Zurich, Zurich, Switzerland, <sup>3</sup> Optentia Research Focus Area, North-West University, Vanderbijlpark, South Africa

The focus of the current study was to examine teachers' well-being in terms of work engagement and burnout by using a person-oriented approach. The participants (n = 149, 70.5% female) were subject-matter teachers from 22 schools from metropolitan Helsinki area in Finland. The first aim was to examine the kinds of profiles we can identify based on work burnout and engagement among teachers. The second aim was to study how the identified profiles differed in job-related demands and resources and personal resources in terms of resilience. Based on the demandsresources model, we expected to find profiles that differ in terms of key resources and demands. The sample was acquired as a convenience sample and the data was collected using online self-report questionnaires. The measures were work engagement, work burnout, work demands/resources (workload and control) and resilience as the personal resource. In addition, changes and effects of the economic circumstances were accounted for with two binary variables assessing the effect on class sizes and material resources. We identified two profiles among teachers: engaged (30%) and engaged-burnout (70%) profiles. We found that those in the engaged profile group had more job and personal resources, such as control and resilience, whereas those in the engaged-burnout profile group experienced more work demands, such as workload.

#### Edited by:

Caterina Fiorilli, Libera Università Maria SS. Assunta, Italy

#### Reviewed by:

Lorenzo Avanzi, University of Trento, Italy Sebastiano Costa, University of Campania Luigi Vanvitelli, Italy

#### \*Correspondence:

Katariina Salmela-Aro katariina.salmela-aro@helsinki.fi

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 12 June 2019 Accepted: 20 September 2019 Published: 04 October 2019

#### Citation:

Salmela-Aro K, Hietajärvi L and Lonka K (2019) Work Burnout and Engagement Profiles Among Teachers. Front. Psychol. 10:2254. doi: 10.3389/fpsyg.2019.02254 Keywords: teachers, work engagement, burnout, person-oriented approach, resilience

# INTRODUCTION

Teacher burnout has been identified as a worldwide problem (Tsouloupas et al., 2010; Dicke et al., 2015, 2017). Studies have focused on teachers' burnout, stress and high attrition, and, thus a lack of teachers. Finland is a rare country in the sense that teacher profession is highly appreciated and wanted profession so far; it has been difficult to get in to teacher education programs that are at the level of Master degree, the requirement of working as a teacher (Lonka, 2018). Still, teachers have been found two decades ago to suffer from burnout levels when compared to workers in other white collar jobs (Kalimo and Hakanen, 2000). This is very alarming as teachers wellbeing is related to students' wellbeing (Koh et al., 1995; Day, 2011; Klusmann et al., 2016). Thus there is a need to identify the amount of teachers suffering from burnout. Moreover, we have to identify the risk as well as protective factors for developing burnout symptoms among teachers.

However, we have to be aware that the majority of teachers are engaged about their work (Kinnunen et al., 1994). In Finland, teachers have high autonomy and control over their work and there are no inspections or yearly standardized tests to monitor teachers (Sahlberg, 2010). On the other hand, a new national curriculum is introduced every 10 years and the profession is

**117**

constantly under reforms and profound changes. Even though the teachers' high competence, their working conditions are becoming more challenging as increasingly heterogeneous pupil population due to immigration and reforms toward inclusive special needs education, endeavors to digitalize schools, as well as cuts and savings in school funding. The profession is thus simultaneously likely to promote both engagement and burnout because of constant new challenges and requirements. Thus, our aim was to take a person-oriented approach and examine simultaneous work engagement and burnout profiles among teachers and their antecedents and consequences in the context of demands-resources model. Under such working conditions is possible to find teachers who are highly engaged, but still may report exhaustion or other symptoms of burnout.

In person-oriented approach compared to variable-oriented approach it is possible to find different profiles of participants and there is still very few studies with regard to occupational wellbeing for educational personnel. A key benefit of a personoriented approach is the possibility to identify the participants endorsing seemingly contradictory measures of work well-being, such as burnout and engagement, simultaneously (Moeller et al., 2018). Latent profile analysis brings the capacity to guide identifying the ideal number of profiles in burnout and engagement (see also Leiter and Maslach, 2016). However, there is, in particular, lack of studies to examine teachers' simultaneous burnout and engagement profiles. Some studies have identified teacher profiles (Klusmann et al., 2008; Collie et al., 2015) but none of the previous studies have examined work burnout and engagement profiles among teachers. Previous studies in educational context, among elementary (Salmela-Aro et al., 2017), high school (Tuominen-Soini and Salmela-Aro, 2014) and higher education (Salmela-Aro and Read, 2016) students rather than teachers have identified three to five profiles, engaged, engaged-burnout and burnout profiles. Moreover, previous studies among occupational context in general have identified mostly two profiles, engaged and burnout (Innanen et al., 2015). Based on the previous studies we expected to identify two to three profiles among teachers, engaged, engagedburnout and possibly even burnout profiles. Based on the professional profile, it was expected that exhaustion would be more common than cynicism, because teachers have good chances for experiencing autonomy, competence and relatedness that support autonomous and intrinsic motivation as well as personal engagement (Deci and Ryan, 2008; Skinner et al., 2014; Rothmann and Cooper, 2015).

The Job Demands-Resources (JD-R) model was used to examine job burnout and engagement profiles (Demerouti et al., 2001; Bakker et al., 2003a; Bakker and Demerouti, 2006). Job demands-resources model proposes a motivational process from job resources to work engagement and an energetic process from demands to job burnout. In the present study, we integrate the energetic and motivational processes by examining simultaneous burnout and engagement profiles among teachers (Hakanen et al., 2006; Upadyaya et al., 2016). Work burnout often reflects employees' reaction to chronic occupational stress, and work burnout has been conceptualized in terms of inadequacy, exhaustion, and cynicism at work (Maslach et al., 2001; Schaufeli et al., 2002). Exhaustion refers to strain and overtaxing work (González-Romá et al., 2006). Cynicism refers to negative work attitude, disinterest and low work meaning (Salmela-Aro et al., 2009). Finally, inadequacy refers to inadequacy at work and decreased accomplishments (Schaufeli et al., 2002). These refer to emotional, cognitive, and behavioral components of burnout (Salmela-Aro and Upadyaya, 2014). In turn, work engagement can be defined as high dedication, energy, and absorption at work (Schaufeli et al., 2002). Resilience and vigor characterizes the energy component (Schaufeli et al., 2002). Meaning, pride, and inspiration refer to work dedication (Schaufeli et al., 2002). Finally, absorption is similar to flow so that time passes quickly at work (Schaufeli et al., 2002). These refer to emotional, cognitive and behavioral components of work engagement (Salmela-Aro and Upadyaya, 2014).

According to the JD-R model (Demerouti et al., 2001; Bakker et al., 2003a) high level of psychological, social, physical, and organizational demands are related to work burnout, whereas psychological, social, physical, and organizational resources are related to work engagement (Xanthopoulou et al., 2009; Alarcon, 2011; Salmela-Aro and Upadyaya, 2014). Key identified work demands are workload, time pressure, and long working hours (Bakker et al., 2003b) but also interpersonal conflicts (Ilies et al., 2015). These have been found to increase teachers' burnout (Pietarinen et al., 2013). Key job-related resources refer to positive team climate, and control. In turn, personal resources refer to employees' self-efficacy, self-esteem, and resilience, and they are associated to work engagement (Bakker and Demerouti, 2008; Salmela-Aro and Upadyaya, 2014). Consequently, work and personal resources can reduce work demands as they help to achieve work-related goals (Demerouti and Bakker, 2011). It has also been found that personal resources may act as a buffer so that the possible negative association from work demands to work burnout is weaker (Hakanen et al., 2006; Bakker et al., 2007; Upadyaya et al., 2016; Spini et al., 2017). Recent studies have revealed that resilience is an important personal resource, as it can act as a protective factor that allows teachers to react using adaptive coping when they face challenges and problems in their work (Skinner et al., 2014). In the present study we focused on the key job demands which might be particularly important among teachers, such as workload and large class size, and job resources which might be particularly important as major motivations in teachers job that increase work engagement, or when lacking increase work burnout, such as control and material resources. Finally, personal resources can buffer and lead to work engagement. The present study included resilience as a personal resource.

The aim of the present study was to identify profiles among teachers based on work burnout and engagement. Based on the previous studies we expected to identify two to three profiles among teachers, engaged, engaged-burnout, and burnout profiles (H1). The second aim was to analyze if the identified profiles differ in job-and personal resources and job demands. Based on the demands-resources model, we expected to profiles differ in terms of resources and demands (H2). We expected those in more engaged profile group to have more job and personal resources, such as control and resilience, whereas those in the burnout

profile group to have more work demands, such as workload than those in the more engaged profile group.

# MATERIALS AND METHODS

#### Participants

The participants were 149 (70.5% female) subject matter teachers from 22 comprehensive and high schools from Metropolitan Helsinki area. The sample was acquired as a convenience sample and the data was collected using online self-report questionnaires. Out of the participants 76% were from high schools, 15.1% from schools that provide both comprehensive and high school education and 8.2% were from comprehensive schools. The average age of the participants were 45.62 years (SD = 8.56) and the average experience as a teacher was 16.93 years (SD = 8.76). Participation in the study was voluntary and informed consent were acquired as a part of the questionnaire. The study protocol was approved by the University of Helsinki Ethical Review Board in the Humanities and Social and Behavioral Sciences.

# Measures

#### Work Engagement

Work Engagement was measured with a short version of the Utrecht Work Engagement Scale, UWES-S (Schaufeli et al., 2006) which consists of nine items measuring work-related energy ("When I work, I feel that I am bursting with energy"), dedication ("I am enthusiastic about my work"), and absorption ("Time flies when I'm working") to be rated on a 7-point scale (0 = not at all to 6 = daily). These items refer to emotional, cognitive and behavioral aspects of work engagement. We used the overall measure of work engagement (Schaufeli et al., 2006). A sum score was formed to measure the employees' overall work engagement. The Cronbach's alpha reliability for the scale was 0.90.

#### Work Burnout

Work Burnout was measured with the Bergen Burnout Inventory (Salmela-Aro et al., 2004; see also Näätänen et al., 2003; Salmela-Aro et al., 2011) which consists of 9 items measuring three factors of job burnout: exhaustion ("I feel overwhelmed by my work", α = 0.73); cynicism about the work meaning of work ("I feel lack of motivation in my work and often think of giving up", α = 0.79), and inadequacy ("I often have feelings of inadequacy in my work", α = 0.80) rated on a 6-point scale (1 = strongly disagree to 6 = strongly agree). These refer to emotional, cognitive and behavioral aspects of work burnout. Sum scores were formed to represent the different components of burnout.

#### Work Demands and Resources

Work demands was measured by workload and work resources as control. They were measured with items taken from the Areas of Worklife Survey (Leiter and Maslach, 2016). Workload consisted of 2 items (e.g., "I don't have time for all the work that needs to be done." α = 0.68). Control consisted of 2 items (e.g., "I have control over how I do my work". α = 0.54). The items were rated on a 5-point likert-type scale (1 = completely disagree; 5 = completely agree).

#### Personal Resources

Personal resources were measured by resilience. Resilience was measured with 8 items (e.g., "When I encounter difficulties in my work, if usually find multiple solutions." α = 0.90) to be rated on a 5-point likert scale (1 = completely disagree to 5 = completely agree) (Smith et al., 2008).

In addition, changes and effects of the economic circumstances were accounted for with two binary variables assessing the effect on class sizes ("Have the class sizes increased due to economic circumstances") and the effect on material resources ("Have the material resources diminished due to economic circumstances and affected general academic performance").

### Analysis Strategy

First, data was screened for possible outliers and missing values in SPSS 24<sup>1</sup> . There were 0.62% of values missing. Little's MCAR test showed that these were missing completely at random (Chi-Square = 139.107, DF = 156, Sig. = 0.830). No outliers were identified. Further analysis were conducted with R version 3.5.3 and RStudio (R Core Team, 2015, 2018). The sum scores were computed with the restriction that 50% of the items in each scale were required to have a valid value in order to be computed, otherwise the sum variable were coded missing (package "sjstats", Lüdecke, 2018). As a measure of internal consistencies there were estimated Cronbach's Alphas (package "MBESS", Kelley, 2018). See **Table 1** for descriptive values of the variables.

Latent profile analyses (see e.g., Vermunt and Magidson, 2002) were conducted to identify latent subgroups of teachers regarding their work engagement and burnout. Analyses were conducted with "tidyLPA" package in R including a simple single imputation and standardization to the variables prior to model estimation as implemented in the package (Rosenberg et al., 2019). To estimate the best fitting model we relied most on Bayesian information criterion (BIC), which has shown to be robust across a variety of conditions (Nylund et al., 2007). Subsequent binomial logistic regressions were conducted to examine the relations of covariates and the profile membership, the statistical significance of the effects were evaluated with the conventional alpha level of p < 0.05 with, however, confidence intervals also presented.

# RESULTS

First, we compared solutions with different restrictions for variances and covariances between different profiles (for details see **Table 2**). This was done to identify the model that would fit our data best. With each specification we estimated models with increasing number of profiles. All models except Model 1 (see **Table 2**) suggested that a two-profile solution would be best and out of these the lowest BIC was estimated with the Model 4 specification in which means, variances and covariances were allowed to be freely estimated across profiles. As the two profile solution was also substantively meaningful we decided to proceed with the two profile solution.

<sup>1</sup>https://www.ibm.com/analytics/data-science/predictive-analytics/spssstatistical-software

#### TABLE 1 | Variable descriptives.

fpsyg-10-02254 October 3, 2019 Time: 18:0 # 4


<sup>∗</sup>Bias-corrected and accelerated bootstrap with 10,000 draws.

TABLE 2 | Model fit indices of latent profile analyses for all models compared.


The model specifications refer to holding variances/covariances equal, fixed to 0 or freely estimated between classes. The selected model specification bolded and the selected model italicized.

The two profiles we identified represented two distinct types of teachers (see **Figure 1** and **Table 3**): Profile 1 was the larger profile (70%) and consisted of teachers that were quite engaged,

although slightly lower than the sample mean (z = −0.35, a small effect). They, however, also experienced more symptoms of burnout than the profile 2. In all burnout symptoms they were higher than the sample mean (z = 0.30 to 0.39, small effects). The profile was labeled as engaged-burnout. Profile 2 (30%) consisted of teachers that were more engaged (z = 0.80, a large effect) and experienced less symptoms of burnout (z = −0.70 to −0.91, large effects) than the profile 1. The profile was labeled as highly engaged. As the entropy of the selected solution was

TABLE 3 | Mean differences between profiles in indicator variables.


z, standardized difference from sample mean.

#### TABLE 4 | Results of the binary logistic regression.

fpsyg-10-02254 October 3, 2019 Time: 18:0 # 5


Estimates show the probability for belonging in profile 2 "Highly Engaged" as compared to profile 1. <sup>∗</sup>Bias-corrected and accelerated bootstrap with 1000 draws.

high (0.92) we saved the most likely profile memberships for subsequent analysis.

We then specified a binomial logistic general linear model in which we assigned the profile membership (profile 2 as compared to profile 1) as dependent variable and gender, workload, control, resilience and the two financial context measures as independent variables (see **Table 4**). Teachers with a higher workload and who reported increase in class sizes due to economic circumstances were more likely to belong to Profile 1 (engaged-burnout), whereas teachers who experienced more control over their work and reported higher resilience were more likely to be assigned to Profile 2 (highly engaged).

# DISCUSSION

The results of the previous study among teachers showed that two profiles could be identified based on teachers' burnout and engagement. These two profiles represented two distinct types of teachers and supported our hypothesis. The first profile was larger (70%) and consisted of teachers that were quite engaged but also experienced more symptoms of burnout than the other group. This group was named as engaged-burnout group. This group is in line with the earlier studies among students in high school which has identified a group of students who are engaged but have at the same time some symptoms of burnout (Salmela-Aro and Read, 2017). This is an important finding as this profile has not been previously identified among teachers. However, the signs of burnout were only emerging and more studies are needed to replicate the findings. In turn, the second profile was smaller (30%) and consisted of teachers that were very engaged and experienced low level of symptoms of burnout. This group was named as highly engaged group. These results show that over two times more teachers belong to the engaged-burnout group compared to the highly engaged group.

Interestingly, the person-oriented approach used in the present study showed that both groups were high in engagement. Teachers seem to be very committed and engaged to their work, whereas many suffer at the same time some signs of exhaustion and inadequacy as a teacher. This is a new but also a worrying result. Earlier studies among students have identified this kind of engaged-exhausted profile to be quite successful in the short–term but it has been found to lead to severe costs in the long run. Earlier studies have shown this group in particular to be risk of depressive symptoms later on (Tuominen-Soini and Salmela-Aro, 2014). Depression is one of the leading risks for early retirement and dropping out of teacher profession and thus these results need to be taken very seriously. This result may indicate that even highly educated, autonomous teachers who experience their work meaningful and who are engaged, may suffer of some symptoms of burnout. It is worth pointing out that, fortunately, cynicism was not very high even in this profile. Teachers value their work and see their work as meaningful.

Supporting the demands-resources model and our hypothesis 2, the results showed further that engaged-burnout and engaged group differed in terms of the key work-related demands and work- and personal resources (Bakker et al., 2003a; Hakanen et al., 2006). Teachers with a higher workload and who reported increase in class sizes due to economic circumstances were more likely to belong to engaged-burnout group, whereas teachers who experienced more control over their work and reported higher resilience were more likely to be assigned to engaged group. Even when facing high challenges and heavy workload, sense of meaning and resilience may help teachers to cope with their work without becoming cynical or feel inadequate. On the other hand, the results should put into a larger context: high workload and increase in class size are school-level problems that may really make teachers to lose control (or at least sense of control) over their own work. The fact that the majority of teachers in this study belonged to the engaged-burnout cluster indicates that the working conditions calls for serious attention by the policy makers and municipals.

#### LIMITATIONS AND FUTURE RESEARCH

The present study has some limitations. The study was based on teachers self-reports, and the sample was quite small. The reliability of some variables was 0.60 and one even below (0.54). By conventional psychometric criteria, any values of coefficient alpha below 0.6 would be regarded as poor, even for relatively heterogeneous constructs that are not regarded as high-stake psychological tests or (Richardson, 2004). Only one Cronbach's Alpha in our study was below this limit. There is always a tradeoff between a wish to increase reliability and simultaneously, to restrict the length of an instrument, because reliability coefficients always become better as the test is prolonged (DeVellis, 1991). In our pilot studies it appeared that teachers were very similar to medical students in terms of not tolerating long series of questions (Lonka et al., 2008; Vedenpää and Lonka, 2014).

We therefore ended up with some scales with only two to three items. To maximize the number of participants, we created an instrument that was reasonably short and still satisfactory in terms of reliability.

In addition, the study was cross-sectional as a design. Future studies needs to be carried out with larger samples and with longitudinal design. Longitudinal design could reveal the possible

long-term dark side of the engaged burnout group of teachers. During recent year, Finnish teachers working conditions have become ever so challenging: a new ambitious and innovative national curriculum was introduced in 2016 with increasingly inclusive school policy and fast digitalization. At the same time the Government cut funding from schools. Thus it is important to monitor the development of well-being with larger numbers of teachers.

Moreover, in the future studies we need to examine at the same time both teachers and students and to reveal possible spill over and buffering effects. In addition, besides teachers, there is a need to study also the principals as they play a key role for empowering and motivating teachers and the school as an organization. The role of principals' servant leadership as a possible buffer for teacher burnout is an interesting future research question (Upadyaya et al., 2016). Intervention studies are strongly needed to identify burnout risk teachers and prevent them from burnout out and leaving the profession.

#### CONCLUSION

The person-oriented study identified two profiles of teachers, engaged-burnout (70%) and highly engaged (30%) ones. The study revealed first time a profile of teachers being engaged but at the same time in a risk of exhaustion and inadequacy. Supporting demands-resources model both personal and work related resources we more typical for the highly engaged group, whereas the work-related demands were more typical among the engaged-burnout group. The results are important from both the perspective of fostering resilience among teachers, but also in looking at the policy issues in the larger context of schools and the educational system. We would also recommend new forms of teachers' crafting options to develop their work in order to help them to keep up with the increasing demands of teacher work in the rapidly changing information society and increasingly heterogeneous student population (Upadyaya et al., 2016). Identifying risk factors for teacher burnout is important. Even those teachers who report to be engaged, but simultaneously also exhausted, may have a risk to develop burnout. It is crucial to invest in teachers' well-being and working conditions. Even excellent teachers have their limits in terms of how much

#### REFERENCES


change they can tolerate at the same time. Cutting funds and introducing ambitious reforms at the same time may not be a good idea. Theoretically, approaching teacher engagement and burnout simultaneously, from the point of view of resources and demands, is important in quickly changing information society. It is important to see the risks on time, before the teachers would really burn out.

### DATA AVAILABILITY STATEMENT

The datasets generated for this study are available on request to the corresponding author.

#### ETHICS STATEMENT

The studies involving human participants were reviewed and approved by the University of Helsinki Ethical Review Board in the Humanities and Social and Behavioral Sciences. The patients/participants provided their written informed consent to participate in this study.

# AUTHOR CONTRIBUTIONS

KS-A planned the design and analyses, and participated in the writing process. LH was responsible for the data analyses and reporting of the statistical results. KL organized the collection of the data as the Principal Investigator of the project and contributed in the writing of the manuscript.

### FUNDING

This research was funded by the Academy of Finland project "Bridging the Gaps—Affective, cognitive, and social consequences of digital revolution for youth development and education," PI KS-A (grant #308351) and PI KL (grant #308352), and the Academy of Finland project #Agents – Young People's Agency in Social Media, PI KS-A (grant #320371) as well as a project "Phenomenal teacher education" funded by Finnish Ministry of Culture and Education (grant #6605844).




**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Salmela-Aro, Hietajärvi and Lonka. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fpsyg-10-02330 October 11, 2019 Time: 16:16 # 1

# Trans-Cultural Validation of the "Academic Flow Scale" (Flow 4D 16) in Arabic Language: Insights for Occupational and Educational Psychology From an Exploratory Study

#### Edited by:

Paula Benevene, Libera Università Maria SS. Assunta, Italy

#### Reviewed by:

Michela Cortini, Università degli Studi G. d'Annunzio Chieti e Pescara, Italy Annamaria Di Sipio, University of Padova, Italy Alessandro De Carlo, Giustino Fortunato University, Italy

#### \*Correspondence:

Nicola Luigi Bragazzi robertobragazzi@gmail.com †These authors have contributed equally to this work as co-last authors

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 07 July 2019 Accepted: 30 September 2019 Published: 15 October 2019

#### Citation:

Chalghaf N, Azaiez C, Krakdiya H, Guelmami N, Re TS, Maldonado Briegas JJ, Zerbetto R, Del Puente G, Garbarino S, Bragazzi NL and Azaiez F (2019) Trans-Cultural Validation of the "Academic Flow Scale" (Flow 4D 16) in Arabic Language: Insights for Occupational and Educational Psychology From an Exploratory Study. Front. Psychol. 10:2330. doi: 10.3389/fpsyg.2019.02330 Nasr Chalghaf1,2,3, Chiraz Azaiez2,3, Hela Krakdiya2,3, Noomen Guelmami1,4,5 , Tania Simona Re6,7,8, Juan José Maldonado Briegas<sup>7</sup> , Riccardo Zerbetto<sup>8</sup> , Giovanni Del Puente<sup>9</sup> , Sergio Garbarino<sup>9</sup> , Nicola Luigi Bragazzi1,6,7,8,9,10 \* † and Fairouz Azaiez1,2,3†

<sup>1</sup> Department of Health Sciences (DISSAL), Postgraduate School of Public Health, University of Genoa, Genoa, Italy, <sup>2</sup> Group for the Study of Development and Social Environment (GEDES), Faculty of Human and Social Science of Tunis, Tunis, Tunisia, <sup>3</sup> Higher Institute of Sport and Physical Education of Sfax, University of Sfax, Sfax, Tunisia, <sup>4</sup> Higher Institute of Sport and Physical Education of Kef, University of Jendouba, Jendouba, Tunisia, <sup>5</sup> Research Unit, Sportive Performance and Physical Rehabilitation, High Institute of Sports and Physical Education of Kef, University of Jendouba, Jendouba, Tunisia, <sup>6</sup> UNESCO Chair "Health Anthropology, Biosphere and Healing Systems", University of Genoa, Genoa, Italy, <sup>7</sup> Department of Psychology and Sociology of Education, University of Extremadura, Badajoz, Spain, <sup>8</sup> Centro Studi Terapia della Gestalt, Milan, Italy, <sup>9</sup> Department of Neuroscience, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health (DINOGMI), University of Genoa, Genoa, Italy, <sup>10</sup> Laboratory for Industrial and Applied Mathematics, Department of Mathematics and Statistics, York University, Toronto, ON, Canada

Background: As an optimal psychological state, flow represents those moments when everything comes together for the performer. Flow is often associated with high levels of performance and is a positive psychological experience.

Aim: Our study aimed to validate the "Academic Flow Scale" (Flow 4D 16) in Arabic language across Tunisian population, and to test its factor structure, in terms of internal consistency/reliability, predictive validity, and sensitivity.

Methods: The population is composed of 320 students (139 men and 181 women) belonging to the University of Sfax, with a mean age of 22.26 years. The students voluntarily responded to the scale of academic flow (Flow 4D 16). Both exploratory (EFA) and confirmatory (CFA) factor analyses were performed.

Results: The four-dimensional alpha coefficients of the Flow 4D 16 indicate an excellent internal consistency, respectively, of 0.902 (Cognitive), 0.959 (Time), 0.974 (Ego) and 0.960 (Well-being). The CFA fit indices were satisfactory.

Conclusion: In summary, the 16-items model (original version) showed for all the indices an excellent fit to the theoretical model, confirming the four-dimensional factor structure among Tunisian student population.

Keywords: trans-cultural validation of a scale, Arabic language, academic flow, occupational psychology, students

# INTRODUCTION

fpsyg-10-02330 October 11, 2019 Time: 16:16 # 2

The concept of "flow" or "psychological flow" can be defined as "the optimal experience" that a subject can make. It generally occurs when an individual is engaged in a specific activity with clear goals and high commitment, facing challenges in proportion to his/her skills, fully mobilizing his/her competencies, and dedicating his/her attention to the task. Indeed, flow occurs when the subject is totally absorbed by what is doing and becomes unconscious of himself/herself. In this case, he/she forgets the time that passes, as well as other potential sources of distraction, even including his/her bodily needs.

The Hungarian psychologist Mihály Csíkszentmihályi has developed the concept of "flow", being considered the founding father and the pioneer of the flow studies. His works published between 1975 and 2000 aimed to study the context of appearance and emergence of the optimal experience peak and the structure of pleasure generated. His investigations are generally based on the descriptions of the experiences of individuals who feel this pleasure in the practice of their activity where the intrinsic reward is considered essential (Nakamura and Csikszentmihalyi, 2002).

There are a lot of flow studies, among which we can take into particular consideration the classification of some scholars (Walker, 2010; Borderie and Michinov, 2014; Borderie, 2015) who distinguish between two types of flow.

The first approach is represented by the model of individual flow: the group to which the individual belongs is considered only as the context of the emergence of the state of flow. The collective entity to which the subject belongs can have just a minimal effect on his/her state.

The second model is characterized by the concept of group flow that corresponds to the collective optimal experience. This takes place when individuals act in co-presence (Johnson and Johnson, 1989; Walker, 2010; Borderie and Michinov, 2014; Borderie, 2015).

Other researchers (Fong et al., 2014) have re-examined, in a comparative analysis, what Csikszentmihalyi (2004) has called "the paradox of work": that is to say, the fact that people experience the flow more often at workplace than during their leisure time. The results found by their meta-analysis do not differ significantly from those obtained for a very short period (only 7 days) among 100 workers working in different occupational contexts and jobs. This could be explained by taking into account the value given to leisure, which is higher so that people tend to prefer leisure, which also correspond to periods of less stress and activation (Engeser and Baumann, 2016).

Many scholarly researches as well as practical applications have taken place after the introduction of the seminal concept of optimal experience. Flow appears particularly in the school world, impacting on different variables and constructs, from well-being to performance, academic achievements, career expectations and future occupational employment (Cortini et al., 2010). Indeed, Csikszentmihalyi and Rathunde (1993), in a longitudinal survey of more than 200 adolescents, showed that the student's optimal feeling at school was the best predictor of his/her subjective engagement and how he/she fully used his/her potential. Larson (2011) has shown how flow can be exploited in order to improve school learning at a higher level, emphasizing the main conditions of the optimal experience including the autonomy, interest and commitment of the learners. Much more, the study of Carpentier et al. (2012) showed that subjects who have an intrinsic and self-controlled motivation, under the influence of a "harmonious passion", can have a greater tendency to live the experience of flow in their favorite activities.

In this context, the relation between flow and positive affect often appears in undergraduate students (Rogatko, 2009). Furthermore, Kulkarni et al. (2015) conducted three experiments with undergraduate university students using a simple and interesting on-screen game. They manipulated the flow level by including more or less flow ingredients. They showed that the presence of the main antecedents (commitment, feedback and balance-skill) produced a higher level of flow (pleasure, interest, concentration) and decreased the defensive behaviors.

Given the importance of the concept of flow in educational contexts and milieus, here briefly overviewed, and given that no instrument exists in Arabic language, the aim of the present study was to provide a trans-cultural validation of the "Academic Flow Scale" (Flow 4D 16).

# MATERIALS AND METHODS

### Participants

The population under study was made up of 320 students (139 men and 181 women) (**Table 1**) belonging to different study sections (beauty arts, law, humanities, economics, sports and physical education, and medicine) of the University of Sfax, Sfax, Tunisia, with a mean age of 22.26 ± 1.86 years. All participants in our study were volunteers, taking part into the study in an anonymous and confidential way. They gave their written, informed consent to be part of the investigation and were carefully and extensively advised about the aim of the study. They were free to withdraw their participation at any moment of the investigation. The study protocol was in-depth reviewed and received the full approval by the ethical committee of the University of Sfax, Sfax, Tunisia.

#### Measurement

Several questionnaires exist to measure the flow. On the one hand, the Mayers, 1978 Flow Scale a self-descriptive measuring instrument, enables scholars to estimate how often a person experiences each of the nine dimensions of flow. This tool can be used for repeated measurements (Delle Fave and Massimini, 1988) in order to evaluate the differences in the state of flow according to the activity practiced at various moments. On the other hand, more recently, a flow scale has been developed, based on questionnaires developed in specific contexts such as sports and physical education (Jackson and Marsh, 1996; Jackson and Eklund, 2002) and psychology/psychotherapy (Demontrond and Gaudreau, 2008). Within this field, there exists a French version of the Flow State Scale (FSS) composed of 32 items divided into four dimensions (Jackson and Eklund, 2002; Fournier et al., 2007).

fpsyg-10-02330 October 11, 2019 Time: 16:16 # 3


TABLE 1 | Distribution of the study population by sex and study sections.

Furthermore, there exists a French version of the Flow scale adapted to specifically study the engagement and involvement of students in collective work, and especially how they live the optimal experience (flow) within an educational context. This measuring instrument of the optimal experience (Flow 4D 16) in an educational context was developed by Heutte and Fenouillet (2010). This tool contains four dimensions (each one is consisting of four items) on a 7-degree Likert scale ranging from 1 ("strongly disagree") to 4 ("indifferent") and 7 ("strongly agree").

In our study, we have chosen this scale to assess the academic flow in a sample of Tunisian students.

#### Procedure

Our procedures for translating and adapting the flow scale (4D 16) validated in French by Heutte and Fenouillet (2010) were based on two subsequent translation steps (namely, forward translation and back translation) of this measurement tool, from the original language to the target language (Arabic). This allows the detection of divergent errors and interpretations of certain ambiguous items of the original version during initial translations. Then, we made a back-translation to the French language from versions translated into Arabic.

The translations were done by two teams of translators. Each team was made up of two translators. Their characteristics and personal qualifications are important in terms of knowledge of both the source language and of the target language.

#### Statistical Analysis

Data normality was checked assessing the skewness and kurtosis. To verify the psychometric quality of the construct, the internal consistency was assessed computing the Cronbach's alpha coefficient. Temporal stability of the questionnaire was calculated (test-retest). Predictive validity was tested by calculating the correlation matrix and the Pearson correlation coefficient.

To investigate the factor structure of the questionnaire, an orthogonal Varimax type exploratory factor analysis (EFA) (Kaiser, 1958) with a principal-component analysis (PCA) was performed on our questionnaire from the 16 items of the tool (Jackson and Marsh, 1996). The item was retained if the loading was satisfactory, there is to say equal to or greater than 0.40 (Archer et al., 1997). The sampling adequacy was measured computing the Kaiser-Meyer-Olkin (KMO).

A confirmatory factor analysis (CFA) of the first order with maximum likelihood estimation made it possible to verify the factorial structure in four dimensions. In order to test the adequacy of the collected data to the theoretical model, it is generally recommended to use several types of indices (Schermelleh-Engel et al., 2003). Thus Roussel et al. (2002) advise to present at least two indices of adjustment per family of indices. These indices make it possible to evaluate to what extent the theoretical model posited a priori correctly reproduces the data. The most common index is the Chisquare (Satorra and Bentler, 1994), which should be not statistically significant. For a further assessment of the degree of fit of the model, we used the Goodness of Fit Index (GFI) (Marsh et al., 1988; Jöreskog and Sörbom, 1996a,b; Schermelleh-Engel et al., 2003), as well as its adjusted value (AGFI), which should be equal to or greater than 0.90 and 0.85, respectively (Schermelleh-Engel et al., 2003). Then we relied on the Root Mean Square Error of Approximation (RMSEA). This index, differently from the GFI, tests the wrong adjustment. It should be less than 0.05 and 0.08 for a good and acceptable fit, respectively, according to some scholars (Jöreskog and Sörbom, 1996a,b; MacCallum et al., 1996), who also suggest to use the ratio between the Chi-square and the number of degrees of freedom in order to distinguish between the "over-adjusted" and "under-adjusted" models. The target threshold generally proposed by Carvalho and Chima (2014) is ≤3. However, some authors (Roussel et al., 2002) agree on an acceptance threshold of ≤2. Furthermore, we utilized also the standardized root mean square residual (SRMR), which should be ≤0.10 for an acceptable fit (Schermelleh-Engel et al., 2003). Moreover, the Comparative Fit Index (CFI), the Not Normed Fit Index (NNFI), the Normed Fit Index (NFI) and the Parsimony NFI (PNFI) are also particularly relevant especially when it comes to comparing different alternative models. NFI value should be ≥0.90, CFI, NFI, NNFI, and PNFI values should be ≥0.95 (Byrne, 1998; Schermelleh-Engel et al., 2003).

Further, a sensitivity analysis (analysis of variance, ANOVA) was done to assess the impact of age, gender and kind of study on the scores of each dimension of the questionnaire.

Exploratory factor analysis was conducted on a random splithalf sample, whereas CFA was performed on the other splithalf sample. EFA was carried out by means of the commercial software "Statistical Package for Social Sciences (SPSS for Windows, version 24, IBM, Armonk, NY, United States), whereas CFA was conducted with AMOS (version 24, IBM, Armonk, NY, United States).

# RESULTS

# Quality of the Construct

fpsyg-10-02330 October 11, 2019 Time: 16:16 # 4

Data were normally distributed in terms of skewness and kurtosis. Furthermore, our results indicated that the scale of students' academic flow (Heutte and Fenouillet, 2010) had a good temporal stability (r = test and re-test = 0.886).

Predictive validity was tested by calculating the Pearson correlation. The results obtained from the correlation matrix between the 16 statements of the Academic Flow indicated that there was a positive correlation at p < 0.001 between most variables. The coefficient r was between 0.110 and 0.916. For some statements, correlations were found to be good, such as the correlation between Item 12 ("I am not concerned about what others might think of me") and Item 10 ("I'm not concerned about the judgment of others") (r = 0.916 at p < 0.01), as well as the correlation between Item 12 ("I am not concerned about what others might think of me") and item 9 ("I do not care what others may think of me") (r = 0.911 at p < 0.01). However, there are items inversely correlated, such as the correlation between Item 2 ("I feel that I control my actions perfectly") and Item 8 ("("It seems to me that time passes slowly or quickly") (r = –0.135 at p < 0.05), However, there are some coefficients that are small, for example the coefficient of the correlation between Item 3 ("At each step, I know what I have to do") and Item 12 ("I am not concerned about what others might think of me") (r = 0.110 at p < 0.05), as well as the coefficient of the correlation between Item 2 ("I feel that I control my actions perfectly") and Item 15 ("This activity is highly entertaining for me") (r = 0.114 at p < 0.05).

# Exploratory Factor Analysis of the Academic Flow Scale

The results of the EFA showed that the students' academic flow scale well reproduced the expected theoretical model (in terms of homogeneity of items) with an overall good internal consistency (α = 0.886). Each dimension (**Table 2**) had an excellent internal consistency, respectively, of 0.902 (Cognitive), 0.959 (Time), 0.974 (Ego), and 0.960 (Well-being). The KMO indicated a good sampling adequacy (KMO = 0.842 at p < 0.001). The eigenvalue of the dimension of altered perception of time was 4.23, which corresponded to 26.47% of the total variance whereas the eigenvalue of the dimension of well-being was 3.45, explaining up to 48.07% of the total variance. The eigenvalue of the dimension of dilation of ego was 3.38 explaining up to 69.20% of the total variance. Finally the eigenvalue of the dimension


(English translation of the items). Item 1. I feel that I am able to deal with the high requirements of the situation. Item 2. I feel that I control my actions perfectly. Item 3. At each step, I know what I have to do. Item 4. I feel like I can control the situation. Item 5. Sometimes time seems to pass differently than usually. Item 6. I feel that time is running fast when I am busy. Item 7. Sometimes things seem to start to run like a slow motion. Item 8. It seems to me that time passes slowly or quickly. Item 9. I do not care what others may think of me. Item 10. I'm not concerned about the judgment of others. Item 11. I am not worried about my appearance. Item 12. I am not concerned about what others might think of me. Item 13. I have a feeling that I am living a pleasant situation. Item 14. My performance of this activity is highly satisfactory. Item 15. This activity is highly entertaining for me. Item 16. When I talk about this activity, I feel a passion that I would like to share.

fpsyg-10-02330 October 11, 2019 Time: 16:16 # 5

of cognitive absorption was 2.89, explaining up to 87. 29% of the total variance.

# Confirmatory Factor Analysis of the Flow Scale

Our model had a statistically significant Chi-square [χ <sup>2</sup> = 121.542, 96 degrees of freedom at p < 0.05]. The GFI (0.956) was satisfactory, the NFI was 0.979, the NNFI was 0.994. Furthermore, the CFI was 0.996, the AGFI was 0.938, the RMSEA was 0.029, the SRMR was 0.007, and the PNFI was 0.994.

In summary, the 16-item model showed for all the indices tested an excellent fit to the theoretical model, confirming the four-dimensional factor structure for the Tunisian student population (**Figure 1**).

#### Sensitivity Analysis

From the findings of the ANOVA, the sensitivity analysis showed that women had a more altered perception of time with respect to men. There was also an age effect on cognitive absorption, whereas kind of study impacted on all domains except for time. On the other hand, in terms of effects of interaction, no significant influence could be detected (**Table 3**).

### DISCUSSION

The flow as it was conceived by Seligman and Csikszentmihalyi (2000) as a state of deep absorption by the task and intense

fpsyg-10-02330 October 11, 2019 Time: 16:16 # 6

TABLE 3 | Impact of sex, age, kind of study and their interaction effects on the dimensions of the Flow Scale (4D 16).


<sup>∗</sup>Statistically significant at p < 0.05, ∗∗Statistically significant at p < 0.01.

concentration of the subject was later conceptualized as a more complex state, a broader provision, and a domainspecific disposition (Jackson et al., 2008). Over the last two decades, researchers in the field of occupational psychology have increasingly focused on the emergence of flow in the context of work in a wide range of occupations and organizational contexts, including scientists (Quinn, 2005), and teachers (Salanova et al., 2006).

In this context, the researchers have investigated the flow at work and its determinants, including individual differences (Eisenberger et al., 2005), factors related to the working environment (Mäkikangas et al., 2010), as well as the consequences of flow at workplace, such as significant improvements in the psychological well-being of employees (Debus et al., 2014), their performance and achievements.

Despite the richness of these studies on the academic flow, we noticed an absence of work on flow among Arabic-speaking students, which has encouraged us to perform a trans-cultural validation of the scale of the academic flow of Heutte and Fenouillet (2010) in Arabic language.

Our results suggested that the Arabic version of the academic flow scale is a valid and reliable scale for assessing flow among students from different academic sections and both sexes. The results here presented indicated that the scale of assessment of students' academic flow (Heutte and Fenouillet, 2010) had a good internal consistency and temporal stability. These results corroborated those found by other scholars (Karageorghis et al., 2000; Jackson and Eklund, 2002; Jackson et al., 2008).

In addition, the results of the EFA showed that this scale well reproduced the expected theoretical model (in terms of homogeneity of the items) with an interesting internal consistency for each dimension extracted.

Furthermore, the results of the CFA showed an excellent fit to the theoretical model, confirming in a satisfactory way the four-dimensional factor structure in a population of Tunisian university students. However, the CFA presented also some error correlations (namely, between e6 and e8, and between e13 and e16). This could be explained taking into account items formulation and, mainly, the overlap in their content.

However, our study is not without limitations. The Arabicspeaking world is quite vast and culturally different, therefore limiting the investigation to Tunisian subjects could influence the general extensibility of the results. Furthermore, only a sample of students from the same university was studied, even though from various courses and studies sections. As such, future studies are warranted to replicate our findings in a more statistically robust way.

#### CONCLUSION

The objective of our study was to trans-culturally adapt and test the factor structure, internal consistency/reliability, predictive validity, and sensitivity of the "Academic Flow Inventory" (4D 16). Our results showed an excellent internal consistency, a good temporal stability (test-retest), a good correlation matrix, good EFA factor loadings, and excellent CFA fit indices. Moreover, the sensitivity analysis investigated the impact of some variables (age, gender, kind of study) on the domain scores. All these findings enable us to conclude that this scale represents a good psychometric tool that can be used to quantitatively assess the academic flow level in a sample of students in the Arabic-speaking world. However, given the above-mentioned shortcomings, future studies are urgently needed, employing more heterogeneous samples from other Arabic-speaking countries and exploring the relationship of the academic flow with other psychological variables and constructs, especially those related to academic achievement, career expectations and future occupational employment.

#### DATA AVAILABILITY STATEMENT

All datasets generated for this study are included in the manuscript/supplementary files.

# ETHICS STATEMENT

All participants in our study were volunteers, taking part in the study in an anonymous and confidential way. They gave their written, informed consent to be part of the investigation and were carefully and extensively advised about the aim of the study. They were free to withdraw their participation at any moment of the investigation. The study protocol was in-depth reviewed and received the full approval by the ethical committee of the University of Sfax, Sfax, Tunisia. The patients/participants provided their written informed consent to participate in this study.

# AUTHOR CONTRIBUTIONS

NC, NB, and FA conceived the experiment. NC, CA, HK, NB, and FA performed the experiment. NC, NG, NB, and FA analyzed the data. All authors wrote the manuscript.

# REFERENCES

fpsyg-10-02330 October 11, 2019 Time: 16:16 # 7


**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Chalghaf, Azaiez, Krakdiya, Guelmami, Re, Maldonado Briegas, Zerbetto, Del Puente, Garbarino, Bragazzi and Azaiez. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Subjective Happiness and Compassion Are Enough to Increase Teachers' Work Engagement?

Simona De Stasio<sup>1</sup> \*, Caterina Fiorilli<sup>1</sup> , Paula Benevene<sup>1</sup> , Francesca Boldrini<sup>1</sup> , Benedetta Ragni<sup>1</sup> , Alessandro Pepe<sup>2</sup> and Juan José Maldonado Briegas<sup>3</sup>

<sup>1</sup> Department of Human Sciences, Libera Università Maria SS. Assunta, Rome, Italy, <sup>2</sup> Department of Human Studies University of Milano-Bicocca, Milan, Italy, <sup>3</sup> Department of Business Management and Sociology, University of Extremadura, Badajoz, Spain

The present quantitative multi-trait cross-sectional study aims to gain a better understanding of the network of relationship between subjective happiness, compassion, levels of work engagement, and proactive strategies (self- and coregulation) in a sample of teachers. Participants were 187 full-time in-service teachers (89% female; age M = 48.5; SD = 7.88) from Rome, Italy. We hypothesized that subjective happiness and compassion of early childhood teachers would be related with work engagement in such a way that subjective happiness would promote the engagement of teachers. In a similar fashion, we theorized that subjective happiness would be positively related to self- and co-regulation strategies and that proactive strategies would be in turn associated to work engagement. As expected, the results revealed that subjective happiness and compassion showed effects on work engagement and that this association among constructs was mediated by the role of proactive strategies (β = 0.22, p < 0.001; β = 0.37, p < 0.001, respectively). Proactive strategies also have a significant direct effect on work engagement (β = 0.56, p < 0.001). The study's findings suggest the importance of investing in the quality of the working environment.

#### Edited by:

Gabriela Topa, National University of Distance Education (UNED), Spain

#### Reviewed by:

Angelica Arace, University of Turin, Italy Elisabetta Sagone, University of Catania, Italy

> \*Correspondence: Simona De Stasio s.destasio@lumsa.it

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 08 July 2019 Accepted: 23 September 2019 Published: 17 October 2019

#### Citation:

De Stasio S, Fiorilli C, Benevene P, Boldrini F, Ragni B, Pepe A and Maldonado Briegas JJ (2019) Subjective Happiness and Compassion Are Enough to Increase Teachers' Work Engagement? Front. Psychol. 10:2268. doi: 10.3389/fpsyg.2019.02268 Keywords: subjective happiness, compassion, work engagement, Italian early childhood teachers, proactive strategies

# INTRODUCTION

Recent studies suggest that the relational nature of the early childhood education work environment bears key implications for teachers' well-being. Teaching is conducted in a social setting that involves not only direct interaction with children and their parents (Kremenitzer and Miller, 2008) but also cooperation with colleagues, head teachers, and other workgroup members (Leiter et al., 2012). According to Johnson et al. (2012), teachers see the school work environment as comprising three main components: working relationships with colleagues, the school principal's leadership, and school culture, all of which can contribute to teacher work satisfaction and retention rates. Teachers' everyday social interactions in the workplace are characterized by major emotional involvement, while positive relational bonds can represent a crucial protective factor for their job satisfaction and well-being (Saarni, 1999; Albanese et al., 2014; Fiorilli et al., 2015; Benevene et al., 2018a). The literature recognizes that individual dispositional variables, such as perceived subjective well-being and compassion toward others, influence the quality of teachers' working experience (Hargreaves, 2000; Morgeson and Humphrey, 2006; Warr, 2007; Boehm and Lyubomirsky, 2008; Lilius et al., 2008; Underwood, 2009; Benevene et al., 2018b). A positive atmosphere at school fosters

job satisfaction and work engagement (Hoy and Spero, 2005), which, in turn, enhance job performance (Xanthopoulou et al., 2009). A number of studies found that teachers benefit from being compassionate, suggesting that affect plays a key role in work engagement (Brief and Weiss, 2002; Barsade et al., 2003; Hareli and Rafaeli, 2008; Robinson et al., 2013). Furthermore, Nislin et al. (2016) reported that good quality teamwork among teachers helps them to select effective strategies for coping with stressful work conditions. It is therefore of interest to investigate whether dispositional variables, such as subjective happiness and compassion, together with socio-contextual factors, such as self/co-regulated proactive strategies, predict the work engagement of early childhood (0–6 years) teachers.

Fredrickson's (2001) "Broaden-and-build theory of positive emotion" suggests that frequent positive affect triggered by subjective happiness and compassion at work will influence teachers' work outcomes. According to this theory, experiencing positive emotions at work contributes to broadening workers' individual mindsets, enabling them to build up their personal resources in terms of enhanced sensitivity and positive attitudes toward their workplace (Fredrickson, 2013). In educational settings, experiencing positive affect can lead early childhood teachers to form positive emotional associations with their workplace, progressively helping them to view it more positively, and fostering their emotional vigor and organizational commitment (Fredrickson, 2001; Fredrickson et al., 2003).

# DISPOSITIONAL VARIABLES AND WORK ENGAGEMENT

Personal experience in the workplace can be influenced by multiple individual factors: among these, subjective happiness and dispositional positive affect toward others have recently been shown to wield a particularly strong influence (Burger and Caldwell, 2000; Morgeson and Humphrey, 2006; Warr, 2007; Boehm and Lyubomirsky, 2008; Lilius et al., 2008).

Happiness may be viewed as a dispositional measure of subjective well-being; it can help to explain why some individuals report greater self-perceived well-being as a function of life changes, while others report the same amount of well-being regardless of life events (Lyubomirsky and Lepper, 1999). Some authors use the term "subjective wellbeing" in preference to "happiness", on the grounds that the former has been more precisely defined in the scientific literature (Diener et al., 2017). In 1999, Lyubomirsky and Lepper introduced the concept of "subjective well-being" as a dispositional factor capable of explaining individual differences in coding, interpreting, and responding emotionally to life events (Lyubomirsky, 2001). A body of research suggests that perceived subjective well-being can influence the way people adapt to situations, events, and everyday life (Lucas, 2007; Luhmann et al., 2012).

As posited by the above-mentioned "Broaden-and-Build-Theory of positive emotions" (Fredrickson, 2001), subjective well-being contributes to predicting how people will manage and experience their work environment (Lyubomirsky et al., 2005). According to Fredrickson (2001), frequent positive affect leads people to broaden their cognitive and behavioral repertoire and thereby to reinforce their store of personal resources such as self-efficacy, resilience, and optimism. Studies have shown that high dispositional subjective well-being is associated with positive work outcomes, superior performance, success, and higher levels of perceived social support from colleagues (Staw and Barsade, 1993; Staw et al., 1994; Burger and Caldwell, 2000; Boehm and Lyubomirsky, 2008). Happier workers are more likely to engage in altruistic and cooperative behaviors, thus enhancing both the general atmosphere and all-around productivity in the workplace (Borman et al., 2001; Lee and Allen, 2002; Miles et al., 2002; Ilies et al., 2006; Boehm and Lyubomirsky, 2008). They also learn to manage workload and stress more effectively, and make better work-related decisions (Staw and Barsade, 1993; Iverson et al., 1998; Morgeson and Humphrey, 2006; Warr, 2007).

Although several studies have identified a relationship between subjective happiness and work engagement, this association has yet to be investigated in-depth in relation to school and educational settings (Bakker and Oerlemans, 2016). In one of the few studies on the topic, Kim and Shin (2016) noted that early education teachers' happiness contributes to predicting their educational strategies, the quality of their interaction with the children, and their ability to foster positive social interaction in the classroom.

Among the dispositional variables that can affect workplace well-being, "compassion" is widely described in the literature as comprising three components: perceiving other people's suffering; reacting to it in terms of empathizing with the other person's discomfort; and offering a behavioral response with the aim of alleviating the other's suffering (Davis, 1983; Clark, 1997; Kanov et al., 2004; Dutton et al., 2006; Miller, 2007). Lilius et al. (2008) reported that compassionate behaviors in the workplace can have long-lasting effects on how individuals experience their work; compassion may be expressed by providing colleagues with emotional support, or by facilitating them in organizing their work, for example, by allowing them more flexible working hours (ibidem).

Although compassionate behaviors are a characteristic feature in many occupational contexts (Eldor and Shoshani, 2016), only a limited number of studies have examined the relationship between compassion and work-related experience (O'Brien, 2006; Tehan and Robinson, 2009).

Early education settings are particularly dependent on compassionate conduct, given that teachers provide crucial care to infants and children (Kremenitzer, 2005; Sutton and Wheatley, 2003; Kremenitzer and Miller, 2008). Compassion in the education sector has mainly been viewed as an expression of a teacher's attitude toward his/her pupils. However, this is an unduly restrictive definition: teachers also express compassion toward other adults such as their teaching colleagues, and thus can also be the recipients of affection and positive emotions. Compassion among teachers can enhance their feelings of emotional connection to their work and strengthen the association between organizational support and work commitment (Eldor and Shoshani, 2016). Some authors have investigated how compassion and proactive coping strategies

are related to one another, as well as their effects on teacher adaptability and job satisfaction, and general classroom atmosphere (Mason et al., 2014; Mauno et al., 2016).

# THE ROLE OF PROACTIVE STRATEGIES IN EARLY CHILDHOOD TEACHERS' WORK ENVIRONMENT

In the school setting, the term "coping strategies" refers to teachers' cognitive and behavioral efforts to reduce, tolerate, or manage work-related stress (Lazarus, 1993; Sharplin et al., 2011; Parker et al., 2012).

"Self-regulated" coping strategies are generated by individual teachers, with a view to managing stressful work situations (Zimmerman, 2002; Boekaerts and Corno, 2005). When teachers manage stressful scenarios by drawing on the social resources made available to them by their community of colleagues, these strategies may be defined as "co-regulated" (Järvelä et al., 2010).

Considering the highly social nature of educational and school settings, it is reasonable to assume that teachers typically coregulate their behavior in order to effectively cope with stressful conditions. Soini et al. (2010) found that, in schools, teachers who feel positively toward their work environment often use externally oriented problem-solving strategies and co-regulate their behavior with colleagues. Teachers who report high levels of support from colleagues are less likely to intend giving up teaching, as compared to those who report finding low levels of social support in the work setting (Pomaki et al., 2010). Proactive strategies help individuals to appropriately direct their behaviors, attain pre-set goals, and reduce their risk of burnout (Xanthopoulou et al., 2009; Devos et al., 2012).

When teachers experience job satisfaction and motivation, this generates a positive social atmosphere (Alhija and Fresko, 2010; Skaalvik and Skaalvik, 2011). According to Fredrickson (2001), positive affect fosters more creative and flexible use of available strategies. Teachers who are engaged with their work are significantly more inclined to deploy active and reflective approaches to problem-solving (Soini et al., 2010). The use of proactive strategies not only reduces the risk of burnout but also contributes to the development of a positive work environment and to increased job involvement. Their degree of efficacy appears to be closely related to the quality of teachers' social interactions, suggesting that co-regulation plays a key role in prompting and sustaining proactive intervention.

# WORK ENGAGEMENT

Work engagement may be defined as "a state of fulfilment characterized by vigor, dedication, and absorption" (Schaufeli et al., 2002). The term "vigor" describes an energetic, effortful, and persistent approach to one's work; "dedication" denotes a high level of involvement, accompanied by a sense of meaningfulness, enthusiasm, and inspiration; the term "absorption" indicates concentration and feeling engrossed in one's work, such that the time spent working passes quickly (ibidem).

Work engagement is a predictor of work-related wellbeing, at both the individual and organizational levels (Bakker et al., 2008). In the school setting, work engagement strongly predicts job performance, reduced intention to quit teaching, positive teacher–student relationships, and student academic achievement (Bakker et al., 2006, 2008; Duckworth et al., 2009; Salanova et al., 2010).

Potential links between individual dispositional variables, such as subjective well-being and levels of work-engagement, have been explored in the literature (Field and Buitendach, 2011; Huynh et al., 2014; Rodríguez-Muñoz et al., 2014; Bakker and Oerlemans, 2016), but inadequately so in relation to educational and school settings (Pillay et al., 2005; Chan, 2009). Indeed, to date, subjective well-being in teachers has been viewed as an outcome of work engagement, rather than one of its potential antecedents (Tadic et al., 2013 ´ ; Sirisunhirun and Dhirathiti, 2015).

Among individual dispositional resources, compassion at work has emerged as a strong predictor of work engagement (Mason et al., 2014; Mauno et al., 2016), being positively associated with teachers' levels of vigor, dedication, and satisfaction. It is also negatively associated with burnout, reflecting its status as a key resource for coping with stress-related burden (Eldor and Shoshani, 2016). Receiving constructive feedback and professional recognition from colleagues and superiors significantly affects job performance (Klusmann et al., 2008; Xanthopoulou et al., 2009), promoting job engagement and satisfaction (Peeters and Rutte, 2005; Kokkinos, 2007; Stoeber and Rennert, 2008).

# AIMS OF THE STUDY

The present quantitative multi-trait cross-sectional study aims to gain a better understanding of the network of relationship between subjective happiness, compassion, levels of work engagement and proactive strategies (self- and co-regulation) in a sample of early childhood teachers. More specifically, we were first interested in evaluating the degree of fit of an integrated structural model with empirical data and then to assess the cumulative network by decomposing total effects in direct and indirect effects among considered variables.

Based on previous research, we expected that subjective happiness and compassion of early childhood teachers would be related with work engagement in such a way that subjective happiness would promote the engagement of teachers. In a similar fashion, we theorized that subjective happiness would be positively related to self- and co-regulation strategies and that proactive strategies would be in turn associated to work engagement. Finally, and most germane for the present paper, the representation of the network of association in an integrated structural model supported the theoretical viewpoints (Borrelli et al., 2014) considering subjective happiness, work engagement as dynamically shaping the social endeavor in which teachers are involved during their daily profession.

# MATERIALS AND METHODS

fpsyg-10-02268 October 15, 2019 Time: 17:22 # 4

#### Participants and Procedure

Our sample was composed of 187 full-time in-service teachers (89% female) from Rome, Italy. Ages ranged from 27 to 63 (M = 48.5; SD = 7.88) In terms of marital status, 56.5% were married, 21.6% were single, 18.4% were separated/divorced, and 1.6% were widowed; 65% of participants had children. Length of teaching experience ranged from 1 to 32 years (M = 17.23 years, SD = 14.23). The study population was a convenience sample and may not be taken as representative of the entire population of Italian teachers given that all participants were based in Central Italy. The authors organized plenary assemblies in kindergarten schools to inform the teachers about the aims of the research and the procedures for completion of the questionnaires. All participants signed informed consent forms and were assured anonymity and confidentiality. The research protocol was approved by the Ethics Committee of LUMSA University, Rome. The original versions of questionnaires were initially translated from English into Italian and then back-translated into English to check the alignment with the original versions.

#### Measures

The Subjective Happiness Scale (Lyubomirsky and Lepper, 1999) is a four-item scale that assesses subjective happiness, using a seven-point Likert scale. The first two items ask people to rate how they are generally happy about their life (1 = not a very happy person, 7 = a very happy person) and how happy they are in comparison with their peers (1 = less happy, 7 = more happy) (e.g., "Compared with most of my peers, I consider myself less happy . . . more happy").

The last two items ask respondents to what extent the characterization of a happy and of an unhappy person describe themselves (1 = not at all, 7 = a great deal). Higher scores on this measure indicate greater subjective happiness. It is utilized in the Italian version adapted by De Stasio et al. (2017). Prior studies have reported Cronbach's alpha coefficients for the SHS from 0.79 to 0.94 (Lyubomirsky and Lepper, 1999).

The Santa Clara Brief Compassion Scale (Hwang et al., 2008) is a five-item scale; it assesses compassion and its link to prosocial behaviors. The scale includes items related to two facets of compassion: "emotionally connecting with other people's suffering" (e.g., "When I hear about someone going through a difficult time, I feel a great deal of compassion for him or her") and "acting to help them" (e.g., "I often have tender feelings toward people when they seem to be in need"). It is a shortened version of Sprecher and Fehr's (2005) Compassionate Love Scale (the correlation between the two scales is r = 0.95) and it refers to non-intimate (i.e., strangers) rather than to close others. It is utilized in the original version translated in Italian through back-translation carried out by the authors. All items were rated on a seven-point Likert-type scale ranging from 1 (completely disagree) to 7 (completely agree), and higher scores are indicative of greater compassion. Cronbach's alpha was 0.87.

The Utrecht Work Engagement Scale [UWES-17; Schaufeli et al., 2006, Italian version of UWES-17, validated by Balducci et al. (2014)] assesses work engagement. The scale is composed of 17 items, grouped into three subscales, namely, vigor (six items), dedication (five items), and absorption (six items) (e.g., "At my job, I feel strong and vigorous"). All items are scored on a sevenpoint frequency rating scale ranging from 0 (never) to 6 (always). Cronbach's alpha was 0.96.

The Proactive Strategy scale (Salmela-Aro, 2009) consists of seven items, measuring two factors of proactive strategies: (a) self-regulation (4 items) and (b) co-regulation (3 items), meaning, respectively, the ability to identify and use resources for coping with stressors and the ability to seek and receive social support from colleagues (e.g., "I can set limits to my work assignments"; "I'm able to support the colleagues who feel strain in their work"). It is utilized in the original version translated in Italian through back-translation carried out by the authors. All items are rated on a seven-point Likert-type scale, ranging from 1 (completely disagree) to 7 (completely agree). Cronbach's alpha was 0.67.

# Analytical Strategy

#### Preliminary Analysis

First, chi-square tests and correlational analysis were conducted to identify potentially confounding interrelationships among participants' demographic characteristics (age and gender) and the measures under study. Second, the statistical distribution of the data was explored for each of the measures. None of the kurtosis and skewness values were in excess of the recommended cutoffs [−1, +1]. Mahalanobis' distance (p < 0.001) was calculated for all scores in order to identify and skip any multivariate outliers. As a result, one extreme multivariate value was omitted from the analyses.

#### Structural Equation Modeling

The cumulative network of relationships among the variables of interest was analyzed via structural equation modeling (SEM), implemented using AMOS (Arbuckle and Wothke, 2006), version 21.0. This analytical strategy involves statistically testing a hypothesized set of direct and indirect paths among variables to evaluate the extent to which it fits the empirical data, yielding a measure known as goodness of fit. In the current study, we followed standard recommendations for the evaluation of a given SEM model (Bagozzi and Edwards, 1998; Hu and Bentler, 1999; Hair et al., 2010) by adopting two different types of fit index: absolute and relative. The absolute indexes selected were χ 2 and normed-χ 2 (NC) [a non-statistically significant χ 2 value and NC values of under 2.0 indicate good fit (Hair et al., 2010)]. The relative indices adopted were the root mean square error of approximation (RMSEA), normed fit index (NFI), non-normed fit index (NNFI), comparative fit index (CFI), and standardized root mean square residual (SRMR). In this case, the thresholds set for good model fit were: RMSEA < 0.07 (Schermelleh-Engel et al., 2003), NFI > 0.95, NNFI > 0.95 (Marsh, 2004), and CFI > 0.95 (Hu and Bentler, 1999). Finally, in keeping with current literature on the use of SEM (e.g., MacKinnon et al., 2004), we estimated confidence limits using both Monte Carlo simulation and bootstrapping methods with a set of random samples (k = 500).

# RESULTS

**Table 1** reports correlations and descriptive statistics (means and standard deviations) for all the variables under study. Subjective happiness was statistically significant and slightly correlated with teachers' self-regulation strategies (r = 0.15, p < 0.005), coregulation strategies (r = 0.19, p < 0.001), and work engagement (r = 0.21, p < 0.001). In a similar fashion, the correlations between compassion and work engagement (r = 0.42, p < 0.001), teachers' self- regulation (r = 0.28, p < 0.001), and co-regulation strategies (r = 0.31, p < 0.001) were positive and statistically significant. Both teachers' self-regulation (r = 0.38, p < 0.001) and co-regulation (r = 0.47, p < 0.001) strategies were significantly correlated with work engagement.

The estimation of the conceptual model (see **Figure 1**) allowed us to test the hypothesis that the pattern of association between measures of subjective happiness, compassion, proactive strategies, and work engagement can be modeled within a single integrated model. SEM revealed that relative indexes of fit were generally robust [χ 2 (33) = 0.279, p = 0.870; NC = 0.14], suggesting a good degree of fit between the conceptual model and the actual empirical data. Further information about the practical significance of the model was provided by analysis of absolute indexes (RMSEA = 0.003, pclose = 0.913, SRMR = 0.001, NFI = 0.998, CFI = 0.999), all of whose values fell within the recommended cutoff points. The analysis of total effects (standardized weights) and the subsequent decomposition in direct and indirect effects estimated by the structural equations model allowed us to explore more specific research hypotheses. The results suggested that subjective happiness wielded a positive total effect [β = 0.22, p = 0.005, 95th C.I. (0.607–1.18)] on work engagement, composed of a non-statistically significant direct effect [β = 0.04, p = 0.590, 95th C.I. (-0.209–0.448)] and of an indirect effect via proactive strategies [β = 0.18, p = 0.006, 95th C.I. (0.420–1.22)]. The latent endogenous variable compassion strategies reported a statistically significant total and positive standardized total effect on work engagement [β = 0.37, p = 0.014, 95th C.I. (0.701–1.28)], composed of both a significant direct effect on work engagement [β = 0.15, p = 0.013, 95th C.I. (2.26– 4.47)] and a significant indirect effect via proactive strategies [β = 0.22, p = 0.009, 95th C.I. (0.148–0.810)]. Compassion strategies also reported a statistically significant direct effect on proactive strategies: β = 0.39, p = 0.011, 95th C.I. (0.047–0.192).

TABLE 1 | Bivariate correlations and descriptive statistics (means and standard deviations) for all the variables under study.


∗∗p < 0.01; <sup>∗</sup>p < 0.05.

With regard to the proactive strategies, the results evidenced their medium, positive, and statistically significant total effect on work engagement [β = 0.56, p = 0.013, 95th C.I. (2.26–4.47)], meaning that the greater teachers use self- and co-regulated strategies, the higher their levels of work engagement.

# DISCUSSION

In the present study, we set out to investigate the network of relationships between subjective happiness, compassion, work engagement, and the use of proactive strategies (both self- and co-regulated) in a sample of early childhood teachers. We found that subjective happiness and compassion had direct positive effects on work engagement. Furthermore, the effect of subjective happiness and compassion on work engagement was mediated by the deployment of proactive strategies. Early childhood teachers' self- and co-regulation (i.e., their ability to identify and use resources for coping with stressors as well as to seek and receive social support from colleagues) also had a direct positive effect on work engagement, lending support to the idea that using such strategies enhances early childhood teachers' degree of involvement in their work. We now discuss the different roles played by the variables under study considering their direct and indirect effects on levels of work engagement.

# The Direct and Indirect Effects of Subjective Happiness and Compassion on Work Engagement

Our data confirm previous findings reported in the literature about, on the one hand, the relationship between positive emotions (triggered by subjective happiness and compassion at work) and proactive strategies and, on the other, the link between the use of proactive strategies and personal involvement in one's work. For example, the present results are in line with studies reporting the beneficial effects of positive emotions on individuals' behavioral and cognitive repertoires (Fredrickson, 2001), as well as the effects of personal and relational proactive strategies on engagement (Salmela-Aro et al., 2011).

Considering that work engagement is the opposite of burnout, the present findings are also in line with previous research identifying subjective happiness and compassion as key personal resources for coping with work-related stress and burnout (Field and Buitendach, 2011; Huynh et al., 2014; Bakker and Oerlemans, 2016). Vice versa, past studies found that depressive symptoms are negatively related to work engagement (Upadyaya et al., 2016), again indirectly confirming the importance of subjective happiness as a contributor to work engagement. Indeed, subjective happiness has been shown to enhance both teaching and broader life experience, generating work engagement by fostering more effective deployment of job resources (Bakker et al., 2014). Nevertheless, to the best of our knowledge, subjective happiness has not previously been investigated as a potential antecedent of work engagement among kindergarten teachers.

A further key finding of this study was that compassion had a positive effect on work engagement. Again, this finding is in line with previous studies showing compassion to be a

correlate of work commitment (Mason et al., 2014; Eldor and Shoshani, 2016). In this respect, Tremblay and Messervey (2011) argued that compassion may generate work engagement by buffering the impact of job demands on work-related strain. More generally, personal resources such as subjective happiness and compassion may determine how job resources—the strongest predictors of work engagement—are perceived (Lorente et al., 2014; Diener et al., 2015).

The effect of compassion and subjective happiness on work engagement is greatly increased when mediated by coping strategies.

Overall, our findings suggest that the self-reported deployment of self- and co-regulated proactive strategies on the part of early childhood teachers is positively associated with their work-related well-being, in terms of self-perceived work engagement. Existing studies suggest that teachers who experience positive involvement in their work are more inclined to draw on active and reflective problemsolving strategies (Soini et al., 2010; Simbula et al., 2011). Furthermore, experiencing positive emotions prompts more creative and flexible use of available coping strategies (Fredrickson, 2001), enabling teachers to tackle challenges in innovative and original ways (Lyubomirsky et al., 2005; Linnenbrink-Garcia et al., 2011).

In keeping with the present findings, adopting proactive strategies has been associated with more effective regulation of behavior and greater adaptability to the work environment on the part of teachers. In fact, the deployment of proactive strategies is related to the ability to manage work-related burden and feel positively involved in it (Butler, 2007; Salmela-Aro et al., 2011; Devos et al., 2012), even when workload is perceived as particularly intense (Pomaki et al., 2010).

The literature suggests that the use of proactive strategies may also be associated with lower levels of perceived stress and more effective use of existing personal and relational resources on the part of teachers (De Caroli and Sagone, 2012; Pietarinen et al., 2013; Bermejo-Toro et al., 2016; Tikkanen et al., 2017). A proactive personal attitude is linked with work engagement and takes the form—for example—of actively seeking support from one's teaching colleagues and setting stimulating goals (Bakker et al., 2012). Pietarinen et al. (2013) examined the relationship in a sample of teachers—between their deployment of proactive strategies, workplace difficulties, and perceived adaptability.

The present findings confirm the role of proactive strategies as protective factors. Again, this bears out previous research suggesting that when teachers are able to use multiple coping strategies, this protects the teacher community from emotional exhaustion (Kyriacou, 2001; Howard and Johnson, 2004; Austin et al., 2005; Carmona et al., 2006).

Thus, self-perceived work engagement may be seen as a social outcome that is dependent on teachers' subjective wellbeing at work and may be significantly enhanced by the deployment of self- and co-regulative strategies on the part of the community of education practitioners. Using proactive strategies allows teachers to simultaneously regulate their own behavior and their working environment, thus enhancing working environment fit.

Ad hoc interventions for fostering the use of proactive strategies should be designed, with a view to reducing teachers' risk of burnout and increasing their engagement and positive involvement in the educational setting (Pietarinen et al., 2013).

# CONCLUSION AND IMPLICATIONS

The present study advances our understanding of early childhood teachers' work engagement, by evaluating the contributions of subjective happiness, compassion, and proactive strategies to work engagement within a single model. It shows that subjective happiness and compassion at work trigger positive feelings,

which in turn contribute to enhancing pre-school teachers' attitudes and work outcomes. Another key outcome is that the dispositional variables happiness and compassion need to be mediated by proactive strategies, undertaken both individually and jointly with colleagues, for teachers to attain a fuller sense of work engagement. A novel aspect of the current research is that it specifically analyzed the work-related well-being of a group of early childhood teachers, a population whose work engagement has been little investigated in the literature to date (Hall-Kenyon et al., 2014; Cumming, 2017). Nonetheless, it is well known that children's well-being is closely related to teachers' wellbeing and work engagement: for example, when teachers perceive their work community in a positive light, this is associated with better classroom teaching quality (McGinty et al., 2008). Early education environments are characterized by multiple stressors including the children's needs, relationships with families and colleagues, and organizational issues (Kelly and Berthelsen, 1995; Rentzou, 2012; Jovanovic, 2013; Faulkner et al., 2014; Nislin et al., 2016), with potentially detrimental repercussions on teacher–child interaction and infant development (e.g., De Schipper et al., 2008).

In conclusion, the current findings suggest that early childhood teachers are more inclined to reinforce their work engagement by drawing on positive workplace relationships than by relying on their dispositional characteristics. Positive interpersonal relationships in the school setting promote work engagement and protect early childhood teachers from the risk

# REFERENCES


of burnout. Based on these findings, we strongly recommend intervention in the domains of personal resources and workrelated well-being in the interest of optimizing early childhood teachers' work engagement.

# DATA AVAILABILITY STATEMENT

All datasets generated for this study are included in the manuscript/supplementary files.

# ETHICS STATEMENT

The study involving human participants was reviewed and approved by the Scientific Board of LUMSA University, Rome. The participants provided their written informed consent to participate in this study.

# AUTHOR CONTRIBUTIONS

SD, CF, and PB designed and carried out the study, contributed to the analysis of the results and to the writing of the manuscript. BR, FB, AP, and JM collected the data, and contributed to the analysis of the results and to the writing of the manuscript. PB, SD, and JM supervised the study design and the manuscript draft.



emotions following the terrorist attacks on the United States on September 11th, 2001. J. Pers. Soc. Psychol. 84, 365–376. doi: 10.1037//0022-3514.84.2.365


and Perennial Issues, ed. R. S. Lazarus, (Mahwah, NJ: Lawrence Erbaum Associates Inc), 366–388.



**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 De Stasio, Fiorilli, Benevene, Boldrini, Ragni, Pepe and Maldonado Briegas. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Teaching in the Suburbs: Participatory Action Research Against Educational Wastage

Santa Parrello<sup>1</sup> \*, Ilaria Iorio<sup>1</sup> , Filomena Carillo<sup>2</sup> and Cesare Moreno<sup>2</sup>

<sup>1</sup> Department of Humanities, University of Naples Federico II, Naples, Italy, <sup>2</sup> Non-profit Association Maestri di Strada, Naples, Italy

#### Edited by:

Simona De Stasio, Libera Università Maria SS. Assunta, Italy

#### Reviewed by:

Viviana Sappa, Swiss Federal Institute for Vocational Education and Training, Switzerland Antti Rajala, University of Helsinki, Finland

> \*Correspondence: Santa Parrello parrello@unina.it

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 20 May 2019 Accepted: 27 September 2019 Published: 18 October 2019

#### Citation:

Parrello S, Iorio I, Carillo F and Moreno C (2019) Teaching in the Suburbs: Participatory Action Research Against Educational Wastage. Front. Psychol. 10:2308. doi: 10.3389/fpsyg.2019.02308 If teaching is a stressful job, it can be even more so in schools in disadvantaged areas, such as the metropolitan suburbs, where the rates of educational wastage are high. Here, teachers often feel ineffective: as a result, there is a reduced sense of well-being at work, which triggers a negative cycle that damages their educational performance. From the literature, it is known that teachers need social support, which has a positive effect on well-being and resilience. For these reasons, the Association "Maestri di Strada" (MdS) has chosen to offer teachers professional social support and to actively involve them through Teacher Participatory Action Research (T-PAR): the "Crossing Educational Boundaries" project. These are the research questions that gave life to the project: do the teachers have resources to analyze the problematic situations they are immersed in and to build improvement strategies? Would a professional social support reinforce their resilience? The objective was the following: to actively engage the teachers in order to generate hypotheses concerning the causes of educational wastage in their schools, and to work with them to plan new methods to lessen the problem. The project was carried out in 12 suburban secondary schools in six Italian cities. This paper illustrates the activities of three cities. All phases of the T-PAR were completed. The teachers organized discussion groups and started workshops in the classes considered at risk. The activities were subject to non-participant observation, and the observation reports underwent semantic-structural analysis. Four clusters emerged the analysis. The results show that the teachers are aware of the importance of a good educational relationship as a way to oppose educational wastage, and, at the same time, they are aware of the difficulties of building it, which they attribute to the mistrust and passiveness of the pupils, and to the demands of the institution. The moments of discouragement shown by the teachers, and their strong emotional engagement in the pupils' difficulties are significant. At the end of the project, a small group of teachers planned and implemented a reflective space in some of the schools.

Keywords: teacher participatory action research, prevention of educational wastage, suburban schools, teachers' resilience, teachers' self-reflection, ALCESTE text analysis

# INTRODUCTION

fpsyg-10-02308 October 17, 2019 Time: 18:23 # 2

Teaching is a stressful and tiring job for a complex variety of reasons linked to multiple factors, including: the nature of the work, which implies a high degree of relationality; personal characteristics, on both a social and emotional level; and school organization, which is influenced by the historical and geographical context (Kyriacou, 2001; Hastings and Bham, 2003; Hakanen et al., 2006; Bizumic et al., 2009; Chang, 2009; Day and Gu, 2009; Zurlo et al., 2016; Steffens et al., 2017; Capone et al., 2019; Parrello et al., 2019a). To this end, an extensive array of literature testifies to the importance of cultivating well-being and resilience in teachers, and identifies protection factors and intervention strategies.

In particular, resilience in teachers is defined as a quality that enables teachers to maintain their commitment to teaching and to their teaching practices, despite challenging conditions and recurring setbacks (Brunetti, 2006). Resilient teachers have been described as those who have the capacity to thrive in difficult circumstances, are skilled in behavior management, are able to empathize with difficult students as well as restrain negative emotions and focus on the positive, and who experience a sense of pride, fulfilment, and increased commitment to their school and profession (Howard and Johnson, 2004). Resilience involves the capacity of an individual teacher to harness personal and contextual resources to navigate through educational challenges and to facilitate the outcome of professional engagement, growth, commitment, enthusiasm, satisfaction, and wellbeing (Beltman, 2015). In recent years, researchers have begun to conceptualize resilience from a social-ecological perspective, wherein resilience is defined as a set of behaviors over time that reflect the interactions between individuals and their environments, in particular the opportunities for personal growth that are available and accessible (Ungar, 2012). Some authors speak of relational resilience (Le Cornu, 2013; Gu, 2014): in particular, reciprocal and mutually supportive personal, professional and peer relationships are important in this process (Sammons et al., 2007). The outcome is that teachers maintain job satisfaction and commitment to their profession (Brunetti, 2006). Many authors do not explicitly examine resilience, but do address the question of what sustains teachers and what enables them to thrive rather than just survive. These papers could be grouped into three categories, each with a different focus: those emphasizing individual factors; contextual factors; and individual perceptions of, and responses to, the specific contexts of teacher work (Beltman et al., 2011).

The urban suburbs are among the contexts considered at risk. If teaching is always difficult, it can be even more so in schools in these disadvantaged areas. In suburban schools, teachers have to deal with adolescents of lower socio-economic status who often experience distress in their personal lives as well as at school (Sommantico et al., 2015; Pellerone et al., 2018; Lavy and Ayuob, 2019), and who have an internalized marginality which becomes a learned sense of powerlessness (David, 2014). Teaching in these environments therefore means: interacting and connecting daily with young people who are struggling with problems and pain, as well as working to decrease educational wastage. In these environments, we find both material and educational poverty: the material poverty of one generation often causes the deprivation of educational opportunities for the next (Jones, 2003; Lott, 2012; OCSE, 2015). In order to break this vicious circle, schools in deprived areas will need more funds and investments. In Italy, on the contrary, schools populated by disadvantaged students tend to have fewer resources (OCSE-PISA, 2012). Consequently, we can identify a socio-economic map showing the distribution of young people with a serious deficiency in the fundamental skills needed to grow and work in the world, to which we can also add a map of early school leavers.

According to UNESCO (1972), educational wastage is the combination of repeated grades and early school leaving. In Italy, this phenomenon reaches alarming levels: in 2014, the "Educational Wastage" Dossier complied by Ministry of Education, University and Research [MIUR] (2014) observed the presence of almost three million young people who did not complete upper secondary education. In 2015, Italy reduced the rate of early school leavers, although the percentage was still higher than the EU average (Istat, 2015). Over time, experts have highlighted different causes for educational wastage. In the '60s and '70s, they focused especially on its socio-economic factors. In the '80s, they additionally considered some of the students' subjective dimensions. Over the same period, researchers began studying the relational dimension concerning students and teachers: increased attention was paid to classroom management styles, to the active or passive role of the students, and to the language used in the educational environment (Batini and Bartolucci, 2016). If we take this relational standpoint, we cannot speak of "wastage" in the singular, but rather of "wastages," plural (Perone, 2006), or of a wasteful system: the situation does not concern only those who leave school entirely, but also those who attend without learning, as in the case of in-school drop-outs (Solomon, 1989). The waste is not limited to the potential of students: it also extends to the work of teachers and to school equipment (Brimer and Pauli, 1971). Contributions from psychoanalysis, which study affective/emotional dynamics, processes of symbolization in inter-generational relationships, and relationships with institutions, belong to the same perspective (Blandino and Granieri, 1995). Depending on the different theoretical approaches, a succession of interventions aimed at strengthening teaching have been employed over the years, for example through the repetition and simplification of the material being taught, as well as interventions aimed at motivating and re-motivating students. There have also been interventions aimed at improving the quality of the educational relationship, which encompassed a variety of the teacher's relational skills, e.g., their emotional skills (Nouwen et al., 2016; Molloy Elreda et al., 2019). Fewer studies have focused on the importance of developing resilience in teachers if they are, in turn, to foster this trait in students (Bobek, 2002; Henderson and Milstein, 2003), even though resilience as a personal characteristic has been studied extensively in at-risk students (McMillan and Reed, 1994; Johnson, 1997; Aronson, 2001). Day and Gu (2014) have cogently argued that "efforts to increase the quality of teaching and raise standards of learning and achievement for all pupils must focus on efforts to build,

sustain and renew teacher resilience, and that these efforts must take place in initial teacher training" (p. 22).

# The Work of the Association MdS to Contain Educational Wastage

The non-profit Association Maestri di Strada (MdS) was born in Naples in 2003 with the Chance Project, a second chance school for drop-outs, recognized as an "activity of excellence" by the European Union: MdS carries out complex socioeducational interventions inside and outside suburban schools, in order to prevent educational wastage and to promote social inclusion (De Rosa et al., 2017). In the last 10 years, in line with several European approaches to reduce early school leaving (Nouwen et al., 2016), MdS has chosen to work especially in schools in the eastern suburbs of Naples: they support teachers in the classroom and share different methods with them.

The environment of the eastern suburbs of Naples is characterized by social and economic marginality, unemployment, environmental neglect, a lack of public services, the widespread presence of organized crime, and educational wastage. In this context, teachers feel ineffective or powerless, and as if their job has lost meaning. It is not easy to work in these environments for MdS, either. The relationships with children, teenagers, families and teachers, all charged with problems and distress, are trying. It is just as stressful to connect with educational or political institutions, because it entails facing different forms of inflexibility. MdS takes care of the well-being of its partners and encourages their professional reflection, communication and cooperation through observations, narrations, and group meetings. Observation (McMahon and Farnfield, 2010) is employed both for on-field educational activities and for reflective group meetings. All partners are also asked to write narrations about their work throughout the year. The multi-vision group (Parrello et al., 2019b) is the method chosen to support professional reflection: it was inspired by the Balint group (Van Roy et al., 2015). The results of the Association's efforts are encouraging: at the end of each school year, 90% of the kids they care for stay in school and succeed in moving on to the next school year (Parrello, 2018).

Spending considerable time inside of schools, MdS partners become witnesses of the hardships of many teachers and suggest activities to support their work. For these reasons, MdS has chosen to offer teachers professional social support, to support their resilience, and to actively engage them through T-PAR methodology.

From the literature, it is known that teachers need social support, both internal and external to the school system. In fact, good professional relationships have a positive effect on wellbeing, sense of engagement, empowerment, self-efficacy and the resilience of teachers in their work (Betoret, 2006; Halbesleben, 2006; Skaalvik and Skaalvik, 2009; Soini et al., 2010; Brouwers et al., 2011; Avanzi et al., 2018). The resilience of teachers is linked to the social support that they receive from colleagues, family and other professionals (Stanford, 2001), as well as to a "relevant, rigorous and responsive" education (Cefai and Cavioni, 2014, p. 144). The T-PAR methodology is a form of teachers' education via Action Research.

Action Research (AR), as it is known, is an investigation model whose main aim is to improve the future skills and activities of the researcher, rather than produce theoretical knowledge (Lewin, 1948). In the '50s, Corey (1953) promoted AR in the United States in the field of education, gaining great cooperation from school districts and teachers: his method was later called Cooperative Action Research. At the end of the '60s, an international network of researchers created Participatory Action Research (PAR) in order to tackle various problems with disadvantaged members of society (Susman and Evered, 1978; Hall, 1992; Fals-Borda, 2001; Kidd and Kral, 2005; Arcidiacono et al., 2016). McTaggart (1997) explains that the dual objective of PAR is, on the one hand, to intervene in the situation being researched, and, on the other, to change the researchers themselves, activating in them a process of transformation that will make them agents of their own changes. From then on, PAR has been applied to many fields and disciplines, and it has assumed a much more critical position with respect to the broader field of AR, given its specific objective of dealing with power imbalances that generate social and personal distress (Arcidiacono et al., 2017; Stapleton, 2018). For this reason, the PAR method is thought to be particularly suited to the educational field (T-PAR), where power issues are a constant influence (Jacobs, 2016). Hooks (1984) discusses the power issue found in traditional educational environments: there are power discrepancies between teacher and pupil that start with teachers grading their pupils; however, this power can be used in non-coercive ways to improve the learning process.

Teachers often perceive a divide between theory and practice, between being able to think and having to do, whereas in PAR, knowledge is built by the people involved in the research process, in a non-hierarchical, democratic environment which ultimately constitutes "a social enterprise" (Savin-Baden and Wimpenny, 2007) that produces "a contextual knowledge" (Pine, 2009, p. 31). PAR can change both teachers and students, and also their mutual perception (Brydon-Miller and Maguire, 2009) and the emotions dominating the classroom (Hooks, 1994). Moreover, according to multiple authors, in-classroom research is in itself a way to promote self-reflection (Schön, 1983; Alber and Nelson, 2002), improving the teachers' self-awareness, their control over their emotions and actions and over their own power, thereby reducing the pressure created by the context and finding resources where they previously only saw limits. More specifically, reflection – as part of the PAR cycle – is a meta-cognitive process that consists of exploring personal beliefs, thoughts and actions in a deliberate, autobiographical, and critical way (Marcosa et al., 2009). Thus, PAR is considered to be a powerful form of professional development for teachers (Johnson and Button, 2000), who are nevertheless usually reluctant to participate in action research: they do not comprehend how research could improve their work, because they lack the knowledge and training to see the connection between theory and practice (Bondy, 2001).

The PAR cycle is composed of the following phases: planning, action, reflection and evaluation. Some common characteristics

of PARs are active participation, open-ended objectives and high levels of commitment (Greenwood et al., 1993; Morales, 2016). Multiple authors report the following various benefits of action research, which often go beyond the goal of the research project: improved teaching practice; enhanced collegiality; feelings of closeness to one another after working on a group research project; and becoming more reflective about the improvement of student performance (Glanz, 2003).

# MATERIALS AND METHODS

In the academic year 2017–2018, MdS carried out the T-PAR project "Crossing Educational Boundaries": the title is a reference to the importance of making education "cross the boundary" and leave the outskirts of society's interests; it also alludes to the necessity of going beyond physical and mental limits, thus letting professionals, disciplines and methods meet in a free and creative way.

These are the research questions that gave life to the project: do the teachers have resources to analyze the problematic situations they are immersed in and to build improvement strategies? Would a professional social support reinforce their resilience?

The objective was the following: to actively engage the teachers in order to generate hypotheses concerning the causes of educational wastage in their schools, and to work with them to plan new methods to lessen the problem.

#### Participants

The schools that participated are all located in the suburbs and have a high rate of educational wastage. 12 suburban schools participated, with about 200 teachers and 20 classes from different Italian cities: two metropolises (Rome, Naples), two mediumsized cities (Bologna, Florence), and two smaller provincial towns from the North and South (Rozzano, Sciacca). This paper will describe the first phase of the PAR project carried out in the first three participating schools – Rome, Bologna, Naples.

#### The School in Rome

Rome is a metropolis (population of city and provinces: almost four and a half million) in central Italy. The school that participated in the project is a Higher Technical Institute in a suburban area. The students are between 14 and 18 years old, and many do not obtain a high school diploma. Here, 15 teachers participated in the PAR.

#### The School in Bologna

Bologna is a medium-sized city (population of city and province: about 1 million) of Northern Italy. The school that participated in the project is a Lower Secondary School (students between 12 and 14 years old) located in a suburban area, where the teenagers attending are mostly children of immigrants. A few years ago, some Italian students also returned to the school, because the principal added a class with a curriculum using high-technology teaching. However, the result was a clear-cut separation between Italians and foreigners. Here, 14 teachers participated in the PAR.

#### The School in Naples

Naples is a metropolis (population of city and province: about 3 million) in Southern Italy. Its suburbs are known for the high rate of educational wastage, unemployment and organized crime. In this Lower Secondary School, teachers have been cooperating with MdS for some time, putting on workshops and accepting the help of its partners. This school's biggest problem is educational wastage in its many forms. Here, 10 teachers participated in the PAR.

#### Procedure

The project was presented to the school principals and teachers. MdS offered a preliminary time schedule, a team of the Association's partners (pedagogues, psychologists, educators, workshop teachers, observers) and a "menu" of possible tools. The five phases of PAR have been planned and implemented, albeit in a specific way, in each school:


In agreement with the teachers, all activities were observed by psychologists from the MdS Association, in the non-participant role (Ahola and Lucas, 1981). Observation is used regularly by researchers to collect data in classrooms (Kawulich, 2012). However, observing is not a natural ability, but rather the result of precise training: for this reason trained psychologists who belong to the association have been involved.

Observers were presented to teachers and students, explaining their role. They took notes during the observation. They had the task of observing freely, with particular regard to the activities and interactions that occurred in the setting: the non-verbal and verbal behaviors, and the conversations between participants.

However, the observer, whether he is aware of it or not, is a deformed and deforming mirror: some consider subjectivity as a risk to be avoided if possible because it is a source of error, while others see it as a resource, as a further component of cognition and therefore they consider subjectivity a main path to knowledge (McMahon and Farnfield, 2010). To limit the threshold of subjectivity inherent in observation, they were asked to use denotative and descriptive language, referring to precise (not generic) situations. They were also asked to include direct speech in the written notes. They were asked to insert their own comments and report them as such.

Observation reports were read out to research participants after the activities. They were then analyzed in a descriptive and

content-oriented way in order to report processes and outcomes resulting from phase 1–3 in each school. Then, the reports related to phases 1–3 were analyzed in a lexical-oriented perspective in order to be used as an input for discussion in phase 4.

In particular, the corpus composed of all reports from the three schools was subjected to textual analysis by ALCESTE software (Analyse des Lexèmes Co-occurrents dans les Énoncés d'un TExte) (Reinert, 1990, 1993). The adoption of computerized textual analysis software enables researchers to overcome several limitations that are likely to occur in the manual coding of text: the time-consuming nature of manual coding, the potential for rater bias, the reliability and accuracy of coding as a result of researcher fatigue, and the practical difficulties of coding large data sets (Illia et al., 2012). It is noted that, on the other hand, this type of analysis reduces the complexity of the meaning making: the data obtained therefore only represent some aspects of this complexity. In particular, we chose the software ALCESTE (Analyse des Lexèmes Co-occurrents dans les Énoncés d'un TExte) (Reinert, 1990, 1993): it is an instrument of statistical analysis that explores the inner organization of a text through the concurrent presence or co-occurrence of several content words. The "positioning text analysis" makes it possible to make sense of a word based on its natural context. It is based on the fundamental assumption that "since the meaning of words is learned through reading and hearing them used in particular combinations, word co-occurrence is a suitable basis for representing meaning" (Ocasio and Joseph, 2005, p. 165). Discourse is conceived as a semantic space, and a word is considered based on the position it takes in this space. The use of ALCESTE is advisable for analyzing a set of data using a nonpredefined dictionary which scans the entire text so as to include idiomatic language.

ALCESTE develops a spreadsheet with the list of words in the columns, and pieces of text or extracts in the rows (a typical extract – not necessarily corresponding to a set of sentences – is between 16 and 19 words long) and, based on this, produces a two-by-two matrix to identify classes of discourse which are very different. In technical terms this procedure is called "hierarchical downward classification of words." The software compares how words co-occur or do not co-occur in each extract and develops a classification tree that is descendant, since the whole text is divided first into two large classes of discourse, each with the most differentiated use of words. Then, for each of these two parts, the software again divides the text into two further parts, which are differentiated, certainly, but less so than the first ones. The software continues this classification until the differences among classes of discourse become too small to be significant. At this third step, the researcher does not yet interpret the results. Instead, next, ALCESTE begins co-occurrence analysis by benchmarking different parts of the text. This benchmark is developed two times, which provides a measure of stability for the comparison (stability index). Thus, here one gets a measure of stability for the descending hierarchical classification. A good measure of stability is said to have been obtained when about 70%<sup>1</sup> of the text is classified in the same way twice (Illia et al., 2012). Only in the second phase of its process does ALCESTE analyze each part of the text separately. This has the added value of managing the potential for human bias in the interpretation of the results: ALCESTE does not impose an interpretation according to which one part of the text is considered different from or similar to others. The researcher analyses quotations only once they have been identified as being either typical or atypical of a specific co-use of words. To put it another way, when the researcher analyses the quotations, s/he does so based on a report of how the language in each quotation is different from or similar to the language used in other quotations. Thus, the human coding is guided by an informative report, and human bias is controlled (Illia et al., 2012). The researcher receives a number of written or visual descriptions of the results. All of the elements listed below have been analyzed by the refinement process described above, until the meaning of each class is clear and the researcher gains a comprehensive understanding of how and why each cluster is distinct. This is the phase in which human interpretation takes place. In the classification tree or dendrogram, composed of macro-areas divided into stable clusters which correspond to lexical universes or "mental rooms" which the authors of the texts "enter" (Reinert, 1993), the distance trees visually represent the descending hierarchical classification. Each cluster is characterized by a specific vocabulary, composed of words which recur significantly more frequently inside the corpus, and by a set of elementary context units (ECUs), obtained through the analysis of punctuation and length (fundamental statistical units for the software). In each cluster, the frequency of words is only a secondary parameter, as frequency is relevant only when it reveals that there is a more frequent use of certain words in a specific part of the text, compared with another part of the text. This is indicated by the Chi<sup>2</sup> for words in a class. The value of Chi<sup>2</sup> is not absolute; it can only be computed within each class. All words listed in the report are relevant for interpretation, even considering that a high positive Chi<sup>2</sup> indicates words that co-occur the most in a cluster (Illia et al., 2012).

## RESULTS

We will describe the phases of the PAR cycle carried out in each school: planning, action, and reflection. The fourth phase (evaluation) is common to all three schools.

#### Phase 1: Planning The School in Rome

The teachers decided to form discussion groups about educational wastage that also included the students, followed by workshops and, finally, new discussion groups to discuss the workshops. During the first meeting, the teachers started by describing the negative characteristics of their pupils: unenthusiastic, passive, inept, uninterested, lazy, dull, distracted, flat. The students were also described in the following terms: they have no passion, they don't care about politics, they have no interests; they don't know how to communicate with each other; they are atomized, isolated, and divided; they don't form a group; they don't know how to communicate with teachers; they have no

<sup>1</sup>According to Reinert (1990), creator of ALCESTE, a value of 50 is sufficient.

perception of their limits and resources; they have an inaccurate perception of (the institution's) space and time; they are often compliant with the teachers, but not genuine; they do not follow the rules. It is not clear whether the characteristics the teachers discussed are subjective or context-bound. Only one teacher said: "it's weird; they shouldn't be bored at that age." Teachers also formulated some interesting theories on the learning/teaching process, along the lines of the following: there is no learning without relationships, but it's difficult here; if a student has specific learning difficulties, the family has to be involved, although they don't usually cooperate; the clever ones have to leave, or they'll waste away. The desire/goal of the teachers is to create an environment of trust and an alliance with the pupils. They listed the educational strategies employed thus far, to no avail, and the obstacles they met. Then they recounted a fragment of daily life in the classroom: "During an hour of substitution, someone in class throws a shoe in the air. The teacher asks who's to blame for this violent and irresponsible gesture. The class joins forces to keep the secret." In the follow-up discussion, the teachers said they fear the aggressiveness of their students, and for this reason they try to avoid any possible conflict. But one teacher said, "In my opinion, there is no education without conflict." Then, they discussed the chance offered by the project to become equipped with a tool for observation and reflection in order to understand why the strategies used thus far do not work.

Later on, the teachers introduced the project to their students. The first meeting with the students followed, where students described themselves as suburban teenagers: as suburban teenagers, different from city center teens. As they said, in the suburbs, you become autonomous early and you start working early; if you leave the suburbs, you're in trouble; the suburbs are freedom and the center is a cage, but it's also true that the suburbs have no rules and are out of control, while the center is protected and follows the rules. They also illustrated theories explaining what is wrong with school, according to them: they want privacy, while adults ask questions and then judge and punish them. Because of this, they hide from the adults; sometimes they pretend to be interested, but it's difficult to have a real dialogue between students and teachers, because the distance is too great. They discussed at length the fact that the teachers address the students with "tu" (informal language) while demanding to be addressed with "lei" (formal language). The teachers actively underline their institutional role. One of the teachers does so aggressively, sarcastically making fun of the student who is talking. The students' desire to not be treated like children and to have more independence and be less controlled emerged. Then came the recounting of a moment of daily school life: "the other day, two girls beat each other up in the hallway. The teachers decided to add more controlling rules applying to all students"; and "all field trips have been cancelled for lack of trust in the students, who are all considered irresponsible." During the discussion, the students said that the teachers forget what it means to be teenagers and don't give second chances, and that it would be good to have teachers who really care about the students and give second chances.

The next meeting began in the midst of many organizational hindrances. The atmosphere was chaotic. Nevertheless, the teachers managed to discuss the issues raised by the students; some teachers critiqued the lack of strictness of present-day schooling, by now "reduced to a playroom." Afterward, they started planning some experimental educational approaches, trying to answer the question: why is it that not even workshopbased pedagogy is successful with students?

Teachers then decided to come up with the following: a workshop on aggressiveness and rules of conduct during law classes; a workshop on the use of technology during English classes; and a lesson "that allows the students to take time to make mistakes and correct them" during maths classes.

#### The School in Bologna

During the first meeting, the teachers started by describing the negative characteristics of their pupils: they are difficult, suburban kids who are left out, disowned and different. They communicate through physical violence, speak a different language, and have no self-awareness, nor awareness of time and space, but are receptive, sometimes accepting the teachers' assistance, letting themselves be guided. Additionally, they immediately pointed out the school's strengths: open and democratic management, good relationships between colleagues, the school's participation in supply requests, funding, and services, new teachers being seen as a resource, and the welcoming and protective environment. The school's problems then followed: too many students drop out of school in the transition to high school, there is aggressiveness in the classroom, and some teachers are insulted. Afterward, they recounted an educational experience carried out the previous year: the teachers offered the students a workshop about Golding's novel Lord of the Flies. The goal was to make the students understand the importance of rules. Their evaluations of the experiment emerged during the discussion: what caused the failure of the workshop were language issues, infantility, and the pace of learning differing between pupils.

They also recounted an experience with the students' parents: the Women's School, aimed at teaching Italian to immigrant mothers. During the discussion, many doubts emerged over the usefulness of involving the parents in this way.

Finally, they described another educational experience: a workshop on the topic of violence against women, based on newspaper readings. During the discussion, the teachers reported to have observed the presence of widespread misogyny and of an alarming aggressive tendency in the classroom.

The teachers proposed to investigate the causes behind the academic failure and school wastage of their pupils, distributing questionnaires or collecting information post hoc. But the methodological proposal, which had initially excited the teachers, was then abandoned. Later on, the teachers decided to introduce interdisciplinary workshops: a workshop about the telegraph, as it is the ancestor of modern immediate, long-distance communication; a workshop teaching how to plan a trip to a foreign country; and a workshop teaching how to build polygons using plastic straws.

#### The School in Naples

As already mentioned, in this Lower Secondary School, teachers have been cooperating with MdS for some time, offering

workshops and accepting the help of its partners. Many times MdS has offered the school a reflective space for teachers, but opposition has been considerable. This school's manifest problem is educational wastage in its many forms.

For the "Crossing Educational Boundaries" Project, we decided to observe some already-existing workshops, characterized by interactive methods and by the presence of a teacher, an expert on the subject, and an educator: a buoyancy workshop; a workshop about food composition; and a workshop about the food pyramid.

#### Phase 2: Action

#### The School in Rome

Because of some new organizational issues, only one workshop was carried out: the one about aggression and rules of conduct, which used triggering images and created spaces of narration and discussion. The workshop was managed by an MdS pedagogue, but the teachers were present too. Some photos were projected: a picture of wolves baring their teeth but not attacking, and another of wolves playing. A debate arose among the students about the reasons behind animal and human aggression, about strategies to restrain it and control emotions, and about civilization and friendship.

#### The School in Bologna

All workshops were carried out and observed: the observer noticed that during the workshop involving learning the basics of the Morse alphabet, the students were divided into groups placed in different classrooms in order to try and communicate from a distance, and as a result many students were bored and some did not participate at all. During the workshop that involved a division into groups in order to use a computer to buy tickets, book accommodations, and plan museum visits and nights out, many conflicts erupted between single students or subgroups for the use of the computer, and some students did not participate at all. During the workshop consisting of building different polygons following the teacher's instructions, using straws of different lengths, only one small group participated, while many students remained inactive.

#### The School in Naples

The first workshop began in a situation of confusion that caused a strong emotional reaction of discouragement in the teacher: "I can't do it anymore! I wasn't trained to educate, it's not my job! I'm sick and tired of shouldering the education of other people's children!" The chaos continued: some pupils stood up and wandered around, some exited the classroom, some strangers entered, one threw pencils, another isolated himself, some provoked others. According to the observer, the teacher became cynical with some students, mocking them for their physical features or for their mistakes. Then she managed to restore calm, and demanded loudly for the approval of the MdS partners: "Here, see, I shut them up." In the second workshop, we found the same chaos, but here one of the students intervened to restore calm: "Shut up, we can't work like this!" After a while, the chaos started again: some left the class, some entered, some joked around, some threw things, some ate, some used mobile phones, some ruined the others' project. At this point, the teacher had a strong emotional reaction and started screaming in turn. The observer noted that when no adult in the classroom decides to stop the actions and talk to the students, chaos grows and the adults explode; when the educator intervenes by stopping the action and listening to the kids, calm is restored.

# Phase 3: Reflection

#### The School in Rome

During the final discussion, the teachers expressed their opinions: the workshop worked, but they don't have time to do things like that, because they have to follow the syllabus. A single teacher says: "Actually, we could, we are free to choose."

#### The School in Bologna

At the end of the workshops in this school, the teachers entrusted the reflection to their students. The observer commented that "they seem to struggle to talk about what they feel more than what they know." Finally, the students said that working in groups is at the same time more tiring and more fun, and that they lacked the time to finish the workshop because the bell rang. A young girl said that today she felt good at school, she felt like she was at home.

After many organizational issues, the teachers suspended the research activities.

#### The School in Naples

A discussion meeting is held with teachers and students, in order to reflect on the workshops' progress: the teachers decided to assign some students the task of cooperating and maintaining order, so that they are responsible. A third workshop started peacefully, then the usual chaos began. The teacher left the class often. This time, the educator took it upon herself to talk to the kids to make them reflect on what was happening and on a sense of the rules for working in a group.

A new discussion with the students followed, where they discussed their own self-representation: some kids said that they are seen as "monsters," and that it's true they are "monsters," and that is why there is an observer.

During the final discussion, the teachers suggested to pupils that they keep diaries about what happens in class, to discuss them together later, but the kids say they don't want to write negative things about their classmates. During the last group discussion, the students gave their evaluation of the workshop activities they took part in, using words like: nice, bad, tranquillity, serenity, happiness, relaxing. A girl adds: "I felt good, like at home."

In this school as well, after some organizational issues, the research activities were temporarily suspended.

#### Phase 4: Evaluation

Subsequently, MdS decided – together with the teachers – to entrust a team of external researchers with the textual analysis of the corpus composed of the observational reports. The initial results have been discussed with the teachers from all schools on the occasion of the International Convention "Crossing Educational Boundaries," held in Naples in October 2017.

The corpus – composed of 18 observational reports – was pre-processed, as it underwent disambiguation and partial lemmatization. Each text was marked by three illustrative variables: city (Roma, Bologna, Napoli), type of school (Lower or Upper Secondary), and type of activity (discussion or workshop). The analyzed characteristics of the textual corpus are illustrated in **Table 1**.

The analysis classified 511 ECUs out of the 732 detected (stability index 69.81%), divided in two macro-areas (A. Students; B. Teachers) and four stable clusters, as shown in the dendrogram of ECU classification (see **Figure 1**). For each cluster, the number and percentage of ECUs were indicated (see **Table 2**). Each cluster was assigned a thematic label on the basis of the specific vocabularies and of the most significant ECUs. Specific vocabularies and illustrative variables of clusters are illustrated in **Table 2**. 2

#### I Macro-Area

The first macro-area identified by ALCESTE contains clusters 1 and 4, regarding, respectively, the words and actions of the students.

#### **Cluster 1 – The students' discourse on marginality (146 ECUs) (28.57%)**

The specific vocabulary of Cluster 1 (see **Table 2**) contains many names of the students<sup>3</sup> who participated in the discussion, and shows their self-representation as linked to the suburb they live in, with all its contradictions. The suburb – compared with the center – is considered a place of freedom, where it is easy to make friends, but also a cage with no institutional control. In school,

2 Since the collected data was in Italian, the resulting translation proved to be particularly difficult: it was necessary to come up with a partial approximation for some of the words where there was no exact correspondence with the English form. <sup>3</sup>All names were changed for privacy.


they feel ill at ease, but this truth is hidden, because they pretend. Addressing teachers with "lei" (Italian formal language) creates too much distance in their relationship, and does not make them feel close. The associated illustrative variables show a significant correlation (see **Table 2**) with the school in Rome, where the students are older.

The content of the cluster becomes clearer by reading the specific ECUs:


In Cluster 1's most significant ECUs, some additional features of student self-representation emerge: exterior features, violent conflict between neighborhoods, and early autonomy outside the school. Students emphasize difficult communication with teachers. They emphasize this difficult communication with teachers, but also express

#### TABLE 2 | Clusters: specific vocabularies<sup>a</sup> and illustrative variables.

fpsyg-10-02308 October 17, 2019 Time: 18:23 # 9


(Continued)

#### TABLE 2 | Continued

fpsyg-10-02308 October 17, 2019 Time: 18:23 # 10


a In the translation from Italian to English it was often necessary to use more than one word to get across the original meaning. The words with a value of Chi<sup>2</sup> up to about 5 are shown. <sup>∗</sup>Non-translatable expression in English that indicates addressing the other in a formal way.

the desire to be guided by adults who are authentically interested: "it's good when people really care about us and can give us real directions, while still leaving us the freedom to choose."

The illustrative variables associated with this cluster show that the discourse is, above all, from the older students and takes place during the workshops (presumably the narrative laboratory on aggressiveness).

#### **Cluster 4 – The students' behavior between disorder and attentiveness (139 ECUs) (27.20%)**

Cluster 4's specific vocabulary (see **Table 2**) contains numerous references to the workshop activities carried out in class (see illustrative variables associated). The names of students, teachers, and MdS educators are repeatedly mentioned; there are many verbs indicating activity and movement, and some words that explicitly refer to space and activities (such as to do, to begin,

to enter, to re-exit, to go out, to come, to sit down, to get up, to arrive, to play, to bring, to throw). There are also many temporal indexes (like now, minutes, while, after, often); some words describe the atmosphere during the workshop, constantly alternating between chaos and involvement (like confusion, calm, disorder, attention, silence).

In Cluster 4's most significant ECUs, we see clearly the constant attention swings of the class:


#### II Macro-Area

The second macro-area identified by the semantic-structural analysis contains clusters 2 and 3, regarding, respectively, the words and actions of the teachers.

#### **Cluster 2 – The teachers' discourse on difficulties (182 ECUs) (35.62%)**

The specific vocabulary of Cluster 2 (see **Table 2**) contains the teachers' words, uttered during the discussion groups (see also the associated illustrative variable). The teachers discussed their professional difficulties, related to the school and to the neighborhood. In particular, they focused on the students' lack of skills, knowledge, awareness, curiosity and spontaneity, and on the need to understand and improve their relationship with them in order to better support their development and learning. Subsequently, the teachers discussed how they can participate in this research in its many possible forms, such as by creating questionnaires and working in groups; they discussed the time they would need, and the topics and experiences that might make sense for them and help them.

The content of the cluster becomes clearer by reading the specific ECUs:


In Cluster 2's most significant ECUs, we see clearly some of the teachers' ideas: in these schools, the kids have a strong sense of suburban identity, which causes some of their difficulties. There have been many tests of alternative teaching methods, aimed at creating an alliance with the students, but to no avail. Finally, there is a need for constant updates and research, especially for those who have been in the school system for a long time. Despite the discouragement, the teachers believe that change is possible.

#### **Cluster 3 – The teachers' behavior in the organization (44 ECUs) (8.61%)**

Cluster 3 is the quantitatively least vast. Its specific vocabulary (see **Table 2**) contains references to the active organization of the research: dates, hours, setting (project, meeting, workshops, etc.), dynamic verbs (to bring about), verbs in the future tense (it will be, they will be), partners to involve (observer, educator), methods (experimentation, workshop observation), criteria for the choice of the classes to involve (to define), and accurate definition of the final phase (feedback, fulfilment, reflections, etc.).

The elementary context units specific to this cluster show commitment to the organization, but also some hindrances that stop the teachers from always being present during the activities they planned:


#### Phase 5: New Planning

After the presentation of the results of this research to all the schools involved in this PAR, on the occasion of the aforementioned international convention, a small group of teachers discussed and planned, in cooperation with MdS, an intervention prototype: S.A.P.E.R.E.<sup>4</sup> (Open Space for Planning And Educational Research). These prototypes are intended to defend resilience through collegiality and self-reflection: they are, in fact, groups of school teachers, educators and psychologists in the association who meet periodically for multi-vision meetings. The multi-vision group is a reflective methodology already successfully used by MdS partners (Parrello et al., 2019b). These prototypes are currently active in some schools in the Neapolitan suburbs, where the Association has made available its economic and professional resources.

#### DISCUSSION

This research is part of the many actions undertaken by MdS in order to contain the phenomenon of educational wastage in its multiple forms.

In deprived environments, school can be an important opportunity for development and social inclusion; instead, it often fails, as shown by the available data. In order to understand the reasons behind this partial failure, MdS deems it essential to actively involve the "main characters" of school life, equipping them with the resources necessary to produce potential change.

Following the theoretical perspectives that see educational wastage mainly as the effect of inadequate relationships, MdS has long since chosen the path of "taking care of the carers," that is, to deal with the well-being of the adults responsible for educating the youth.

This project was born out of the experience gained in many Italian suburban schools: here – as it is also known from the literature – the difficulties of the students, and the high dispersion rate, make teachers feel ineffective or, worse, powerless, thus reducing their well-being at work and triggering a negative cycle damaging their educational performances.

In particular, in the school in Rome, the teachers decided to let the students talk from the start. This is an important, infrequent choice that involves students, and it allows the comparison of the discourses of adults and teenagers. Both ascribe the causes of academic failure to a difficult educational relationship and to lack of communication, but end up blaming each other for it; both complain about the strict limits of the institution. Both wish to find a way to communicate and change some of the school's rules. A single teacher – a voice from outside the box – ultimately reminds the colleagues: "we are free to choose!" In the school in Bologna, the teachers reported impeccable functioning, and focused on the different features of workshop-based pedagogy, potentially able to create a serene climate in which learning is made easier. In the school in Naples, the well-established habit of being observed for MdS projects allowed the teachers to show their own strong emotional reactions, caused by the constant oscillation of the classes between chaos and attention. The discouraged teacher said, "I wasn't trained to educate, it's not my job!" Each school therefore seems to give emphasis to one aspect: the active involvement of students as co-researchers (Fielding, 2004), attention to alternative pedagogy (Esi, 2015), and/or the importance of the teacher's emotional competence (Nouwen et al., 2016).

The semantic-structural analyses of the observational reports highlight additional important elements.

The students' discourse (cluster 1) underlines the specificity of the environment. The suburb – compared with the center – is considered a place of freedom, a place which requires early autonomy, but is also a cage: some kids say that those who tried to "trespass" ended up in danger. In school they are not at ease, but they pretend. They emphasize difficult communication with teachers. They would like less controlling and more authentically interested teachers. In workshops, their behavior is inconsistent (cluster 4): they swing continuously

T-PAR is a powerful form of professional development for teachers: it actively involves them, improving their awareness of themselves, of their emotions and actions, and of their power in that environment. However, T-PAR is not easily realized in schools: in the first place, because its multiple, openended goals would require the availability of the educational institution to accept modifications and unexpected changes. Secondly, because, in the beginning, it adds to the workload of the teachers without offering immediate advantages; and thirdly, because it can necessitate additional funding along the way, according to the paths undertaken. For these reasons, the first result to be emphasized here is therefore the realization of T-PAR in all its phases. In particular, the initial engagement phase was rather delicate: it was necessary to prevent the teachers from perceiving the research as if it were imposed by their superiors requesting their voluntary participation. The largest phase was planning in all schools, while the action phase has often encountered organizational obstacles, and the phase of reflection was always very short. But all phases – with their strengths and difficulties – were implemented in each school and documented by the observational reports.

<sup>4</sup> "Sapere" means knowledge in Italian.

from moments of attention and participation to moments of indifference and disruptive action; they are able to undertake the work, but they either do not want or cannot devote themselves to it with consistency. Perhaps even in the workshops realized in the three schools, where there is no asymmetry caused by the frontal lesson format, the issue of who controls and holds the power never disappears, and therefore ultimately carries weight. The teachers' discourse (cluster 2) underlines the specificity of the environment, too: in the suburbs, it is difficult to teach and to apply theories, because it is hard to build a good educational relationship; here, according to the teachers, young people seem to have no awareness of their limits and possibilities, and they take a defensive stance, in order to oppose or to attack the adults. The observation of teachers' behavior (cluster 3) shows how difficult it is to free themselves from the rigid and careless mechanism of the institution; organizational problems arise constantly, impeding full participation in the research. In all clusters, the lack of emotional vocabulary is striking. Teachers and students never mention emotions, while – as we have seen – the observers record multiple instances of behaviors definable as "acting out": sarcastic displays, shouting, people leaving the classroom, disrupting actions from adults and adolescents alike.

These results suggest that, as assumed, the teachers used the professional support offered by MdS and their resources to analyze the problematic situations in which they are immersed. The general objective of the project was achieved, and hypotheses have been generated regarding the causes of educational wastage, identifying two potential areas of intervention: the improvement of teacher-student communication and of emotional competence. The second objective of the project was reached by a small group of teachers from various schools who, after the evaluation phase, started to plan actions to reduce the problem. Are they particularly "resilient" teachers? Remembering that resilience involves the capacity of an individual teacher to harness personal and contextual resources to navigate through educational challenges, we wonder if the project has supported their resilience and why it has failed to involve all the other teachers in the final phase (phase 5). In the context of the teaching profession, resilience may be conceptualized as a capacity, a process and also as an outcome: a dynamic process whereby characteristics of individual teachers and of their personal and professional contexts interact over time as teachers use specific strategies (Beltman, 2015). In particular, time and its management are fundamental resources: the most common professional challenge is lack of time due to heavy workloads and non-teaching duties such as paperwork or meetings (Castro et al., 2009; Beltman et al., 2011). The teachers denounced the "dual level" of the institutional mandate: the organization is separated from the exigencies of the educational relationship; and its scheduling and regulations fail to take into account the specificities of individuals and contexts, and are more and more oriented toward formality. Ultimately, school tends to hinder change. But, according to Enriquez (1972), the institution can become the place to cling onto in order to defend oneself from multiple anxieties.

"It follows that when the defensive component prevails, norms, role, schedules, technologies, etc., are used not as tools of work and development, but with the aim of never changing, of controlling everything and expecting everything, so to prevent the unexpected, or change, from creating anxiety. (. . . ) Within organizations, everything is planned and thought-out (and happens consciously) as if the members and their relationships were moved solely by ideas and rational moments (. . . ). However, in school there are also emotional, libidinal, affective factors (. . . ). If the effect of these feelings over whether we support or obstruct each other in our cooperation and partnership with colleagues and pupils are not recognized, not only do we lose an occasion to work creatively, that is, to learn, but above all we contribute to keep the Institution unchangeable, namely we contribute to represent its defensive feature" (Blandino and Granieri, 1995, 95–96).

Ultimately, school is a place of paradoxes and contradictions. In order to avoid sacrificing any needs of the people and of the institution, pupils and teachers become masters of balancing acts: oscillating between participation and absence, cooperation and insubordination. However, this incessant oscillation can be seen in a positive perspective, as the attempt of both adults and young people to live within the school authentically, voicing their ambivalence. Of course, it is an attempt that requires energy and does not change the institution.

In conclusion, the results of this phase of the research contribute to the construction of a more precise picture of the situation of suburban schools, where the rates of educational wastage are high: the teachers who participated in our T-PAR are prepared on a theoretical level, are sensitive to the many problems of their pupils, and feel responsible as educators, but are also very distressed by the formal requests of the educational institution. In these conditions, they struggle to put their theoretical preparation to use, and fail to gain the trust of these young people and to give meaning to their job. What they get in return is a state of dissatisfaction that threatens their resilience.

Caring for horizontal and vertical relationships, the construction of an atmosphere of community within the school, and systematic reflection that gives space to the emotional and unconscious world should all be at the center of any project against educational wastage.

#### LIMITS AND PROSPECTS

The specific limitations of this study involve the use of a single tool monitoring the intervention, that is observation, and the use of a single type of analysis. It would also be useful to complete this study by adding different analyses of the other schools' research material.

In any case, these initial results propose school policies not limited to broadening the curriculum of suburban schools at risk of educational wastage: there is, additionally, the need to use further resources to improve the educational environment and to allow the teachers to systematically exercise professional self-reflection. To do so, it would be very useful to monitor the phase of research currently underway, which is the realization of intervention prototypes based on the use of reflective tools, in turn based on the model of the multi-vision group of MdS.

### DATA AVAILABILITY STATEMENT

The datasets generated for this study are available upon request from the corresponding author.

# ETHICS STATEMENT

fpsyg-10-02308 October 17, 2019 Time: 18:23 # 14

This study was carried out in accordance with the recommendations of the Declaration of Helsinki, and was approved by the Ethical Committee for Psychological Research (CERP) of the Department of Humanities of the University of Naples Federico II, prot. number 24/2019. The minors involved were only taking part in school activities that were included by the schools in their learning programs; the schools obtained an oral informed consent from the families at the beginning of the year. The CERP did not request further parental consent.

#### REFERENCES


# AUTHOR CONTRIBUTIONS

SP and CM designed the study and supervised the research. SP, II, and FC analyzed the data and wrote up the first draft of the manuscript. All authors interpreted the results, and approved the final version of the manuscript.

# FUNDING

This work was partially funded by MIUR, the Italian Ministry of Education and Research.

# ACKNOWLEDGMENTS

We wish to thank Dr. Chiara Cigliano for the English translation of the article.




**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Parrello, Iorio, Carillo and Moreno. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

Trans-cultural Adaptation and Validation of the "Teacher Job Satisfaction Scale" in Arabic Language Among Sports and Physical Education Teachers ("Teacher of Physical Education Job Satisfaction Inventory"—TPEJSI): Insights for Sports, Educational, and Occupational Psychology

Nasr Chalghaf 1,2,3, Noomen Guelmami 1,4,5, Tania Simona Re<sup>6</sup> , Juan José Maldonado Briegas <sup>6</sup> , Sergio Garbarino<sup>7</sup> , Fairouz Azaiez 1,2,3 and Nicola L. Bragazzi 1,6,7 \*

*<sup>1</sup> Department of Health Sciences (DISSAL), Postgraduate School of Public Health, University of Genoa, Genoa, Italy, <sup>2</sup> Group for the Study of Development and Social Environment (GEDES), Faculty of Human and Social Science of Tunis, Tunis, Tunisia, <sup>3</sup> Higher Institute of Sport and Physical Education of Sfax, University of Sfax, Sfax, Tunisia, <sup>4</sup> Higher Institute of Sport and Physical Education of Kef, University of Jendouba, Jendouba, Tunisia, <sup>5</sup> Research Unit, Sportive Performance and Physical Rehabilitation, High Institute of Sports and Physical Education of Kef, University of Jendouba, Jendouba, Tunisia, <sup>6</sup> Department of Psychology and Sociology of Education, University of Extremadura, Badajoz, Spain, <sup>7</sup> Department of Neuroscience, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health (DINOGMI), University of Genoa, Genoa, Italy*

Background: Job satisfaction is largely associated with organizational aspects, including improved working environments, worker's well-being and more effective performance. There are many definitions regarding job satisfaction in the existing scholarly literature: it can be expressed as a positive emotional state, a positive impact of job-related experiences on individuals, and employees' perceptions regarding their jobs.

Aims: No reliable scales in Arabic language to assess job satisfaction in the sports and physical education field exist.This study aimed to trans-culturally adapt and validate the Pepe's "Teacher Job Satisfaction Scale" 9 items (TJSS-9), creating the "Teacher of Physical Education Job Satisfaction Inventory" (TPEJSI) in Arabic language. This scale was administered to a Tunisian population of sports and physical education teachers and analyzed according to the Pepe's theoretical model. More in detail, this investigation systematically tested its factor structure, in terms of internal consistency/reliability, predictive validity, sensitivity and convergent validity.

Methods: A total of 417 Tunisian teachers of sports and physical education participated voluntarily in this study. The sample comprised of 258 males and of 159 females. More in detail, 189 were teachers teaching in primary schools of physical education, 105 teaching

#### Edited by:

*Wong Yau Ho Paul, Tung Wah College, Hong Kong*

#### Reviewed by:

*Laura Dal Corso, University of Padova, Italy Lea Ferrari, University of Padova, Italy*

#### \*Correspondence:

*Nicola L. Bragazzi robertobragazzi@gmail.com*

#### Specialty section:

*This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology*

Received: *01 April 2019* Accepted: *17 September 2019* Published: *22 October 2019*

#### Citation:

*Chalghaf N, Guelmami N, Re TS, Maldonado Briegas JJ, Garbarino S, Azaiez F and Bragazzi NL (2019) Trans-cultural Adaptation and Validation of the "Teacher Job Satisfaction Scale" in Arabic Language Among Sports and Physical Education Teachers ("Teacher of Physical Education Job Satisfaction Inventory"—TPEJSI): Insights for Sports, Educational, and Occupational Psychology. Front. Psychol. 10:2234. doi: 10.3389/fpsyg.2019.02234*

**158**

in secondary schools, and 123 were university teachers. Both exploratory (EFA) and confirmatory (CFA) factor analyses were performed on random-split halves of the sample.

Results: The three-dimensional alpha coefficients of the TPEJSI were all >0.80: for satisfaction with colleagues, alpha was 0.865; for satisfaction with parents, alpha was 0.856 and for satisfaction with students alpha was 0.860. The CFA fit indices were satisfactory.

Conclusions: Given the good EFA factor loadings, the CFA fit indices, the correlation matrix, the sensitivity analysis, the convergent validity and the excellent internal consistency, it can be concluded that the TPEJSI is a good psychometric tool that can be used to quantitatively assess the job satisfaction level among teachers of sports and physical education in the Arabic-speaking world.

Keywords: trans-cultural adaptation and validation of a questionnaire, Arabic language, sports psychology, occupational psychology, job satisfaction, teachers

# BACKGROUND

Job satisfaction is a multi-disciplinary concept that can be studied utilizing a variety of theoretical approaches and frameworks: it can be addressed from a psychological or sociological standpoint, as well as from the perspective of educational sciences and management (Hongying, 2007). As such, there exist different definitions of job satisfaction. According to Spector (1997), it represents a variable that measures the person's attitude to work, including the different facets of the job. Job satisfaction can be, as well, viewed as the result of the employee's interaction with and perceptions of his/her workplace and its surrounding environment (Carriere and Bourque, 2009; Rehman et al., 2013). Further, other researchers have defined job satisfaction as a positive emotional state (Inandi et al., 2013), a positive impact of the job-related experiences on individuals (Filiz, 2014), and employees' perceptions regarding their jobs and occupations (Altinkurt and Yilmaz, 2014).

In the literature, job satisfaction has been found to be inversely associated with absenteeism (Hanebuth, 2008), leaving the profession/retiring (MacIntosh and Doherty, 2010; Tschopp et al., 2014) and psychological stress (Moen et al., 2013).

By scanning the literature, it has been seen that there are very few studies investigating Arabic-speaking teachers' opinions about their job. In occupational psychology, having a reliable tool to measure the job satisfaction of employees is of crucial importance.

For this reason, in order to quantitatively evaluate job satisfaction among teachers, Pepe (2011) has developed the 9-item "Teachers Job Satisfaction Scale" (TJSS-9) as a psychometrically sound tool to be used in the academic field. More in detail, the TJSS-9 has three dimensions: namely, (i) the satisfaction of the colleagues (3 items), (ii) the satisfaction of the parents (3 items) and (iii) the satisfaction with the behaviors of the students (3 items). The current version of the instrument (9 items) was developed by Pepe (2011) from an original set of 35 elements covering six different dimensions: (i) satisfaction with all colleagues, (ii) satisfaction of colleagues, (iii) management satisfaction, (iv) satisfaction of parents, (v) satisfaction with students' behaviors, and (vi) responsibility.

### AIMS

To the best of our knowledge and by scanning the existing scholarly literature, there is no psychometrically validated tool in Arabic language to assess the job satisfaction among sports and physical education teachers. Therefore, the major objective of the present study was to trans-culturally adapt and validate the TJSS-9 according to the Pepe's 3-dimensional theoretical model for sports and physical education teachers in Tunisia and to test its factor structure, in terms of internal consistency/reliability, predictive validity, and sensitivity. This study should be considered as a pilot, exploratory investigation, the results of which can pave the way for the development of further ad hoc instruments, specifically intended for sports and physical education teachers, correlating the job satisfaction level with other variables and constructs of interest.

# MATERIALS AND METHODS

#### Participants

A total of 417 sports and physical education teachers answered the questionnaire (159 and 258 males and females, respectively). Participants were divided according to their age into four categories: T1 (age < 35; n = 113), T2 (35 ≤ age < 40; n = 125), T3 (40 ≤ age < 50; n = 100), and T4 (age ≥ 50; n = 79). Depending on their degrees, teachers are classified as physical education teachers (having completed a 2 year post-baccalaureate degree and having a 2-year university degree, n = 189), specialized secondary school teachers (having completed a 4-year training cycle after the baccalaureate and having a master degree, n = 105), and high sport institute teachers (n = 123). Teaching volumes varied according (22, 18, and 12 h per week, respectively).

#### Ethical Approval

The study protocol of the present investigation received ethical clearance from the UNESCO Chair "Health Anthropology Biosphere and Healing Systems," University of Genoa, Genoa (Italy), the Higher Institute of Sport and Physical Education of Sfax, Sfax (Tunisia), and the Higher Institute of Sport and Physical Education of Kef, Kef (Tunisia).

The project was approved by the Ethical Committee of the University of Sfax, Sfax, Tunisia.

All participants to the present study provided written, informed consent. Teachers were extensively informed about the purposes and procedures of the study, and were advised that the results would be made available to them upon completion of the study only in aggregate form, with no possibility to trace back to the single teacher's scores, thus ensuring anonymity and preserving the privacy of each participant.

The present investigation was carried out in accordance with the ethical principles of the 1964 Helsinki declaration and its subsequent amendments.

# Procedure

Teachers who agreed to participate in the study were instructed how to proceed and complete the survey procedures required by the present study. Following the agreement of the primary and secondary school principals, copies of the transculturally adapted TJSS-9 were distributed to teachers at their work sites in off-peak hours over a 2-month period, ensuring a proper duration (∼30 min) in order to answer the questionnaire thoroughly.

# Statistical Analysis

#### Descriptive Analysis

Before commencing any statistical analysis (data handling, pre-processing, and analysis), data were visually inspected for potential outliers. Normality of data distribution was checked using the Kolmogorov-Smirnov test. Questionnaire scores were also checked for skewness and kurtosis, computing the Mardia's multivariate statistics.

For descriptive purposes, we calculated the mean and standard deviation (SD) of each score. The alpha level was set a priori at P ≤ 0.05.

#### Internal Consistency/Reliability

The internal consistency of the instrument was examined computing the Cronbach's alpha coefficient for all the 3 dimensions of the scale. More in detail, in order to properly interpret the alpha coefficient, the following rule of thumb was used (Nunnally, 1978): the coefficient was considered excellent if the estimate was >0.90, whereas it was deemed good in the range 0.80–0.90, acceptable in the range 0.70–0.80, questionable or adequate in the range 0.60–0.70, poor in the range 0.50–0.60 and unacceptable if <0.50.

#### Inferential Statistics—Sensitivity Analysis

The sensitivity of the instrument was tested by performing a multivariate analysis of variance (ANOVA), examining the impact of teachers' grade, gender, age, and their interaction effects on the TJSS-9 3 dimensions and total scores.

#### Exploratory Factor Analysis (EFA)

The factor structure was initially tested by carrying out an exploratory factor analysis (EFA) with the principal component analysis (PCA) and a varimax rotation with Kaiser Normalization. More in detail, varimax rotation was preferred to other kinds of rotation in that this approach, differently from the others, enables to minimize factor complexity while, at the same time, maximizing the variance of factor loadings (Tabachnick and Fidell, 2013).

Before proceeding with the EFA, the Kaiser-Meyer-Olkin (KMO) measure was computed in order to investigate the sampling adequacy. Ideal values of the KMO should be >0.60. Once verified the sampling adequacy, an EFA iterative strategy was implemented in the present study. More in detail, different PCA runs were conducted. First, an exploratory/preliminary PCA was conducted on the 9 items of the questionnaire without any rotation, in order: (i) to check if PCA could be judged an appropriate technique for the correlation matrix by assessing whether the correlations among items were satisfactory (that is to say, reporting values >0.30), and (ii) to control for the factorability of the correlation matrix computing the Bartlett's test of sphericity. In cases of statistical significance, this test enables to reject the null hypothesis that the correlations in the correlation matrix are zero and the matrix is an identity matrix.

The likely number of factors was found by: (i) calculating the number of factors with eigenvalues >1 (Field, 2009; Tabachnick and Fidell, 2013), and (ii) visually inspecting the Cattell's screeplot. After checking the factor loadings, items were deleted in cases of unsatisfactory loading (that is to say, values <0.45). Moreover, items were not retained and suppressed if their factor loading conflicted with a sound theoretical explanation (Field, 2009; Tabachnick and Fidell, 2013).

Different PCAs with varimax rotation runs were, therefore, carried out in an iterative way, as previously explained, until a satisfactory, clearly interpretable solution was finally obtained. Cases of cross-loading were interpreted according to salience and explained variance, with theoretical considerations also being taken into account (Field, 2009; Tabachnick and Fidell, 2013).

#### Confirmatory Factor Analysis (CFA)

Then the model was tested by carrying out a confirmatory factor analysis (CFA). As suggested and recommended by many scholars (Hu and Bentler, 1995; Schumacker and Lomax, 2004), a wide range of fit indices was calculated and reported, namely: (i) discrepancy indices (including the chi-squared and the Steiger-Lind's Root Mean Square Error of Approximation or RMSEA), (ii) tests comparing the target model with the null model (like the Tucker-Lewis' Index or TLI, the Bentler's Comparative Fit Index or CFI, and the James-Mulaik-Brett's Parsimony Goodness-of-Fit Index or PGFI), and (iii) information theory goodnessof-fit measures (such as the Joreskog's goodness-of-fit index or GFI).

Regarding the cut-off and threshold values for discrepancy indices, the p-value associated with the chi-squared test should exceed 0.05. As far as the RMSEA is concerned, values higher than 0.10 indicate poor fitting models (Steiger, 2000). Concerning the cut-off and threshold values for tests that compare the target model with the null model, TLI should exceed 0.90 according to Byrne (1994) or 0.95 according to Schumacker and Lomax (2004) as well as according to Hu and Bentler (1995). PGFI is derived from NFI, correcting and compensating for model parsimony. CFI should exceed 0.95 (Bentler, 1990; Hu and Bentler, 1999) or 0.90 according to other scholars. Finally, regarding the cutoff and threshold values for information theory goodness-of-fit measures, GFI should be higher than 0.90 (Byrne, 1994).

# Convergent Validity

Convergent validity was measured by computing the Average Variance Extracted (AVE). More in detail, AVE measures the amount of variance that can be explained by the construct under study in relation to the amount of variance due to measurement error. Ideal values of AVE should be >0.5.

# Reliability

Reliability was assessed by calculating the composite reliability values, which ideally should be >0.7.

# Statistical Software

All statistical analyses were carried out using the commercial software "Statistical Package for the Social Sciences" (IBM SPSS software for Windows, version 21.0, IBM Corp., Armonk, NY, USA; released 2012) whereas the CFA was performed by utilizing the commercial software "Analysis of a moment structures" (Amos software for Windows, version 21.0, IBM, SPSS, Chicago, USA) (Arbuckle, 2012a,b).

EFA was conducted on a random-split half of the sample, whereas CFA was run on the other split half of the sample.

For all statistical analyses, figures with p-value less than 0.05 were considered statistically significant.

# RESULTS

# Univariate and Multivariate Normality

Data were normally distributed in terms of skewness and kurtosis, whereas the multivariate Mardia test revealed violation of normality of distribution.

#### Adaptation of the Psychometric Instrument: The TPEJSI

The ad hoc devised psychometric instrument is made up of 9 items (3 items for each dimension), and the scores of the dimensions are obtained by averaging the items scores. The answers are coded on a 5-point Likert scale (see **Table 1**).

#### The Sensitivity Analysis

To test the sensitivity univariate analyses were conducted by gender, age, and grade for each dimension of the instrument. The descriptive statistics for the three dimensions are shown in **Table 2** and the comparative statistics of the ANOVAs are shown in **Table 3**.

TABLE 1 | Items of the *ad hoc* devised psychometric tool to quantitatively assess the job satisfaction level among the teachers of sports and physical education in the Arabic-speaking world, the "Teacher Physical Education Job Satisfaction Inventory" (TPEJSI).

# Teacher Physical Education Job Satisfaction Inventory (TPEJSI)


The analysis of the variance showed significant differences in the first dimension by age and grade. Also, a very significant difference by grade was found in the third dimension. However, no interactions effect between gender, age, and grade was shown for all the three dimensions.

#### Exploratory Factor Analysis

The KMO and Bartlett sphericity test provided a very significant chi-squared = 1773.817 value at p < 0.001, with a total explained variance of 78.41%. EFA made it possible to extract three factors that explained, respectively up of 26.31, 26.16, and 25.93% of the total variance. **Table 4** shows the matrix of components after rotation.

TABLE 2 | Descriptive statistics reporting the scores of the "Teacher of Physical Education Burnout Inventory" (TPEJSI) for each dimension found performing the exploratory factor analysis (EFA). Scores are broken down by teachers' grade, gender, and age group.


*CS, colleague's satisfaction; PS, parent's satisfaction; SSB, satisfaction with student's behavior; SD, standard deviation; TPS, teacher of primary school; TSS, teacher of secondary school; UT, university teacher.*

TABLE 3 | Univariate analyses for the three dimensions of the "Teacher of Physical Education Job Satisfaction Inventory" (TPEJSI).


\**Statistically significant at p* < *0.05;* \*\**Statistically significant at p* < *0.01.*

*CS, colleague's satisfaction; PS, parent's satisfaction; SSB, satisfaction with student's behavior.*

#### Internal Consistency

The alpha coefficients of the three factors of the TPEJSI were all >0.80: for the dimension of satisfaction of the colleagues, alpha was 0.865; for the dimension of satisfaction of the parents, TABLE 4 | Factor loadings for the 3-factor solution of the "Teacher of Physical Education Burnout Inventory" (TPEJSI) questionnaire.


*Exploratory factor analysis was performed with varimax rotation and Kaiser Normalization. CS, colleague's satisfaction; PS, parent's satisfaction; SSB, satisfaction with student's behavior.*

TABLE 5 | Fit indices obtained from the confirmatory factor analysis (CFA) of the "Teacher of Physical Education Job Satisfaction Inventory" (TPEJSI).


*CFI, Comparative Fit Index; GFI, goodness-of-fit index; PGFI, Parsimony Goodness-of-Fit Index; RMSEA, Root Mean Square Error of Approximation; TLI, Tucker-Lewis' Index.*

alpha was 0.856 and, finally, the dimension of satisfaction with behaviors of the students, alpha was 0.860.

#### Confirmatory Analysis of the Structure of the TPEJSI

We performed a confirmatory factor analysis (CFA) for the measuring instrument using the robust method of "Maximum Likelihood Estimate."

The results of the CFA indices showed a factorial structure consistent with the theoretical model tested for the developed version of the instrument (see **Table 5**).

The robustness of an item is given by a high load factor. Comrey and Lee (2013) suggests that a factorial weight >0.71 is considered excellent, >0.63 is considered very good, >0.55 is considered acceptable and <0.45 is considered poor.

In this study, CFA of the 9 items of TPEJSI showed excellent factorial weights (see **Figure 1**).

#### Convergent Validity of the Instrument

For the three dimensions, AVE values were 0.79, 0.78, and 0.78, respectively, indicating a good convergent validity of the instrument.

#### Reliability of the Instrument

For the three dimensions, composite reliability coefficients were 0.92, 0.91, and 0.91, respectively, indicating a good reliability of the instrument.

#### DISCUSSION

The objective of the present study was to trans-culturally adapt, validate, and test the factor structure, internal

consistency/reliability, predictive validity, sensitivity, and convergent validity of an ad hoc job satisfaction measurement scale for the Arabic-speaking world across sports and physical education teachers, devised according to the Pepe's 3-dimensional theoretical model. Once adapted, the 9-item tool was validated in a convenience sample of sports and physical education teachers in Tunisia, using both EFA and CFA. EFA factor loadings were good and the CFA fit indices satisfactory. The internal consistency/reliability of the 3 dimensions was found to be excellent.

To the best of our knowledge, no study has attempted so far to validate an adapted version of the TJSS-9 on a specific population of teachers in the Arabic-speaking world. The recent work of Pepe et al. (2017) has tested the invariance of the TJSS-9 in six countries: the Netherlands, the United States, Russia, China, Italy, and Palestine for a total of 2,819 active teachers. CFA and multi-group invariance tests were applied. TJSS-9 displayed robust psychometric properties and did not differ significantly between groups in terms of measurement invariance.

In agreement with our results, Pan et al. (2015) tested through a linear hierarchical regression model the relationship of job satisfaction with several demographic and work-related variables, such as gender, age, marital status, educational level, and job position. Results showed that only age and marital status were significant independent predictors of job satisfaction among Chinese university teachers. Furthermore, the effect of gender on the job satisfaction of British academics has been studied by Oshagbemi (2000), reporting that gender does not directly affect the job satisfaction of university teachers. However, the effect of gender interaction and grade was statistically significant (at p < 0.05). Overall, university women with higher grades tended to have higher job satisfaction scores. In another study by Wang and Lee (2009) conducted among Chinese university teachers, female teachers had slightly higher levels of satisfaction than men. Furthermore, teachers with the highest rank had the highest job satisfaction. Also, other scholars (De Nobile and McCormick, 2008; Demirta¸s, 2010; Andersen, 2011; Magee, 2013) showed that gender can be either directly or indirectly associated with job satisfaction. While some investigations found that female teachers tended to exhibit a greater job satisfaction than their maile counterparts, other studies reported that gender was found to have a low level impact on teachers' job satisfaction. A recently published meta-analytical study conducted by Aydin et al. (2012) found an overall effect size of −0.02, in favor of male teachers.

Other studies carried out in Europe showed that female teachers were more satisfied in their job even though they were disadvantaged in terms of opportunities such us expectations about income, recruitment, and career advancement (Klassen and Chiu, 2010; Aydin et al., 2012; Saiti and Papadopoulos, 2015). Kaur and Sidana (2011) showed that the level of job satisfaction of male teachers was higher than their female counterparts.

Concerning the potential impact of the grade, many studies showed that job satisfaction levels presented significant differences between teachers teaching in primary, secondary or high schools. In this context, Demirta¸s (2010) reported that the job satisfaction of teachers teaching in elementary schools was higher than the level of teachers teaching in secondary schools or in universities. Mhozya (2007) explored job satisfaction of primary school teachers in Botswana and found that a significant number of teachers were not satisfied with their salary with respect to the workload. Also, Indhumathi (2011) conducted a study among teachers of a secondary school, assessing the relationship between job satisfaction and performance. Author was able to prove, on the one hand, a significant association between job satisfaction and performance, and, on the other hand, significant differences in terms of job satisfaction across teachers based on their grade. On the contrary, Yazici and Altun (2013) could not find apparent differences among lecturers in universities in Turkey.

Furthermore, a considerable number of international studies have highlighted links between teachers' job satisfaction, workload and other related variables. For instance, Collie et al. (2012) reported that teacher job satisfaction was directly related to the perceived workload and the sense of effectiveness in teaching. In another study by Hoigaard et al. (2012), authors were able to replicate such findings, showed that teacher job satisfaction was positively related to teacher effectiveness and professional engagement.

However, the present study is not without limitations. The Arabic-speaking world is rather vast, heterogeneous and culturally different, therefore limiting the investigation to Tunisian subjects could influence the general extensibility of our results. As such, future studies from other Arabic-speaking

#### REFERENCES

Altinkurt, Y., and Yilmaz, K. (2014). Relationship between occupational professionalism of teachers and their job satisfaction. Sakarya Univ. J. Educ. 4, 57–71. doi: 10.19126/suje.46033

countries are warranted to replicate our findings in a more statistically robust way.

# CONCLUSION

Our study aimed to trans-culturally adapt, validate and test the factor structure, internal consistency/reliability, predictive validity, sensitivity and convergent validity of a job satisfaction measurement inventory across Arabicspeaking sports and physical education teachers. Given the good EFA factor loadings, the CFA fit indices, the correlation matrix, the sensitivity analysis and the excellent internal consistency, it can be concluded that the TPEJSI is a good psychometric tool that can be used to quantitatively assess the job satisfaction level across teachers of physical education in the Arabic-speaking world. However, given the abovementioned shortcomings, future studies in the field are urgently needed, also exploring the relationship of teacher job satisfaction with other psychological variables and constructs of interest.

#### DATA AVAILABILITY STATEMENT

The raw data supporting the conclusions of this manuscript will be made available by the authors, without undue reservation, to any qualified researcher.

# ETHICS STATEMENT

The study protocol of the present investigation received ethical clearance from the UNESCO Chair Health Anthropology Biosphere and Healing Systems, University of Genoa, Genoa (Italy), the Higher Institute of Sport and Physical Education of Sfax, Sfax (Tunisia), and the Higher Institute of Sport and Physical Education of Kef, Kef (Tunisia). All participants to the present study provided written, informed consent. Teachers were extensively informed about the purposes and procedure of the study, and were advised that the results would be made available to them upon completion of the study only in aggregate form, with no possibility to trace back to the single teacher's scores, thus ensuring anonymity and preserving the privacy of each participant.

#### AUTHOR CONTRIBUTIONS

NC, SG, FA, and NB conceived the experiment. NG and NB collected and analyzed data. NC, NG, SG, FA, and NB drafted the manuscript. NC, NG, SG, TR, JM, FA, and NB critically reviewed and approved the final version of the manuscript.

Andersen, L. B. (2011). "Teacher diversity: do male and female teachers have different self-efficacy and job satisfaction?" in 33rd EGPA Conference in Bucharest 7-10 September (Bucharest).

Arbuckle, J. L. (2012a). IBM SPSS Amos 21. Chicago, IL: Amos Development Corporation.


**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Chalghaf, Guelmami, Re, Maldonado Briegas, Garbarino, Azaiez and Bragazzi. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Teachers Between Job Satisfaction and Burnout Syndrome: What Makes Difference in Czech Elementary Schools

Irena Smetackova\*, Ida Viktorova, Veronika Pavlas Martanova, Anna Pachova, Veronika Francova and Stanislav Stech

Department of Psychology, Faculty of Education, Charles University, Prague, Czechia

As has been shown by several studies, teaching is a highly stressful occupation (Johnson et al., 2005), and most teachers experience work stress. Long-term stress decreases job satisfaction and can result in chronic exhaustion which can develop into burnout syndrome. Implications of burnout syndrome are strongly negative both for the personal and professional life of teachers. As burnout syndrome puts teachers' well-being, quality of the teaching process and relationships with students at risk, it is important to seek ways to avoid teachersÉ' burnout. Many studies have confirmed the protective role that coping strategies play in managing stressful situations, teacher's selfefficacy or social support. In previous studies, a negative connection was found between burnout syndrome and job satisfaction. Job satisfaction is a crucial element in the definition of wellbeing. We find wellbeing rather not as a resource for burnout prevention but as an opposite state to burnout syndrome. The paper presents a quantitative survey on burnout syndrome and related variables among Czech elementary school teachers (n = 2,394). According to burnout score, two subgroups were selected – teachers with no burnout manifestations and teachers with developing/developed burnout syndrome. Through the statistical analysis, these two groups were compared in job satisfaction, self-efficacy, coping strategies, and social support. The study shows significant differences between the two groups of teachers in all examined variables. The strongest impact on burnout syndrome was found for negative coping.

Keywords: burnout syndrome, teachers, self-efficacy, social support, job satisfaction

# INTRODUCTION

As has been shown by several studies, teaching is a highly stressful occupation (Johnson et al., 2005). Most teachers experience work stress from stable distinctive conditions of teacher occupation and from specific demands related to technology and authority in the new era (Chan and Hui, 1995; Caprara et al., 2006). Stress comes when a person does not manage the task ahead. S/he has to put a lot of effort into mastering it. At the same time, s/he must also process the negative emotions they experiences due to failure (Lazarus, 1993; Hobfoll, 2002). Both for the first and second reasons, exposure to long-term stress leads to exhaustion. Long-term stress decreases job satisfaction and

#### Edited by:

Simona De Stasio, Libera Università Maria SS. Assunta, Italy

#### Reviewed by:

Marcello Nonnis, University of Cagliari, Italy Alessandro De Carlo, Giustino Fortunato University, Italy

\*Correspondence: Irena Smetackova irena.smetackova@pedf.cuni.cz

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 09 July 2019 Accepted: 24 September 2019 Published: 29 October 2019

#### Citation:

Smetackova I, Viktorova I, Pavlas Martanova V, Pachova A, Francova V and Stech S (2019) Teachers Between Job Satisfaction and Burnout Syndrome: What Makes Difference in Czech Elementary Schools. Front. Psychol. 10:2287. doi: 10.3389/fpsyg.2019.02287

**166**

can result in chronic exhaustion which can develop into burnout syndrome (Leiter et al., 2014), especially in occupations with incessant human interactions and a high level of responsibility for others (Brouwers and Tomic, 2000; Johnson et al., 2005).

Burnout is included in the 11th Revision of the International Classification of Diseases (ICD-11) as an occupational phenomenon. WHO defines burnout syndrome as "resulting from chronic workplace stress that has not been successfully managed. It is characterized by three dimensions: (1) feelings of energy depletion or exhaustion; (2) increased the mental distance from one's job, or feelings of negativism or cynicism related to one's job; and (3) reduced professional efficacy. Burnout refers specifically to phenomena in the occupational context and should not be applied to describe experiences in other areas of life" (ICD-11).

Implications of burnout syndrome are strongly negative both for the personal and professional life of the teacher. Many studies confirm that teachers with high level of burnout syndrome are not able to establish positive relationships with students, to understand students' needs, to stay in contact with trends in their field and to provide pedagogically effective lessons (Greenglass et al., 1996; Yong and Yue, 2007; Shen et al., 2015). Other studies show that teachers with a high level of burnout syndrome suffer from somatic problems such as back-pains or headache, from low self-esteem, a lack of meaningful-life orientation, interpersonal conflicts, and low social support (Hultell et al., 2013; Kidger et al., 2016). In many cases, burnout syndrome results in depression (Dilekmen and Erdem, 2013). It can result in a high rate of absences, dropouts and early retirements of burnout teachers, which increases costs (Brackett et al., 2010). As burnout syndrome puts teachers' well-being, quality of the teaching process and relationships with students at risk, it is important to seek ways to avoid teachersÉ' burnout. Loonstra et al. (2009) argue that teacher burnout is a barrier to a higher quality of the educational system and thus it is necessary to pay scientific and political attention to this.

The burnout syndrome can be considered as a diagnostic category for medical reasons, as an investigatory category for psychological or medical research or as a real-life category for an individual description of subjective state. Despite some theoretical problems referring to burnout syndrome as a diagnostic and investigatory category (Kaschka et al., 2011; Bianchi et al., 2015), we find this term very useful. In our experience and in some studies (Pyhältö et al., 2011), teachers often described their professional experience and feelings with using expressions such as lack of flame or no spark or out of battery, etc. Such expressions refer to burnout. Thus, it means that burnout syndrome represents an in vivo code. Conversely, teachers with a high level of energy and enthusiasm self-report being satisfied in their profession and even in their life (Skaalvik and Skaalvik, 2007, 2014). They use the following statements: I am still happy with students; it makes me feel self-realized; I find contentment in teaching, etc.

Based on our psychological experience and on literature (Van Horn et al., 2004; Pyhältö et al., 2011; Skaalvik and Skaalvik, 2014), we conclude that burnout syndrome stands in the opposition to the job satisfaction. However, we can replace job satisfaction with life satisfaction which is used in definitions of well-being (Diener and Suh, 1997). Diener and Suh (1997)state that "well-being consists of three interrelated components: life satisfaction, pleasant affect, and unpleasant affect. Affect refers to pleasant and unpleasant moods and emotions, whereas life satisfaction refers to a cognitive sense of satisfaction with life" (Diener and Suh, 1997, p. 200). Besides of life satisfaction, two other dimensions of well-being: the ability to fulfill goals and happiness (Pollard and Lee, 2003). Dodge et al. (2012) put in the center of well-being a state of equilibrium between resources and challenges in psychological, social and physical ways.

The theoretical background for our study covers not only the conceptualization of well-being by Dodge et al. (2012) but also the conservation of resources theory by Hobfoll (2002). This theory assumes that individuals seek to obtain, retain, and protect resources. Resources are socio-culturally influenced and consist not only of personality traits and social relationships but also of materials or life conditions (Hobfoll, 2002). Despite gaining resources that brings well-being, losing resources brings stress, including burnout (Grandey and Cropanzano, 1999). Based on the conservation of resources theory, the central motivation for individuals refers to the acquisition and facilitation of resources which is heightened especially in moments when resource loss occurs or are threatened. "Resources are seen as the essential elements of people's stress resistance armamentarium, loss of resources tends to lead to resource loss cycles that have increased strength and speed." (Hobfoll, 2002, p. 312). Individuals with a high level of resources are less likely to encounter stressful situations and thus their well-being is not negatively affected. The well-being of teachers is a widely studied field. Most studies find lower wellbeing scores of teachers when compared to working population samples (Johnson et al., 2005; Kidger et al., 2016).

There are three theoretical assumptions in our study which come from previous findings presented above: (1) the burnout syndrome is caused by chronic stress, (2) the stress results from threatened or lost resources that establish unbalance between resources and challenges, (3) the balance between resources and challenges is a fundamental feature of well-being. Based both on research findings and on our psychological experience, we postulate burnout syndrome and well-being as the opposites. If teachers report high burnout, they do not report high wellbeing and vice versa. This postulation brings us to questioning differences between teachers with a high level of burnout syndrome and with a high level of well-being. The goal of the presented study is to compare those two groups of teachers to find out in which characteristics they differ. Some of these differences can be a field for the prevention of burnout syndrome among teachers. The prevention of burnout syndrome makes a frame of our study because we argue similarly to Loonstra et al. (2009) that teacher burnout is a barrier for the higher quality of the educational system.

Burnout prevention has been studied for more than three decades. During this period, the scientific attention has been moved from demographic characteristics, like gender or age, to more subtle and psychological relevant characteristics. The latest research findings confirm the protective role of positive, problem-centered coping strategies (Austin et al., 2005),

strong professional self-efficacy (Skaalvik and Skaalvik, 2007) and collegial social support (Martin and Dowson, 2009; Buchwald, 2013).

Coping strategies are defined as the patterns of an individual's responses to stress, which remains relatively stable – either over time or in different situations (Frydenberg, 2008) and as the degree to which an individual is able to manage stress experienced in certain situations (Brouwers and Tomic, 2000). Most authors agree that some coping strategies help to avoid stress-producing situations due to changing the source of stress, while others do not because they are focusing just on the actual regulation of negative emotions (Tobin et al., 1989; Lazarus, 1993; Kyriacou, 2001; Marroquin et al., 2010). In our study, the typology by Janke and Erdmann (2003) was used. It distinguishes six coping strategies as positive strategies because they reduce stress (play down, guilt denial, substitutional satisfaction, situation control, reaction control, and positive self-instruction); four are considered negative strategies because they do not remove the source of stress and thus the level of stress increases over the long-term (flight tendency, rumination, resignation, and selfaccusation). Two strategies are characterized as neutral because their effects are ambivalent and depend on the situational context (need for social support and active avoidance). Kepalaite (2013) found that teachers who used positive reappraisal (e.g., situation control) were inclined to use other positive strategies more often as well (e.g., positive self-instruction); and vice versa, teachers who used negative strategies (e.g., escape) also used other negative strategies more often (e.g., distancing). Studies by Chwalisz et al. (1992), Austin et al. (2005), Marroquin et al. (2010), and Shin et al. (2014) show that negative strategies are associated with higher burnout levels.

Teacher self-efficacy is defined as the set of beliefs that teachers hold about their professional competencies and their efficiency to plan, organize, and carry out different educational activities and to solve challenging situations (Guskey and Passaro, 1994; Brouwers and Tomic, 2000; Bandura, 2006; Skaalvik and Skaalvik, 2014). Guskey and Passaro (1994) emphasize that teachers with high self-efficacy believe in their personal influence, power, and impact on the educational process, including students' learning (internal dimension) and factors outside the classroom (external dimension). Several studies have confirmed a moderate negative correlation between teacher self-efficacy and burnout (Brouwers and Tomic, 2000; Skaalvik and Skaalvik, 2007).

Social support at the workplace is defined as the set of activities by colleagues, managers, and leaders which are interpreted as helpful or intended to be helpful (Deelstra et al., 2003; Harris et al., 2007). Social support covers a variety of interpersonal behaviors among colleagues (such as providing emotional support, assisting, mentoring, etc.) that enhance the psychological or behavioral functioning of individual workers. Most studies find social support at the workplace to be positively predictive of job satisfaction and well-being (Harris et al., 2007) and negatively predictive of burnout syndrome (Myung-Yong and Harrison, 1998).

Earlier studies have confirmed the positive effects of each characteristic described above on the prevention of teacher burnout syndrome. Teachers reporting high self-efficacy scored lower in burnout scale (Skaalvik and Skaalvik, 2007, 2014), teachers reporting positive, problem-centered coping strategies score lower in the burnout scale (Austin et al., 2005) and teachers reporting strong workplace social support score lower on the burnout scale (Harris et al., 2007). However, there is still a lack of empirical evidence about mutual relationships between all three variables among teachers with burnout syndrome and with well-being.

With regard to previous knowledge in the field, our study attempts to find out what relationships between burnout syndrome on the one hand and self-efficacy, coping strategies, social support, and job satisfaction on the other hand exist. The overall aim of the presented study focuses on a comparison between teachers with a high and low score in burnout syndrome in the following four dimensions: selfefficacy, coping strategies, social support, and job satisfaction. We assume that teachers without burnout syndrome have stronger self-efficacy (hypothesis 1), use more often positive coping strategies (hypothesis 2) and get better workplace social support from colleagues and principals (hypothesis 3). Because of opposition between burnout syndrome and wellbeing, we assume that teachers with developing or developed burnout syndrome feel more often unsatisfied in their profession (hypothesis 4).

# MATERIALS AND METHODS

The study has quantitative correlation design using self-report questionnaires within cross-sectional approach. The study was conducted among elementary school teachers in the Czechia. The Czech educational system is specific in three ways (Education at a Glance, 2016; Ministry of Education, 2017): in its organization (primary and lower-secondary educational levels are combined in elementary schools; however, specific university training is required for each level), in its financing (schools have been underfinanced for many years and teachers' salaries are roughly equal to the national average but significantly below the national average for university graduates), and in a gradual deterioration of student achievement (as indicated by TIMMS and PISA tests).

#### Participants

Teachers were asked to participate in the study via an email distributed to all elementary school principals in the Czechia and via educational journals. The sample consisted of teachers who were willing to participate in the research, which may introduce a certain bias. However, a comparison of all teachers working in elementary schools in the Czechia in 2016 (according to 2016 Ministry of Education statistics) showed that the sample used in this study was representative in terms of gender and educational levels. An online version of the questionnaire was made available to respondents for completion for 13 weeks during the winter of 2016 and spring of 2017. All respondents were informed about the aims of the study and the ways how the results will become public. They were also ensured that their answers will be anonymized. The ethical guidelines were followed both in collecting and analyzing of data.

The study included 2,394 teachers who work in Czech elementary schools. The set of respondents consisted of 358 male teachers (15%) and 2,036 female teachers (85%). The average number of years of teaching experience was 21.42 (SD = 10.67). Czech elementary schools operate on two educational levels: the primary level is for children from 6 to 11 years of age and the lower-secondary level for students from 11 to 15 years. Each educational level requires specific teacher training. The sample included 976 primary education teachers (41%), 859 lower-secondary education teachers (36%), and 559 teachers specializing in both levels (23%).

Using burnout scale, the sample was divided in three subgroups: teachers with no burnout (n = 383), teachers at risk of burnout (n = 1,545) and teachers with developing or developed burnout (n = 464). In the analyses, sub-groups with no manifestation and with strong manifestation of burnout were compared. This groups were similar in terms of age, length of teaching practice, teacher training and school size (no significant difference was found in these variables).

#### Instruments

Burnout was measured using the Shirom–Melamed Burnout Scale (Shirom and Melamed, 2006) which had been validated for the Czech population. The Shirom–Melamed Burnout Scale conceptualizes burnout along three dimensions: physical fatigue, emotional exhaustion and cognitive weariness. The SMBM is comprised of 14 items divided into three sub-scales. The examples of particular items are: I have no energy for going to work in the morning; I feel like my "batteries" are "dead"; I feel I'm not focused in my thinking. Each sub-scale consists of several items evaluated on a 7-point Likert scale ranging from "never or almost never" to "always or almost always." High summary scores are indicative of burnout. The reliability measured by Cronbach's alpha was 0.939. The factor structure of the 13-item SMBS was explored by means of exploratory factor analysis with Varimax rotation and eigenvalues greater than 1. The analysis extracted six factors that were consistent with the theoretical SMBS model. These factors explained 81% of the variance in the equation. The expected factor loadings were greater than 0.7 for all items. Thus, the SMBS proved to be an excellent tool for measuring the sample's burnout.

Coping strategies were measured by the Stress Coping Style Questionnaire (SVF 78) by Janke and Erdmann (2003) which had been standardized for the Czech population. The measurements were based on the following assumptions. The questionnaire used a common introductory sentence: When someone upsets me, something disturbs me, or I am thrown off balance in some way, . . . Each of the 78 items completes this sentence, for example: I tell myself, "It's not that bad." The respondent selfevaluates each assertion based on the probability of his/her reaction using a 5-point scale ranging from "not at all" to "very likely." The 78 items are divided into 12 sub-scales representing the following ways of reacting to stressful situations: play down, guilt denial, substitutional satisfaction, situation control, reaction control, positive self-instruction, need for social support, active avoidance, flight tendency, rumination, resignation, and self-accusation. To interpret questionnaire scores, Janke and Erdmann (2003) distinguish between positive coping strategies (six sub-scales), negative coping strategies (four sub-scales) and two neutral coping strategies. In our study, the reliability of the SVF78 measured by Cronbach's alpha was 0.89, which is satisfactory.

Teacher self-efficacy was measured using the Czech Teacher Self-Efficacy Scale (CTSES; Smetácková et al., 2017 ˇ ) which was developed specifically for the Czech educational system per Bandura's guidelines (Bandura, 2006). The CTSES includes 45 items measured on a 5-point scale ranging from "never" to "always." Each item starts with a formulation: "I am convinced that I can. . ." and the particular professional skill follows, for example: "adapt school tasks and instructions to be clear for all children"; "cooperate with other teachers on making educational process more effective." The reliability of the scale measured by Cronbach's alpha was 0.948. The CTSES is comprised of six sub-scales: pedagogical approach, student diversity, collaboration with parents, discipline, influence on school management, cooperation among teachers, and professional development.

Social support was measured by a newly created Scale of Perceived Quality of Workplace Relationships that contained eight statements (e.g., At school I have people with whom I can deal with my work difficulties), with whom they expressed their agreement on a four-point scale from "definitely." Scale reliability was 0.8 (measured by Cronbach's alpha).

Job satisfaction was measured by a battery of items focusing on 16 aspects of teacher occupation such as relationships with colleagues, relationships with management, students, curricula, Ministry of Education etc. Respondents were asked: In which level are you satisfied with each of following areas of your profession? The level of satisfaction was expressed on a four-point scale from "very satisfied" to "very dissatisfied." Scale reliability was 0.85 (measured by Cronbach's alpha).

# Data Analysis

Data were analyzed using IBM SPSS analytics software. In the first step, the collected data were mapped to the CTSES, the Shirom– Melamed Burnout Scale, the SVF78, the Scale of Perceived Quality of Workplace Relationships and the Job Satisfaction Scale. In all five cases, the structuring of the data allowed for the use of parametric tests. We also measured reliability using Cronbach's alpha to verify that the scales had the same validity for this sample as for other studies. In the second step, mean values on all scales were determined including total mean values, sub-scale mean values and mean values in each sub-group. Differences between the mean values were determined through a t-test. In the third step, correlations for all three scales and their sub-scales were analyzed. In the last step, we performed a hierarchical regression analysis to establish the statistical significance of relationships among variables. In all particular steps, we compare two groups of respondents – teachers with high scores in the burnout scale (corresponding to present, a serious and very serious manifestation of burnout syndrome based on the SMBM norms) and teachers without burnout symptoms (corresponding to the total lack of manifestation of burnout).

# RESULTS

In the first step, descriptive statistics for all variables were computed to explore and describe characteristics of Czech teachers. The results are shown in **Table 1**. On average, teachers reported mild burnout. No burnout was reported by 16% of teachers, very mild burnout by 32% of teachers, mild burnout by 33% of teachers, present burnout by 15% of teachers, serious burnout by 4% of teachers and very serious burnout by 1% of teachers. There was no gender gap in the total scores. The strongest burnout was reported on the physical scale, with milder burnout on the cognitive scale and the lowest on the emotional scale. While male teachers reported stronger emotional burnout, female teachers reported stronger physical burnout.

The results showed that teachers used positive coping strategies more often than negative coping strategies. The most common coping strategy was substantial gratification, which belongs to positive strategies. An often used strategy was also searching for social support which belongs to neutral strategies. As an independent variable, the workplace social support was found to be also rather higher. Most teachers reported also high satisfaction with relationships with colleagues and principals. The job satisfaction, in general, was moderately high, especially in aspects related to school characteristics instead of educational system characteristics.

With regard to self-efficacy, the study showed that Czech teachers have moderately weak self-efficacy, contrary to our expectations. Most teachers perceived themselves as professionally competent sometimes or rarely. The highest level of self-efficacy was reported in the dimensions of Collaboration with parents and Professional development. On the contrary, the lowest self-efficacy was reported in the Pedagogical approach, which is considered to be a crucial aspect of teaching beliefs. Thus, these findings have important consequences for at-risk aspects of teacher identity.

In the second step, the mutual relationships among examined characteristics were measured. **Table 2** shows correlations between the study variables. All correlations with burnout syndrome were strongly significant. However, the correlation with social support was too weak and thus probably ecologically invalid. Correlations with three other variables were sufficiently strong both in terms of statistic findings and in reality. The results confirm that burnout syndrome is lower when teachers have strong self-efficacy, feel satisfied, use positive coping and avoid negative coping, and vice versa.

TABLE 1 | Means, standard deviations, minimum, and maximum (N = 2,394).


#### TABLE 2 | Bivariate correlations (N = 2,394).


Correlations were significant, ∗∗p < 0.001.

The correlation between teacher burnout syndrome and teacher self-efficacy was −0.293. Teachers with high self-efficacy scored low in burnout symptoms and vice versa. All three dimensions of burnout showed a significant negative correlation with teacher self-efficacy. However, the strongest value was reported for the emotional dimension (−0.375). Teachers with high self-efficacy scored low in emotional burnout, which could have the most negative impact on relationships between teachers and students and thus on student well-being in schools.

Burnout syndrome was strongly correlated to both types of coping strategies. The correlation between positive coping and burnout symptoms was 0.259 and the correlation between negative coping and burnout symptoms was even as high as 0.553. The more teachers used negative coping strategies, the higher they scored for burnout symptoms. The more teachers used positive coping strategies, the lower they scored for burnout symptoms. Positive and negative coping strategies are not mutually exclusive; a person can use both positive and negative coping strategies during the same period of his/her life. However, a negative correlation was found between these two types of coping strategies (−0.291). This implies that most teachers tend to use only one coping style.

A significant correlation between both coping styles and teacher self-efficacy was found as well. Positive coping correlates positively (0.354) and negative coping correlates negatively (−0.214). This means that teachers with high self-efficacy score high in positive coping but rather low in negative coping. All sub-scales of teacher self-efficacy show similar patterns.

Moreover, a significant correlation between job satisfaction and all other variables was found. Teachers with higher job satisfaction reported not only lower burnout syndrome, but also stronger self-efficacy, more frequent use of positive coping and avoiding of negative coping and receiving better social support (although, this relation was relatively weak).

In the third step, respondents were divided into three subgroups according to global score in the Shirom–Melamed Burnout Measurement. The norms indicated teachers with no burnout manifestation and five more levels of increasing burnout manifestations. The three highest levels – presented, serious and very serious burnout manifestations – were put together and established the sub-group called "developing and develop burnout." Two subgroups with no burnout and with existing burnout were compared in the following statistical analyses. The group with no burnout reported the global score M = 21.45

TABLE 3 | Means, standard deviations, and significance of difference (N = 847).


TABLE 4 | Hierarchical regression analysis for predicting burnout syndrome.


(SD = 3.588), while the group with developing/developed burnout syndrome reported the global score M = 60.89 (SD = 7.502). The gap between global scores is significant, p < 0.001. In the following steps, the variables between identified subgroups were compared. The means and their statistical significance are shown in **Table 3**.

There is a significant difference between burnout and nonburnout teachers in all variables. Teachers with high score in burnout scale have lower job satisfaction, receive lower social support at workplace, report lower self-efficacy, use less often positive coping strategies and use more often negative coping strategies.

In the fourth step, in order to determine which variable shows a higher degree of burnout predictability, multiple linear regression was conducted using SPSS software. In a model with the burnout level as a dependent variable, total self-efficacy scores, positive coping, negative coping, social support and job satisfaction were tested in a regression equation as independent variables. All details are presented in **Table 4**.

The results of the hierarchical regression analysis showed that four variables in the model were significantly related to burnout symptoms. The only variable without significant link was Social support. Teachers' burnout level decreased by 2.30 points for each standard score of total job satisfaction, by 0.13 point of each standard score of total self-efficacy, by 0.05 point for each standard score of total positive coping, while the burnout level increased by 0.467 points for each standard score of negative coping.

The same model was tested separately for subgroups of teachers with no burnout and with developing/developed burnout. R 2 value was 0.037 for no-burnout-group and 0.109 for burnout group. A significant degree of burnout predictability was found only for self-efficacy in the no-burnout-group (p < 0.046), while for negative coping (p < 0.001) and job satisfaction (p < 0.024) in the burnout-group.

#### DISCUSSION

The aim of the study was to test whether there is a relationship between teacher burnout syndrome, self-efficacy, coping strategies, workplace social support, and job satisfaction. In our study, we apply a perspective of the conservation of resources theory by Hobfoll (2002) and the perspective of wellbeing as a balance between resources and challenges (Dodge et al., 2012). Based on these theories, we search for professional resources which can teachers use to preserve their pedagogical energy and competencies. In the listed variables, we have searched for differences or accordance with teachers with a high level of burnout symptoms and without any burnout symptoms.

Earlier studies have found strong evidence for a connection between burnout syndrome and self-efficacy (Brouwers and Tomic, 2000; Skaalvik and Skaalvik, 2007), for a connection between burnout syndrome and coping style (Chan and Hui, 1995) and for a connection between burnout syndrome and social support (Myung-Yong and Harrison, 1998). Some studies also found evidence for strong protective function of job satisfaction to burnout syndrome (Harris et al., 2007). However, connections between all these variables have not been analyzed sufficiently to date. The lack of empirical findings exists from educational systems out of United States and Western Europe. Current knowledge about prevention of burnout syndrome and promoting teacher wellbeing need to be completed and compared with research from other countries, like the Czechia as post-communist country. Every educational system, with its specific values and organizational structure, introduces different challenges and accepted or rejected ways of responding to them (Chan and Hui, 1995). The Czechia as a new research location in this field can be therefore taken as a validation of previous findings on the burnout prevention. The presented study results in the conclusion that examined variables as concepts universally exist, however, the ways how they are manifested and how they can be measured are partially culturedependent (Bandura, 2006).

Our study worked with four hypotheses derived from previous research. We expected that teachers with developing or developed burnout syndrome have stronger self-efficacy (hypothesis 1), use more often positive coping strategies (hypothesis 2) and get better workplace social support from colleagues and principals (hypothesis 3). With regards to the opposition between burnout syndrome and wellbeing, we assume that teachers with developing or developed burnout syndrome feel more often unsatisfied in their profession (hypothesis 4). All four hypotheses were confirmed.

Our study has confirmed that there is a close relationship between burnout and job satisfaction and also self-efficacy, coping strategies and social support among teachers. Teachers with lower burnout scores reported stronger self-efficacy, more frequent using of positive coping strategies, avoiding negative coping strategies and better workplace social support. This result is relevant for the conservation of resources theory (Hobfoll, 2002) in rationalizing how specific psychological resources may shield teachers against the hazards of work burnout.

Most findings from our study were in correspondence with previous studies on self-efficacy. Austin et al. (2005) found that teachers who reported lower stress levels used positive coping more often, while teachers who reported higher stress levels used negative coping more often. Skaalvik and Skaalvik (2007, 2014) showed that teachers with high self-efficacy scored lower on burnout symptoms and vice versa. Brouwers and Tomic (2000) reported that teacher self-efficacy was significantly related to their burnout levels, however, only with respect to the depersonalization and personal-accomplishment dimensions of burnout, and not with respect to the emotionalexhaustion dimension. In our study, all three dimensions of burnout syndrome – physical, emotional, and cognitive – were significantly related both to self-efficacy and to coping strategies.

Also, our study is in agreement with previous findings on coping strategies. Studies by Chwalisz et al. (1992), Austin et al. (2005), Marroquin et al. (2010), and Shin et al. (2014) showed that using positive, problem-centered coping is linked to lower score in burnout, while using negative, emotion-centered coping is linked to higher score in burnout. However, based on these studies, we assumed that protective function of positive coping is stronger than inhibition caused by negative coping. Our study showed that the connection between burnout and negative coping was stronger. It means that the avoidance of negative coping helps in prevention against burnout syndrome more effective than using positive coping.

Many previous studies showed that social support at workplace had positive effect for job satisfaction and for prevention of burnout syndrome (Harris et al., 2007; Leiter et al., 2014). Our study confirms this result. However, the connection between social support and burnout syndrome was not as strong as we expected. The social support does not have a predictability degree for burnout score in either the noburnout-group nor burnout-group. In a separate analysis, the significant predictability showed only self-efficacy for teachers without any burnout symptoms and job satisfaction and negative coping for teachers with developing/developed burnout. We find this finding challenging, and encourage further studies in this field because it has a high relevance for supporting teachers' wellbeing.

The presented study has several limitations. One limitation of this study is that we measured only four variables, however, there are other dimensions of the school and individual context (personal traits or collective efficacy e.g.) which should be analyzed in future research. As another limitation to be considered is the fact that the study was conducted just in the Czechia. However, it is not only the national study with a regional contribution. Czech teachers had not previously participated in any international research about this issue. The Czech educational system is different from systems in the United States, Western Europe, and Scandinavia, where the majority of research has been conducted. This study in the Czech as a new research location in this field can be therefore taken as a validation of previous findings on the correlation between burnout and self-efficacy and burnout and coping strategies. The presented study results in the conclusion that these concepts universally exist, however, the ways how these concepts are manifested and how they can be measured are culture-dependent (Bandura, 2006).

Cultural dependency was the reason for the development of the CTSES. However, using a new instrument to measure the selfefficacy which has been yet not compared to other measurements on self-efficacy can be considered as the third limitation of the study. Therefore, it is not known whether the level of self-efficacy among Czech teachers is higher or lower than in other countries. The advantage is that the instrument measuring the selfefficacy specifically addressed teacher occupation. Conversely, the measurement of coping strategies is general and does not respect the school environment. The using of specific measurement of teacher coping strategies would be helpful in future research. We used the SVF78 to measure coping strategies of teachers, although this instrument is too general. For example, distancing represents an effort to break away from a situation and not take it seriously, which may be a functional strategy in situations where an individual has no real control over the problem. However, in the case of teaching, this strategy results in an unprofessional approach where students, their needs and the context in which the education process takes place are ignored. As our study showed, it is important to imply the perspectives specific to certain professions, in our case teaching. We consider future development of a specific measurement for teachers' coping strategies important, especial for Central European educational systems which differ significantly from school systems in the United States and other countries.

# AUTHOR'S NOTE

The research was conducted by a team of psychologists from the Charles University, Prague, Czechia, namely: IS, IV, VP, VF, AP, SS, Radek Ptacek, Petra Topkova, and Jiri Raboch.

# DATA AVAILABILITY STATEMENT

The datasets generated for this study are available on request to the corresponding author.

# ETHICS STATEMENT

Ethics approval was not required as per the local legislation and national guidelines.

# AUTHOR CONTRIBUTIONS

All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication.

# FUNDING

This research was financially supported by the Czech Science Foundation, grant no. GA16-21302S.

# REFERENCES

fpsyg-10-02287 October 29, 2019 Time: 15:16 # 8


**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Smetackova, Viktorova, Pavlas Martanova, Pachova, Francova and Stech. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Effect of Teachers' Happiness on Teachers' Health. The Mediating Role of Happiness at Work

Paula Benevene<sup>1</sup> \*, Simona De Stasio<sup>1</sup> , Caterina Fiorilli<sup>1</sup> , Ilaria Buonomo<sup>1</sup> , Benedetta Ragni<sup>1</sup> , Juan José Maldonado Briegas<sup>2</sup> and Daniela Barni<sup>1</sup>

<sup>1</sup> Department of Human Studies, Libera Università Maria SS. Assunta, Rome, Italy, <sup>2</sup> Department of Business Management and Sociology, University of Extremadura, Badajoz, Spain

The present study aims to expand the understanding of the effects of dispositional happiness and self-esteem, as dispositional traits, on the health of teachers, as well as to understand the role played by the working environment in generating positive affection, thus mediating between the dispositional traits and teachers' health. Two hundred and eighty-two full-time in-service teachers (93.6% female) from Rome (Italy) took part in this study. Their ages ranged from 26 to 55 (M = 40.49 years, SD = 5.93). Participants' teaching experience ranged from 1 to 31 years (M = 9.95 years, SD = 5.65). 30.6% of participants taught in kindergarten (for children aged 0–5 years), 42.6% in primary schools (for children aged 6–11 years), 15.8% in middle schools and 10.9% in high schools. A questionnaire was administered, containing: the Subjective Happiness Scale (SHS); the Rosenberg Self-Esteem Scale (RSES); The adapted version for teachers of the School Children Happiness Inventory (Ivens, 2007); the Physical and Mental Health Scales (SF12). The data were analyzed using the MPLUS software, version 8. Our results showed that teacher happiness at work partially mediates the relationship between dispositional happiness and teacher health, and fully mediates the relationship between self-esteem and teacher health. To the best of our knowledge, the mediational role of teacher happiness has not been addressed before, concerning these dimensions. At the same time, our findings confirmed the role of self-esteem in endorsing health-related behaviors, thus promoting physical and mental health. Moreover, according to our study findings, when teachers acknowledge their workplace as a context in which they feel happy, the impact of dispositional happiness and self-esteem on health conditions is higher. Effective measures to promote teachers' well-being are discussed.

Keywords: teachers' happiness, teachers' health, happiness, self-esteem, happiness at work, happiness at school, subjective happiness

# INTRODUCTION

Teachers' health has received much more attention over the past century (Skaalvik and Skaalvik, 2011), since schoolteachers have emerged as a category of workers highly prone to a variety of psychological, mental and physical problems, as a consequence of the stress and attrition they are dealing with every day (Borrelli et al., 2014; Benevene and Fiorilli, 2015). Consistently, there is extensive academic literature showing that stress is one of the leading causes of teachers'

#### Edited by:

Badri Bajaj, Jaypee Institute of Information Technology, India

#### Reviewed by:

Marianela Denegri, University of La Frontera, Chile Claudio Barbaranelli, Sapienza University of Rome, Italy

> \*Correspondence: Paula Benevene benevene@lumsa.it

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 09 June 2019 Accepted: 15 October 2019 Published: 31 October 2019

#### Citation:

Benevene P, De Stasio S, Fiorilli C, Buonomo I, Ragni B, Briegas JJM and Barni D (2019) Effect of Teachers' Happiness on Teachers' Health. The Mediating Role of Happiness at Work. Front. Psychol. 10:2449. doi: 10.3389/fpsyg.2019.02449

**174**

sickness absence, ill-health retirement and turnover (Fiorilli et al., 2015). High rates of teachers' attrition are found both among Western and non-Western countries, notwithstanding the relevant cultural, organizational and educational differences across the countries (Chang, 2009; Hong, 2010; Fiorilli et al., 2015). On the other hand, the approach of the positive psychology has shown that dispositional traits have an impact on how teachers, and more in general, individuals, successfully manage the challenges they have to face in their workplace and cope with stressful events (Xanthopoulou et al., 2007; De Stasio et al., 2017). It is also a fact that happiness in the workplace is often influenced by one's own working experiences events. There is evidence that individuals may experience a higher level of happiness than usual days, compared with their baselines, when they feel fulfilled in their basic needs for competence, autonomy and relatedness in major activities in their workplace (Sheldon et al., 1996; Reis et al., 2000; Buonomo et al., 2019). It has to be noted that happiness at work has been mainly approached in relation to the effects of the characteristics of the organization or the job on individuals, thus devoting little attention to the cognitive and affective processes. From these studies, it has emerged that the main factors leading to a higher degree of happiness at work are: positive interpersonal relationship, both with colleagues and superiors (Macuka et al., 2017); performing activities perceived as meaningful and fruitful; recognition for the achievements (Fisher, 2010), fair treatment (Chaiprasit and Santidhiraku, 2011) and a positive family and work-life balance (Golden et al., 2013). On the other hand, the effect of the stable attributes of individuals on happiness at work has recently been recognized, thus showing the need for deepening the knowledge on how dispositional, psychological variables are linked to job happiness (Fisher, 2010; Warr, 2014). Namely, individual characteristics identified as antecedent happiness in the workplace are: personality traits such as extraversion, responsibility and agreeability (Warr, 2014), as well as positive dispositional affect (Fisher, 2010; Seligman, 2011).

The present study aims to deepen the understanding of the effect of dispositional happiness and self-esteem, as dispositional traits, on the health of a group of teachers, as well as the understanding of the role played by the working environment in generating positive affection, thus mediating between the dispositional traits and teachers' health. Teachers represent a group of employees that is still poorly addressed by the literature in this regard, therefore, addressing this topic might offer the opportunity for identifying effective measures to promote teachers' well-being (Fisher, 2010; Kok et al., 2013).

# Happiness as a Dispositional Trait Linked to Teachers' Health

Happiness is probably the dispositional trait most studied in relation to individuals' health (Fredrickson and Levenson, 1998), despite the paucity of studies carried out among teachers. The construct of dispositional, subjective happiness, refers the global judgment of oneself as a happy or an unhappy person, which tends to be rather stable across years, in spite of life changes and daily routines (Diener et al., 1997; Lyubomirsky and Lepper, 1999; Lucas, 2007; Luhmann et al., 2012).

A considerable number of studies have proven the protective role in facing life challenges and therefore generating positive health outcomes, such as subjectively perceived health, disease rates and severity, as well as mortality. For instance, the effect of happiness on longevity among the healthy population is similar to that of smoking or not, while negative emotions are often predictors of worse health outcomes (Diener and Seligman, 2004; Veenhoven, 2008). Dispositional happiness protects individuals' health at least in three ways. Firstly, chronic unhappiness activates the flight-flight response, which is the physiological reaction that occurs in response to a perceived harmful event, attack, or threat to survival. This activation in the long run is related to harmful effects, such as higher blood pressure and a lower immune response. Furthermore, studies carried out in the field of psychosomatic medicine have proven the protective effects of positive mental states against illness, since they enhance better immune response (Cohen et al., 1995; Robinson et al., 2012). Secondly, happy people are more inclined to live healthier. Happy people tend to watch their weight, do not exceed in smoking and drinking, and engage in sports. Moreover, they are quicker at identifying symptoms of illness and more successful in coping with threatening information (Graham, 2008; Angner et al., 2009; Van Cappellen et al., 2018; Steptoe, 2019). Thirdly, happiness may promote health through the development of a more extensive repertoire of positive actions to cope with life events (Zautra, 2006). Happy people are more able to develop and maintain supportive social networks and to make better choices in life, since "they are more open to the world and more self-confident. they are also less likely to fall victim to the pattern of onedimensional thinking in distress, which might hamper choice" (Veenhoven, 2008, p. 459).

The effects of happiness on individuals' health may be explained in the light of the Fredrickson's (2013) broadenand-build theory, which posits that positive affect helps to 'build' resources that are effective in creating healthier living conditions. More specifically, "certain discrete positive emotions—including joy, interest, contentment, pride, and love—although phenomenologically distinct, all share the ability to broaden people's momentary thought-action repertoires and build their enduring personal resources, ranging from physical and intellectual resources to social and psychological resources." (Fredrickson, 2001, p. 220). Thus, experiencing positive emotions tends to expand the spectrum of the options to be considered; conversely, negative emotion may narrow the strategies available to the mind. Moreover, the resources generated by positive emotions last over time, and therefore can be retained for future use, in different emotional states (Fredrickson and Branigan, 2005). As far as the school is considered as a workplace, the paucity of previous studies carried out on the subjective happiness and health among teachers, belonging to different countries and educational systems, has confirmed the link between these two constructs. Namely, from a comparative study, carried out in Hong Kong and Italy among two groups of kindergarten teachers, it emerged a negative correlation of self-esteem and subjective well-being with mental health complaints. Another study carried out in India, among K-12

teachers in the state of Kerala, proved the correlation of selfesteem and subjective happiness on teachers' health (Benevene et al., 2018a,b). However, no previous study has addressed the mediating role of positive affect at works between dispositional happiness and teachers' health. The relevance of deepening the knowledge about this dynamic lies in the opportunities that increased awareness of this subject offers, in terms of preventing teachers' illness and enhancing their health, through training and positive school management.

Therefore, the following hypothesis was formulated:

H1. Teachers' subjective happiness is related to teachers' health.

### Self-Esteem as a Dispositional Trait Linked to Teachers' Health

Like subjective happiness, self-esteem is also a dispositional trait that highly contributes to individuals' health and quality of life (Antonucci and Jackson, 1983; Evans, 1997). Unsurprisingly, there are pieces of evidence that self-esteem is highly linked with happiness (Hornsey et al., 2018). Self-esteem is described as the evaluative and affective dimension of self-concept, as it refers to the global appraisal of one's overall positive or negative value (Brown and Marshall, 2001; Leary and Baumeister, 2004; Rosenberg et al., 2006).

Self-esteem operates as a protective factor, buffering against the impact of adverse life events. There is evidence that it actively promotes the healthy functioning of individuals in managing their life aspects, generating positive social behavior, achievements, satisfaction, fulfillment, as well as the ability to cope with diseases like cancer and heart disease (Baumeister et al., 2003). Conversely, people with low self-esteem are at a high risk of developing depression, burnout, anorexia nervosa, bulimia, anxiety, violence, substance abuse and highrisk behaviors (Schimmack and Diener, 2003; Schmitz et al., 2003; Collie et al., 2012; Sowislo and Orth, 2013). The protective nature of self-esteem has become evident through the studies where chronically ill individuals were observed. They showed that high self-esteem, in combination with enjoying close relationships, directly protects individuals from developing depressive symptoms (Penninx et al., 1998; Prince et al., 2007). From these studies, it emerged that self-esteem prevents from developing stress or physical disease in situations where individuals might experience fear and uncertainty (Mann et al., 2004). More in general, self-esteem has been addressed by many studies on stress and coping, demonstrating how this factor buffers the impact of stressors. In fact, according to the transactional model of stress and coping developed by Lazarus and Folkman (1984), high self-esteem mitigates the perceived threats and improves the quality of the coping strategies. As Lazarus and Folkman state: "Viewing yourself positively can also be regarded as an essential psychological resource for coping. We include in this category those general and specific beliefs that serve as a basis for hope and that sustain coping efforts in the face of the most adverse condition" (Lazarus and Folkman, 1984, p. 159). Thus, self-esteem can be seen as an internal moderator of stressors (Mann et al., 2004) and, more in general, as a crucial component of individual adaptability, thereby enabling resilience in facing difficulties (Hobfoll, 2002; Betoret, 2006; Tsouloupas et al., 2010). Unsurprisingly, from many studies it has arisen that self-esteem is a crucial resource for coping with the challenges of teaching, playing a leading role in teachers' health (Mäkikangas et al., 2004; Benevene et al., 2018a). Conversely, low self-esteem is associated with mental health problems, such as depression and burnout (Roberts and Kendler, 1999; Schimmack and Diener, 2003; Collie et al., 2012).

Therefore, the following hypothesis was formulated:

H2. Teachers' self-esteem is related to teachers' health.

# The Mediating Role of Happiness at Work Between Dispositional Traits and Teachers' Health

Despite their dispositional traits, teachers have to face and cope with specific working conditions, which might result in either challenging or enhancing their happiness. Teachers do not operate in a vacuum and schools are an emotional arena, able to elicit positive or negative feelings in the interaction of individuals with their working environment. As Fineman (2012) noted, teaching experience may generate significant and repeated unpleasant emotions, eventually leading to high level of stress and negative consequence on the general state of health. In this regard, the transactional theory of Lazarus and Folkman (1984) posits the dynamic, mutually reciprocal, bidirectional interactions of teachers with their workplace, including factors such as teacher perceived leadership style, attribution of student misbehaviors, and perceived exchange of investments and outcomes (Bibou-Nakou et al., 1999; Friedman, 2000; Evers et al., 2004; Van Horn et al., 2004). This theory underlines the cognitive process through which individuals mold their interaction with the specific characteristics of their environment, in the light of their assessment of the stimuli and their coping strategies. Therefore, happiness and selfesteem, as dispositional resources, not only impact individuals' health, but also how individuals respond to their working environment, activating further resources to face the challenges of their workplace (Sutton et al., 2009). There is evidence to prove that people with positive thoughts are healthier and more successful in their working life; they are more productive and efficient, generating a competitive advantage for their organizations (Momeni et al., 2011). Therefore, it can be hypothesized that happiness and self-esteem, as dispositional traits, may be linked with happiness in the workplace, namely teachers' happiness at work. On the other hand, the interaction between individuals and their working environment, as the transactional theory, also suggests the workplace may elicit positive or negative feelings and emotions, therefore generating positive or negative affect at work, which in turn has an effect on teachers' health (Argyle, 1997; Fredrickson, 2000). In fact, positive thoughts at work help individuals to enjoy a healthier life (Raei Dehaghi, 2012). Unsurprisingly, the impact of happiness on the health of employees is strictly linked with low levels of absenteeism and sick leave and high performance (Peterson and Park, 2011).

Therefore, the following hypotheses were formulated:


At the same time, acknowledging the multiple theories that account for relationships between positive emotions and selfesteem, which differ from the ones considered in our study, will allow for considering alternative associations among our predictive and mediating variables, as described in the data analysis strategy. More specifically, while the measured variables are suitable to verify the hypothesis that self-esteem is a predictor of happiness (e.g., Ford et al., 2016), they are not suitable to compare our model with others, considering more complex relationships (e.g., Alessandri et al., 2012; Barbaranelli et al., 2019).

#### MATERIALS AND METHODS

#### Participants and Procedure

Two hundred and eighty-two full-time in-service teachers (93.6% female) from Rome (Italy) took part in this study. Their ages ranged from 26 to 55 (M = 40.49 years, SD = 5.93). Regarding demographic variables (marital status and children), 54.7% of teachers had a partner, 38.1% were single, 6.8% were separated/divorced, while only one teacher was widowed. Furthermore, sixty-four percent of participants had children. Regarding job-related dimensions, participants' teaching years of experience ranged from 1 to 31 (M = 9.95, SD = 5.65). Teachers' age and years of experience are positively correlated (Pearson's r = 0.365, p = 0.000). 30.6% of participants taught in kindergartens (for children aged 0–5 years), 42.6% in primary schools (for children aged 6–11 years), 15.8% in middle schools and 10.9% in high schools. Thus, the teachers involved constitute a convenience sample, which cannot be considered as being representative of the entire population of Italian teachers. Data were gathered by the research group at the end of school board meetings, on a voluntary basis in an individual setting. The entire process was anonymous. Participants took part to the study after having received written information on Italian privacy regulations and having signed informed consent. The research was conducted following the APA's ethical principles and code of conduct (American Psychological Association, 2016). When an Italian validation was not available, the original versions of questionnaires were initially translated from English into Italian and then back-translated into English to check the alignment with the original versions.

#### Measures

The Subjective Happiness Scale (SHS; Lyubomirsky and Lepper, 1999; Italian version by Iani et al., 2014) is a 4-item scale aimed at assessing subjective happiness, using a 7-point Likert scale. The first two items ask people to rate how they happy they are about their life in general (1 = not a very happy person, 7 = a very happy person) and how happy they are in comparison with their peers (1 = less happy, 7 = happier); the last two items ask respondents to what extent the characterization of a happy and an unhappy person describe themselves (1 = not at all, 7 = a great deal). Higher scores on this measure indicate higher subjective happiness. Cronbach's alpha was 0.81.

The Rosenberg Self-Esteem Scale (RSES, Rosenberg, 1965; Italian version by Prezza et al., 1997) consists of 10 items and assesses self-esteem, with a 4-point Likert scale (1 = I strongly disagree, 4 = I strongly agree). Higher scores indicate higher selfesteem. Scores between 15 and 25 are within normal range; scores below 15 suggest low self-esteem. Cronbach's alpha was 0.81.

Teacher Happiness at work was evaluated with the adapted version for teachers of the School Children Happiness Inventory (Ivens, 2007). The adapted version consists of 33 items, measured on a 4-point Likert scale (1 = I strongly disagree, 4 = I strongly agree). Higher scores indicate high perceived happiness at school. The adapted version, despite not being validated yet, showed good reliability (Cronbach's alpha 0.93).

The 12-item SF Health Survey (SF12; Ware et al., 1996; Italian version by Kodraliu et al., 2001) is a short survey composed of 12 questions, selected from the SF-36 Health Survey (Ware et al., 1996). Items are grouped into two sub-scales: Physical health (PH, 6 items), assessing limitations in physical activity and functioning, pain and overall health; Mental health (MH, 6 items) assessing mental health, vitality, and social functioning. The scale comprehends items assessed with dichotomous questions, as well as measured with 3- to 6-point Likert scales. Overall, the score ranges from 0 to 100, where higher scores represent better health. Cronbach's alpha was 0.83.

#### Data Analysis

First, a Confirmatory Factor Analysis (CFA, Kline, 2015) was performed in order to examine the measurement model with MPlus version 8 (Muthén and Muthén, 2017).

To enhance the reliability and parsimony of our model, item parcels were created for self-esteem measure (10 items) and teacher happiness at work measure (33 items). Each factor was defined by three parcels, to obtain less free parameters to estimate and to reduce the sources of sampling error (Little et al., 2002, 2013; Coffman and MacCallum, 2005), and each parcel was created by sequentially summing items assigned based on the highest to lowest item-total corrected correlations (Little et al., 2002, 2013; Coffman and MacCallum, 2005). The Robust Maximum Likelihood Approach (MLR) was used to deal with non-normality in data (Wang and Wang, 2012).

Next, the structural model (Model 1) was tested by using the structural equation modeling (SEM) approach (Kline, 2011). The model was conceptualized by using subjective happiness (as measured by SHS), self-esteem (as measured by RSES), happiness at work (as measured by the adapted version of TSCHI), and health (as operationalized by physical and mental health). We hypothesized both direct and indirect (through teachers' happiness at work) effects of subjective wellbeing and self-esteem on health. Moreover, we tested an alternative model (Model 2) in order to investigate whether a model, in which dispositional and work-related happiness mediated the effect of self-esteem, better predicted health outcomes than a model in which self-esteem and subjective happiness were considered as dispositional variables.

In order to compare the two models, given that Model 2 was non-nested within our hypothesized model, we used three information criteria: AIC, BIC and Sample-Size Adjusted BIC. Lower values of these indices indicate a better model (Wang and Wang, 2012).

The following procedures of data exploration were applied: (a) uni- and multivariate outlier analysis (Mahalanobis's distance was set to p < 0.001, Gath and Hayes, 2006); (b) score distribution analysis (skewness and kurtosis cut-off points were set to [−2; +2] (George and Mallery, 2003); (c) missing value analyses (missing values were skipped listwise, Little, 1992). At the end of these procedures, we obtained the sample described above.

#### RESULTS

#### Measurement Model

The measurement model showed a good fit to the data: χ2(48) = 66.974, p = 0.036, CFI = 0.977, TLI = 0.968, RMSEA = 0.056 (90% CI = 0.015–0.086), p = 0.357, SRMR = 0.054., confirming validity and distinguishability of the four theoretical constructs.

The correlations among the studied variables are presented in **Table 1**. Teacher happiness at work was significantly correlated with teachers' subjective happiness (r = 0.48), perceived selfesteem (r = 0.47) and health (r = 0.63). Moreover, teachers' health was significantly correlated with teachers' subjective happiness (r = 0.47) and perceived self-esteem (r = 0.40). Sociodemographic and job-related variables are not shown, as their associations with the variables of interest are not significant.

#### Final Model

Model 1 (**Figure 1**), hypothesizing both direct and indirect (through teachers' happiness at work) effects of subjective happiness and self-esteem on health, proved to be an adequate fit to the data: χ2(48) = 66.974, p = 0.036, CFI = 0.977, TLI = 0.968, RMSEA = 0.056 (90% CI = 0.015–0.086), p = 0.357, SRMR = 0.054. Overall, subjective happiness was associated with both teachers' health (b = 0.23, p = 0.029) and teachers' positive affects at work (b = 0.390, p = 0.000). Furthermore, selfesteem resulted associated with teachers' positive affects at work (b = 0.25, p = 0.005), while the association between self-esteem with teachers' health was not statistically significant. Finally, teachers' happiness at work showed a significant direct effect on teachers' health (b = 0.57, p = 0.000). The percentages of variance explained were 53.6% for teachers' health and 30.6% for teachers'


happiness at work. Subjective happiness and self-esteem were positively correlated (r = 0.44, p = 0.000).

In line with our hypothesis (H3 and H4), teachers' happiness at work mediated the effects of both subjective happiness and self-esteem on teachers' health. Specifically, teachers' happiness at work partially mediated the effects of subjective happiness on health (bDIRECT = 0.23, p = 0.029; bINDIRECT = 0.22, p = 0.001; total indirect effect:0.45, p = 0.000) and fully mediated the effect of self-esteem on teachers' health (bDIRECT = ns.; bINDIRECT = 0.15, p = 0.021; total indirect effect: 0.18, p = 0.041; Hayes, 2013).

The alternative model (Model 2) examined both direct and indirect (through teachers' happiness at work and subjective happiness) effects of self-esteem on health. The model showed a poor fit to the data χ2(49) = 82.169, p = 0.002, CFI = 0.960, TLI = 0.946, RMSEA = 0.073 (90% CI = 0.044–0.100), p = 0.087, SRMR = 0.101. Moreover, comparing the two models, Model 1 showed lower values on the three information criteria (AIC = 5803.565, BIC = 5922.689, Sample-Size Adjusted BIC = 5789.872) than Model 2 (AIC = 5817.272, BIC = 5933.560, Sample-Size Adjusted BIC = 5803.905). For these reasons we concluded that Model 2 could be rejected, and we retained Model 1 as the final model of the study.

#### DISCUSSION

Our findings show that teacher happiness at work partially mediates the relationship between dispositional happiness and teacher health, and fully mediates the relationship between selfesteem and teacher health. To the best of our knowledge, the mediational role of teacher happiness has not been addressed before with regard to these dimensions.

With regard to the role of dispositional happiness and selfesteem in predicting health, we confirmed, in a sample of Italian teachers, that individuals who tend to experience positive emotions about themselves and their life events are more likely to build healthy life conditions (Fredrickson, 2001; Pressman et al., 2018). Our results confirmed the broaden and build theory (Fredrickson, 2001), which assumes that experiencing and acknowledging positive emotions leads to several positive outcomes, in terms of physical (Fredrickson et al., 2000; Gloria et al., 2013) and mental health (Tugade and Fredrickson, 2007; Tugade, 2011), and, more generally, in terms of positive selfperceptions (Gloria et al., 2013). According to Pressman et al. (2018), in fact, acknowledging positive affects heightens the likelihood of pursuing positive health behaviors, and influences the patterns of physiological arousal, thus leading to better health outcomes. At the same time, our findings confirmed the role of self-esteem in endorsing health-related behaviors (Mäkikangas et al., 2004), thus promoting physical and mental health (Mann et al., 2004; Macinnes, 2006; Benevene et al., 2018a), despite only with the mediation of the workplace context.

At the same time, findings of our study shed light on the role of happiness at work. More specifically, it emerged that the impact of dispositional happiness and self-esteem on health conditions is higher when teachers acknowledge their workplace as a context in which they feel happy. This result is consistent

with Lyubomirsky's model of hedonic adaptation (Lyubomirsky, 2012; Dutt et al., 2013). According to this author, being prone to happiness is not enough when it comes to prolonging the positive effects of happiness over time. The hedonic adaptation effect suggests that people tend to get used to happiness and positive emotions, unless they experience more variety and frequency of positive events in their life. Consistently, the teaching profession includes several dimensions: didactical, pedagogical, caring, relational (Beijaard et al., 2000; O'Connor, 2008; Hanna et al., 2019). Furthermore, the ways teachers perceive these dimensions differs according to personal history, teaching context, professional experiences, and so on (e.g., Beijaard et al., 2000; Buonomo et al., 2017; Fiorilli et al., 2017). These findings suggest that teachers, considering the heterogeneous nature of their profession, may experience a variety of duties, relationships and tasks during their daily job. This is confirmed by studies about organizational citizenship behaviors and extra-mansions tasks in teachers (Somech and Drach-Zahavy, 2000; Belogolovsky and Somech, 2010). Thus, our results could be explained by Lyubomirsky's theory about the need to live in social contexts perceived as various and stimulating.

Moreover, Bao and Lyubomirsky (2014) stated that positive experiences occurring in contexts in which individuals pursue intrinsic, self-determined goals boost the effects of positive experiences in the long term. Several studies have shown that teachers are prone to motivate their professional choice based on personal and social utility values, considering teaching as a mission and, thus, a way to express themselves (Schwarz, 1999; Watt and Richardson, 2007; Fokkens-Bruinsma and Canrinus, 2012).

Consistently, Pressman et al., 2018 stated that the influence of positive affect on health is partially mediated by individuals' social, intellectual, psychological and physical resources. With this regard, other authors claim that positive workplaces support these resources, as they are inherently interesting, empowering and supportive for employees (Henry, 2015). Therefore, it is likely that the teachers who recognize their workplaces as sources of happiness, have higher chances to feel healthy. Previous studies have included several jobrelated sources of happiness in teachers: students' achievement and recognition, positive management of work demands, colleagues' and principal's support (Hargreaves, 2000; Tadic´ et al., 2013; Borrelli et al., 2014). Consistently, interventions aiming at promoting positive workplaces usually address job enhancement, intrinsic motivation, flexible working conditions, social exchanges, positive recognition, and participation (Henry, 2015). Living in such workplaces, in turn, promotes workers' happiness and physical health (Bono et al., 2013; Salanova et al., 2013; Gilbert et al., 2017). These effects have been found in teachers, too (Bradshaw et al., 2008; Waters, 2011; Ross et al., 2012).

Overall, our findings suggest the need for policies and interventions aimed to promote positive emotions at school. At the same time, positive emotions should not be considered as

the final aim of interventions in the workplace. Current literature suggests that happiness itself does not promote positive health outcomes in the long term (Lyubomirsky, 2012). Organizational changes are needed to pursue an effective health promotion. International surveys showed that teachers are at a high risk of developing illness and burnout disorders, other than attrition, low self-efficacy, and dissatisfaction (OECD, 2014). While the most recent models regarding the effect of positive emotions on health outcomes showed the buffering role of positive affect in stress conditions (Pressman et al., 2018), international surveys show that teachers at a lower risk for burnout and dissatisfaction receive positive, valuable feedback from colleagues and mentors, have a satisfied supervisor, and feel they can improve students' ways of learning and thinking (OECD, 2014). These data suggest that positive interventions could strengthen positive organizational dimensions and teachers' psychological resources, promoting teachers' healthy behaviors and conditions.

At the same time, the study is not without limitations. Firstly, this is a cross-sectional study, thus it is highly exposed to autoregressive biases (Maxwell and Cole, 2007). Secondly, the composition of our sample does not allow us to consider differences among teachers working in different school levels, despite several studies (e.g., OECD, 2014; Ainley and Carstens, 2018), showed that teachers are differentially exposed to risk and protective factors according to their school level. At the same time, while the high percentage of women in the sample reflects the general feminization of this profession (Drudy, 2008), this may have influenced our results, as suggested by previous studies on gender differences in health

#### REFERENCES


prevention and intervention (Gyllensten and Palmer, 2005; Hiller et al., 2017).

#### DATA AVAILABILITY STATEMENT

The datasets generated for this study are available on request to the corresponding author.

#### ETHICS STATEMENT

The study involving human participants were reviewed and approved by the Scientific Board of LUMSA University, Rome. The participants provided their written informed consent to participate in this study.

#### AUTHOR CONTRIBUTIONS

PB, SD, and CF designed and carried out the study, contributed to the analysis of the results and to the writing of the manuscript. IB, BR, JB, and DB collected the data, and contributed to the analysis of the results and to the writing of the manuscript. PB, SD, and DB supervised the study design and the manuscript draft.

#### FUNDING

The research was developed thanks to the LUMSA Research Funds.




Schools, Work, and Society, eds S. I. Donaldson, M. Csikszentmihalyi, and J. Nakamura, (Abingdon: Routledge), 49–62.



Schwarz, G. E. (1999). Teaching as vocation: enabling ethical practice. Educ. Forum. 63, 23–29. doi: 10.1080/00131729808984383


**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Benevene, De Stasio, Fiorilli, Buonomo, Ragni, Briegas and Barni. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# The Role of Digital School-Home Communication in Teacher Well-Being

#### Anne-Mari Kuusimäki<sup>1</sup> \*, Lotta Uusitalo-Malmivaara<sup>1</sup> and Kirsi Tirri<sup>2</sup>

<sup>1</sup> Faculty of Educational Sciences, University of Helsinki, Helsinki, Finland, <sup>2</sup> Department of Education, Helsinki Collegium for Advanced Studies, University of Helsinki, Helsinki, Finland

Well-functioning communication is crucial in all work communities. A respectful and trusty partnership between parents and teachers in schools is essential not only for pupils but also for the well-being of the whole organization. Communication is at the heart of such a partnership. Although most parent-teacher communication nowadays takes place on digital platforms, not much is known about the specific role of digital communication (DC) in building parent–teacher partnerships. In an attempt to find out we asked 400 Finnish parents and 80 teachers about their experiences of communicating digitally and the kind of matters they discussed, and how they thought feedback on pupils should be expressed. The data was content-analyzed. Three categories related to DC content were identified: study-related matters, behavioral issues, and sensitive issues. Parental and teachers' expectations of how pupil feedback should be expressed also fell into three categories: a good balance of encouraging and corrective feedback, more encouraging feedback, and a constant emphasis on the child's weaknesses. These results have research and practical implications for teacher well-being.

Keywords: parent–teacher partnership, digital communication, teacher well-being, teacher education, Finnish schools

# INTRODUCTION

The aim in this paper is to find out how digital communication (DC) can benefit teacher well-being and parent–teacher partnerships in Finnish schools. Parents (N = 400) and teachers (N = 80) from Finland responded to open questions related to their experiences of DC as part of a larger study. A new, specially designed 14-item Digital Communication Scale (DCS) was used in a recent study (Kuusimäki et al., 2019) to elicit the opinions of Finnish parents (N = 1123) and teachers (N = 118) on DC in urban and rural areas. According to the results, the parents and teachers were satisfied with the communication, which they perceived as supporting the parent–teacher partnership and providing valuable information on the development of pupils and on issues related to their schooling. However, the parents stated that the feedback they received about their children was less encouraging than the teachers thought it was. On the other hand, the teachers experienced more ambiguity in DC than the parents did. In general, rural parents and teachers were more positive about the effect of DC on their collaboration than their urban peers (Kuusimäki et al., 2019).

#### Edited by:

Wong Yau Ho Paul, Tung Wah College, Hong Kong

#### Reviewed by:

Lai Ping Chung, The Education University of Hong Kong, Hong Kong Dan Yu, Hong Kong Institute of Technology, Hong Kong

#### \*Correspondence:

Anne-Mari Kuusimäki anne-mari.kuusimaki@helsinki.fi

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 06 June 2019 Accepted: 20 September 2019 Published: 14 November 2019

#### Citation:

Kuusimäki A-M, Uusitalo-Malmivaara L and Tirri K (2019) The Role of Digital School-Home Communication in Teacher Well-Being. Front. Psychol. 10:2257. doi: 10.3389/fpsyg.2019.02257

Our aim in this article is to build on the results of that study and explore further the aspects of parent-teacher communication that can enhance teacher well-being. Our specific emphasis is on the nature of the DC that can support the well-being of teachers in Finnish schools.

Teacher stress and burnout are rather well-studied research areas (Maslach et al., 1996; Stoeber and Rennert, 2008; Benevene and Fiorilli, 2015; Benevene et al., 2018). Work engagement and professional dedication are sources of life satisfaction, but they may also be burdening (Hakanen et al., 2018). According to the European Working Conditions Survey (2015), employees in human-service jobs such as education report higher work engagement than those in several other job areas. Demands on schools and the growing heterogeneity of pupils and their homes seem to increase the challenges in the work of teachers (Chang, 2009; Skaalvik and Skaalvik, 2011). In a survey conducted by The Trade Union of Education in Finland (OAJ, 2017), 59% of teachers reported too heavy a workload, and 43% suffered from stress quite or very often. Starting from the beginning of the 2010s, well-being instead of ill-being has been the focus of a plentitude of studies (e.g., De Pablos-Pons et al., 2012). Teacher well-being is positively related to a good working community (Konu and Rimpelä, 2002), a fair share of tasks, support from the principal, and smooth collaboration with parents (Skaalvik and Skaalvik, 2011).

Teacher wellbeing is a critical issue in Finnish society, particularly given the decline in applications for teacher education. For the past 40 years Finnish teacher education has attracted gifted students whose first choice was the teaching profession (Tirri, 2014), a trend that is unique to Finland and could be attributed to historical and societal factors. The profession has thus far been more respected in Finland than in other countries. The current decline in applications may reflect a shift in the respect that the teaching profession attracts, or the circumstances in schools with diverse families and increasing demands for teacher competence in DC. Furthermore, the implementation of the new Finnish national curriculum (NBE, 2014), which among other developmental goals requires close school-home collaboration, has resulted in the feeling among many teachers that they do not have enough resources to manage well in their jobs (OAJ, 2017).

Joyce Epstein's seminal work (Epstein et al., 2009) on school– home collaboration underlines the shared responsibility between parents, teachers, and communities in taking care of pupils' holistic development. She established the importance of frequent interaction between schools and homes in Theory of Overlapping Spheres of Influence (Epstein, 1995). According to Epstein (2011), a successful parent–teacher partnership has several positive effects on the well-being and academic achievements of pupils (see also Samdal and Torsheim, 2012). Parents play an essential role in a school community, and a well-functioning partnership between parents and teachers constitutes the basis of the child's social learning environment (see also, Samdal, 1998; Konu and Rimpelä, 2002). Moreover, the more contact with the school and involvement in their children's studies parents have, the more likely are the children to do their homework, succeed in tests and develop positive attitudes toward school (Bauch, 1998; Freytag, 2001). Thus, parents' strong support to their child's studies can have a major impact on a teacher's work and wellbeing. Smooth communication between parents and teachers is a prerequisite for a fruitful companionship.

# Digital Communication

Ways of maintaining contact between schools and homes include parental evenings, face-to-face meetings, phone calls, printed letters, e-mails, text messages, and school web pages. The most commonly used communication channel in Finland nowadays is a digital platform through which the great majority of information is delivered. On the platform, teachers can inform parents about the timetables and events in the school, provide shorter or longer feedback on pupils' progress and discuss diverse topics with parents (Kuusimäki et al., 2019). Overall, digital platforms enable exchanging information about all running issues but also about more personal matters such as pupil progress, possible conflicts in school or learning difficulties (Carr et al., 2015; Palts and Kalmus, 2015). At its best, parent–teacher partnership is built with respectful two-way communication with frequent, trusty interaction strengthening the idea of striving toward common goals (Epstein et al., 2009). DC ensures rapid messaging, but it simultaneously has its challenges. As Epstein et al. (2009) found, essential elements of efficient parent–teacher communication include clarity, readability, frequency, quality, effectivity, and informativity. We also identified these aspects of effective DC in our previous study (Kuusimäki et al., 2019). Juniu (2009) adds four aspects that should be considered in an optimal parent-teacher collaboration: (a) positive communication including messages of encouragement from teachers to parents, (b) personalized communication, letters or feedback on the pupils' work, (c) proactive communication, and (d) partnership and collaboration between parent and teacher (Juniu, 2009). Below we explore these aspects further with regard to Finnish parents and teachers.

# Advantages and Disadvantages of DC for Teacher Well-Being

Digital communication has many advantages over traditional forms of communication, facilitating timely online information exchange and immediate feedback between parents and teachers, for example. It also makes it easier to intervene in pupils' learning and behavioral problems at an early stage, and thereby to improve study outcomes (Carr et al., 2015). Complementing the positive outcomes among pupils, effective and frequent two-way communication supports positive partnership among parents and teachers in that both receive information about what is going on at school and at home (Chu, 2018). DC could also encourage parents living separately to participate in school activities, both being able to access the digital platform (Palts and Kalmus, 2015), and it allows parents and teachers to communicate at times that suit both parties. Other advantages include easy information storage and being able to communicate with the entire school community simultaneously (Ramirez, 2001; Palts and Kalmus, 2015), as well as enhancing work satisfaction, motivation, and the

pedagogical atmosphere among teachers by allowing continuous parental support of their daily work with pupils (Scheopner, 2010; De Pablos-Pons et al., 2012; Oostdam and Hooge, 2013).

However, communication between parents and teachers is not always seamless. Given the limited expertise in DC among teachers, the lack of time during the school day and the chances of being misunderstood, problems may well arise. DC lacks exact rules, it is time-consuming, and it follows teachers from school to home thereby blurring boundaries between work and leisure (Agger, 2011; Palts and Kalmus, 2015). The idea of always being available and within reach may seriously affect teacher management of working time, work-related stress, and consequently teacher well-being (Skaalvik and Skaalvik, 2011). Moreover, experiences of not being trusted or of being criticized by parents may cause anxiety and feelings of inadequacy among teachers (Skaalvik and Skaalvik, 2011). Consequently, good communication skills are needed to prevent unnecessary conflict with parents and to build a positive partnership, mutual respect, and trust among all families (Epstein, 2013). Teachers require pedagogical sensitivity when pondering upon whether DC is the appropriate way to contact parents, especially when delicate matters need to be discussed. It should also be remembered that parental opportunities and competences with regard to the use of technology depend on age, education, income, and cultural background. Poor language skills may be a barrier, resulting in unequal communication opportunities. Schools should ensure that all parents understand the information they receive (Carr et al., 2015). DC also demands emotional and writing competence, given that misunderstandings occur more easily than in face-to-face communication (Palts and Kalmus, 2015).

In addition to reflecting on the advantages and disadvantages of DC, one needs to consider the content and the quantity of messages. Parents require frequent information regarding their children's overall progress with their studies, namely their grades, attendance record and homework (Freytag, 2001; Fusco, 2004; Weinstein, 2005), thereby enabling them to provide more active support of schoolwork. Another significant element of DC concerns the socio-emotional development of pupils. Parents should have an open and proactive dialog with teachers about their children's behavior and about issues affecting well-being such as unhappiness or problems making friends (Juniu, 2009). Without open dialog, conflicts may arise. Common agreement about the content and quality of DC should be negotiated in schools to ensure a congruent policy on informing parents.

Parents need (NBE, 2014) and want (Kuusimäki et al., 2019) supportive feedback on their children's development and studies. However, it seems as if feedback is unevenly given and gender-dependent. In Finland, Oinas et al. (2017) analyzed 211,003 digital teacher-to-parent quick-markings evidencing that boys more than girls get negative feedback and the majority of pupils get encouragement only occasionally. According to Oinas et al. (2017), teachers' feedback on pupil performance should be realistic and concrete so that parents know how to support their children in learning. Additionally, digital feedback should be sensitive, building common understanding, creating a positive learning environment, and increase school adjustment (Reddy et al., 2003). Hence, demands for a thriving DC are high and they can easily add stress in teachers. In particular, being criticized by parents online may severely affect teacher well-being.

The aim of the current study is to provide more detailed information on the views of parents and teachers on DC and how they contribute to teacher well-being. We build on the results of our previous study on digital parent-teacher communication (Kuusimäki et al., 2019) and explore further the aspects of contents of messages and form of feedback they deliver.

Specifically, we aim to find out:


#### MATERIALS AND METHODS

#### Ethics Statement

In our research, we followed the guidelines of the Finnish Advisory Board on Research Integrity (2009). This means that we respected voluntariness, anonymity, and confidentiality related to our participants and informed them in advance of the details of our study and asked for inform consents from them. According to Kvale (1996), qualitative research needs to include the following aspects to acknowledge research ethics: inform consent, confidentiality, and knowledge of the consequences of the study. First, an invitation to participate in the study was sent to two selected municipal education administrators in April 2016 as part of the larger study. We asked the permission from administrators to approach the schools selected to our study. After that, the administrators sent an informative letter to school principals and asked their consent to approach the teachers and parents of their schools. The administrators were then asked to send principals the web link to the questionnaire and principals were asked to forward the link to the teachers and the parents of the children in their schools. The administrators received two reminders about the study and the online questionnaire was open 3 weeks in total. The link was sent through the digital platform that schools are using to communicate with parents. Parents and teachers have private passwords to enter the digital platform in order to ensure the privacy of communication and in this case answering the questionnaire. Participants were informed about the voluntariness and preserving the privacy of each participant.

Our study did not deal with animals or any vulnerable groups, nor did it involve risks for participant well-being, or use of biomedical devices, or invasive investigation tools. Our study did not need ethics approval, according to our national regulations as well as to the Ethical board of the University of Helsinki.

#### Participants and Procedure

The participants of the current study were 400 parents and 80 teachers from Finland. These respondents answered to open qualitative questions related to their experiences of DC as part of a larger quantitative study. Cities selected to this study were samples of rural and urban areas, giving a relatively

diverse overview of one of the largest and one of the mediumsized municipalities in Finland. Participants were parents and teachers of pupils from grades 1–9 in comprehensive schools. **Table 1** gives the background information about the participants, including gender, overall attitude to DC, and the frequency of communicating in that way. In general, parents and teachers were positive toward DC, although the teachers would have liked more frequent contact than the parents. 14% of the teachers answered that "DC increases my workload" and 12% responded "DC is difficult because of misinterprepatations with parents."

Both quantitative measures and qualitative, open-ended questions were included in the questionnaire to provide a thorough picture of the parents' and the teachers' experiences. Below we analyze and discuss the qualitative findings: the quantitative questionnaire and the main quantitative outcomes are reported elsewhere (Kuusimäki et al., 2019). In order to get the widest range of answers, we asked three open questions from parents and teachers. Specifically, parents and teachers were asked to respond to the following questions: (1) What kind of matters would you like to be communicated digitally? (2) What kind of matters would you not like to be communicated digitally? (3) What else would you like to say about the nature of DC?

# Analysis

We subjected the qualitative data to content analysis, the purpose of which is to make replicable and valid inferences from a text (Krippendorf, 2004). The unit of analysis may vary from words to entire interviews (Elo and Kyngäs, 2008). The analysis unit for this study was the aggregate statements that parents and teachers gave to the open-ended questions about their views and expectations to content of DC. The length of responses varied from a few words to several sentences. The analysis was carried out in an inductive-oriented manner as all the codes were derived from the data (Elo and Kyngäs, 2008). The purpose was not to test any theory, but rather to reveal different conceptions of DC that parents and teachers expressed in their own words.

The analysis proceeded in three phases. First, we reduced all the aggregate statements to codes (see Examples 1 and 2). The codes were based on the data, and every time a new topic emerged a new code was created. This coding phase included multiple readings of the data. Statements that included many kinds of codes were divided among different categories accordingly.

TABLE 1 | Participants, attitudes to DC, and wanted frequency of contacts via DC.


Example 1: "If a child is being late, forgetting school materials or disturbing constantly, parents have to be informed." (aggregate statement from a parent)


Example 2: "If pupils forget occasionally to do homework or they arrive a bit late to lesson (and these are not symptoms of something more serious), it's not worth sending parents a message." (aggregate statement from a teacher)


The subcategories thus formed were further combined in three main categories (**Table 2**). The main categories were named according to contents of subcategories (Elo and Kyngäs, 2008), in these example cases (1) and (2), the main category was named Behavioral issues. The subcategories were formed when possible. Responses concerning quality of feedback formed only three main categories (**Table 3**). The first author coded the data. The

TABLE 2 | The views and expectations of parents and teachers concerning the contents of DC.


TABLE 3 | Parents' and teachers' expectations of how pupil feedback should be expressed in DC.


parental data was categorized first, and this categorization guided the analysis of the teacher data. To increase the reliability of the categorizations the second author independently rated 10% of the parental data, and the third author independently rated 50% of the teachers' data. Interrater reliability was confirmed by means of Cohen's Kappa (0.71–1.0), which was calculated separately for each category (Cohen, 1960).

# RESULTS

# The Views and Expectations of Parents and Teachers Concerning the Contents of DC

The content analysis revealed three main categories (see **Table 2**). The biggest one (n = 188) concerned Study-related matters in both groups, including issues such as homework, test dates, evaluation, and absences. The following examples reflect the definitions of the codes that were classified under this category: "Information about homework, the timetable and the contents of tests" (Parent) and "For example, information about goals, aims, and evaluation" (Teacher). Some teachers saw the digital platform as an evaluation tool, as this example points out: "The digital platform should be developed to archive evaluation about project works, group work, and pupils' attendance at lessons. It could give continuous information to parents about pupils' development in learning during the semester." The category also included various issues related to learning that require cooperation from parents and teachers and parental support. These statements refer to this category: "I need information about how to support my child in his studies, in what subjects and in what way. I would also like to know what they are studying every week, so I could discuss the contents with my child" (Parent) and "The teacher's weekly digital letter, which contains all kinds of educational and pedagogical issues to be discussed at home" (Teacher). Accordingly, teachers in difficult situations appreciated parents' cooperation and DC could deliver "information about pupils' challenges where parents' support is needed" (Teacher). Studyrelated matters also included information about class/school events. The following examples reflect the definitions used to describe this category: "I need information about my child's class events and changes in the timetables" (Parent), and "Information about trips, events and timetables" (Teacher).

The second largest category, Behavioral issues (n = 58), included aspects such as continuous misbehavior and infrequent misbehavior. The following statements exemplify continuous misbehavior: "If being late for lessons, forgetting school materials and misbehavior are becoming frequent, parents have to be informed" (Parent), and "If a child clearly has problems with interaction and in social situations" (Teacher). Informing parents about frequent misbehavior was seen important as this example reveals: "Parents have to be informed about pupils' ongoing disturbance at lessons or about having continuous conflicts with other children in order to avoid negative surprises concerning child's behavior" (Teacher). The following statements describe the teachers' contentment to DC: "The use of DC has considerably decreased misbehavior at lessons, as parents can immediately see teachers' feedback about the pupil's behavior."

The statements in category Infrequent misbehavior covered issues that parents and teachers agreed were minor, that did not need to be communicated digitally. Infrequent misbehavior is described as follows: "I don't need information about minor behavioral matters such as 'he went outside without a jacket'. . . some teachers are in the wrong profession and DC shows that" (Parent). These following statements exemplify the teachers' views of informing Infrequent misbehavior: "I don't send messages about minor things that are part of my educational work with a child" and "To my opinion, teachers don't have to inform parents every day if pupils forget to do homework or forget their school materials. This burdens teachers' work too much."

The third category, Sensitive issues (n = 51), concerned conflicts and health matters. Parents and teachers agreed that the issues included in this category should be dealt with by phone or in face-to-face discussions, and not digitally. The following examples describe conflicts: "In bullying situations or conflicts we want to be contacted by phone immediately" (Parent), and "Discussions about conflicts by phone or inviting the parents to the school, because written communication can lead to misunderstandings" (Teacher). The following examples refer to health issues: "I don't want to receive messages form a teacher evaluating a child's personality, or sending sensitive information about their mental health" (Parent), and "Information about confidential matters concerning a pupil's overall situation in school" (Teacher) or "Pupils personal issues, such as special education plans or psychologist's statements, have to be informed otherwise than in DC" (Teacher).

# Parental and Teachers' Expectations of How Pupil Feedback Should Be Expressed in DC

It is not only the contents of DC that affect the parent–teacher partnership but also the nature of the feedback: 78 respondents emphasized the importance of giving positive feedback to children. The following statements illustrate this point: "We would like to have more positive feedback. Receiving encouraging feedback is really important for our youngsters" (Parent), and "Digital platforms should be developed to give positive feedback faster and more easily" (Teacher), or "DC could be more positive. Encouraging feedback uplifts pupils. Constant negative feedback strengthens the negative image about the child." (Teacher)

This issue was also present in the parents' responses concerning a constant emphasis on child's weaknesses (n = 32), as these parent's statements exemplifies: "At this moment we only get negative feedback. Only one teacher sometimes sends positive messages, and it has encouraged our child to become more actively involved," and "My child is constantly criticized by certain teachers because no one controls DC. Some teachers see only problems."

Both parents and teachers (n = 29) also emphasized the importance of maintaining a good balance in digital feedback, and the following statements illustrate the definitions of balanced feedback: "I find that DC works great! It is important to get

both positive and 'negative' feedback and I hope I will get information in good time about the things that need to be developed (in the child's behavior or studies)" (Parent), and "It is important to give realistic and not just positive feedback, otherwise parents may have too rosy an image of the child's studying and behavior" (Teacher), or "Digital feedback have to be in balance, focusing only on negative or positive gives parents wrong information about the child's progress. I think DC is a good supplementary tool for communication, but meetings and phone calls are more important."(Teacher)

# DISCUSSION

This article presents the results of an analysis of the views and expectations of Finnish parents and teachers concerning the content of digital messages and the nature of feedback in DC. We found three main areas of its relevance or non-relevance. It seems that most issues concerning school-related matters or behavioral issues can be appropriately dealt digitally. The results are in line with our previous study, which found that parents and teachers are overall satisfied with DC and that it serves their partnership well by providing versatile information about pupils' studies and happenings at school (Kuusimäki et al., 2019). Effective and frequent two-way communication supports positive partnership among parents and teachers by conveying information about what is going on at school and at home (Epstein et al., 2009; Chu, 2018). It also allows them to intervene in pupils' learning and behavioral problems at an early stage (Carr et al., 2015). However, as evidenced in the present study, teachers should be careful when reporting problems in a pupil's behavior. The parents seemed unwilling to get information about minor and infrequent misbehavior. On the other hand, parents appreciate having frequent and versatile information about their child's studies (Kuusimäki et al., 2019). Overall, parents and teachers working together in order to share the responsibilities of a child's learning and growth can greatly decrease teacher workload (Epstein et al., 2009). The constant and immediate support from parents can affect positively on teacher's everyday work and well-being.

Parents want to have information about their child's conflicts at school (Kuusimäki et al., 2019). However, according to the present study, both parents and teachers felt that sensitive issues with pupils, like constant conflicts and health issues, should be communicated face-to-face or by phone. Choosing the right channel to communicate is essential because parents and teachers should have open and proactive communication about sensitive issues affecting children's well-being (Juniu, 2009). It is very important that teachers recognize the conflicts and sensitive matters that are best dealt with by phone or in face-toface meetings (Palts and Kalmus, 2015). Communicating about sensitive issues concerning a pupil's conflicts or health issues can cause misunderstandings between parents and teachers. This can be a major issue affecting the partnership and teacher's well-being because experiences of not being trusted or being criticized by parents may cause anxiety among teachers (Skaalvik and Skaalvik, 2011). Given the large number of sensitive matters that require attention in this study, it would seem that DC alone does not suffice to foster the parent–teacher partnership, and that teachers still need to meet parents personally. The growing demands connected to using the appropriate communication channel for various kinds of information can directly affect the teacher's workload and well-being.

In this study, parents and teachers expressed the need for more balanced and encouraging feedback on pupils. It appears from our results that there is too much emphasis still on a child's weaknesses. This result confirms the outcomes from our previous study about parents wanting more encouraging feedback about their children (Kuusimäki et al., 2019). The Finnish study by Oinas et al. (2017) showed that the distribution of feedback is uneven and the majority of pupils received only occasional encouragement. Parents need supportive and encouraging feedback about their child's studies and development (NBE, 2014). More positive feedback and encouragement might increase well-being in school communities and promote teacher well-being by building partnership and strengthening social relationships with parents.

The specific aim of the present study was to find out what kind of contents parents and teachers wish to communicate digitally and how pupil feedback should be expressed. The results were studied with regard to teacher well-being. To conclude, respectful and trusty DC that also supports teacher well-being contains the elements of frequency, clarity, prudence, proactivity, and encouraging feedback. These findings are in line with the previous studies of parent–teacher communication (Epstein et al., 2009; Juniu, 2009). We claim that DC can be one positive factor building parent–teacher partnership and enhancing teacher wellbeing. In order to build up fluent DC, there is a need of more studied information of the expectations of parents and teachers on the content and of frequency of DC. By raising teachers' awareness of parents' views, teachers can enhance their DC competences. Finnish teacher education lacks training in communication competences and in usage of DC (Alanko, 2018). It is obvious that DC needs to be addressed more carefully in future teacher education. "Best practices" in DC ought to be taught for student teachers as well as for teachers in the field (Epstein, 2018). This is not only to avoid difficulties but to promote partnerships, mutual support and well-being. According to studies, adequate teaching competences in using information and communication technologies (ICT) is factor in teacher wellbeing (De Pablos-Pons et al., 2012). With good communication skills, it is possible to enhance positive partnership, mutual respect, and trust among teachers and all families (Epstein, 2013). Yet communication skills are not enough; schools need to reflect on common policies in DC and in what time teachers conduct communication with parents, so it does not become another burden on the teacher well-being.

#### Limitations

This study provides new, qualitative information on the nature of DC between Finnish parents and teachers. The findings add more detailed information to that reported in our previous study on general trends in DC in Finnish schools (Kuusimäki et al., 2019).

We believe the results are reliable, and that they contribute to the discussion on digital home-school communication. However, some limitations should be mentioned. First, it seems that those who chose to participate in this study have a predominantly positive attitude toward DC, which may have biased the responses underlining its advantages. In order to generalize our results, having more municipalities participating in the study would give a more comprehensive picture. In the future, studies on parent–teacher communication need to include more variety in locations and school contexts. Second, there was much less data on teachers than on parents, which may reflect the heavy workload of teachers. The study was executed in May, which is the last month before summer holiday in Finland. In the future, by implementing the study earlier in semester and increasing response time, there could be more responses from teachers. Teachers' voice could be more in evidence in this study. Third, in some cases the coding was challenging due to the richness of the informants' statements. However, the three authors coded the data independently, and the kappa values indicated a good inter-rater reliability for the established categories.

Future studies should yield more information about the effect of DC on teacher well-being. Good practices

#### REFERENCES


need to be established to develop a healthy work and life balance.

#### DATA AVAILABILITY STATEMENT

The data used to support the findings of this study are available on request to the corresponding author.

# ETHICS STATEMENT

In our study, all national regulations on research ethics (Finnish Advisory Board on Research Integrity, 2009) were carefully followed. Ethics approval was not needed, according to our national regulations as well as to the Ethical board of the University of Helsinki.

#### AUTHOR CONTRIBUTIONS

A-MK, LU-M, and KT developed the research project. A-MK and LU-M developed the administration procedure. A-MK carried out the data analysis with the contributions of LU-M and KT.



**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Kuusimäki, Uusitalo-Malmivaara and Tirri. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# An Opportunity to Grow or a Label? Performance Appraisal Justice and Performance Appraisal Satisfaction to Increase Teachers' Well-Being

Laura Dal Corso<sup>1</sup> \*, Alessandro De Carlo<sup>2</sup> , Francesca Carluccio<sup>3</sup> , Damiano Girardi<sup>1</sup> and Alessandra Falco<sup>1</sup>

<sup>1</sup> Department of Philosophy, Sociology, Education and Applied Psychology, University of Padua, Padua, Italy, <sup>2</sup> Giustino Fortunato University, Benevento, Italy, <sup>3</sup> Department of Human Science (Communication, Training, Psychology), LUMSA University, Rome, Italy

Performance management is a key factor to enhance professional development and improve teaching quality. This process is successful only if teachers perceive it as fair, clear, and effective: namely, if it is satisfying. Carefully considering teachers' attributions in the performance appraisal process is fundamental to better clarify the relations between performance management and positive individual outcomes. Therefore, this study aims to investigate the effects of perceived performance appraisal justice on teachers' well-being, in terms of job performance, job satisfaction, and life satisfaction, hypothesizing the mediation role of performance appraisal satisfaction. Data from a sample of Italian teachers were analyzed through structural equation modeling. Results confirm the mediation role of performance appraisal satisfaction. In particular, perceived performance appraisal justice was positively associated to performance appraisal satisfaction, which, in turn, was positively associated with job performance, job satisfaction, life satisfaction. Consequently, performance appraisal satisfaction totally mediated the relations between performance appraisal justice and the outcomes considered. Findings are relevant for two reasons. First, they contribute to better understanding the performance management process in educational settings – an issue requiring further attention. Second, they contribute to highlighting the importance of performance management efficacy, which is essential not only to improve individual well-being but also to enhance teaching quality.

Keywords: performance appraisal satisfaction, performance appraisal justice, teacher performance, teacher job satisfaction, teacher life satisfaction

# INTRODUCTION

Performance appraisal is one of the most important HR management tools and its efficient implementation is one of the greatest HR professionals' challenges, particularly in terms of validity and reliability (Gupta and Kumar, 2013; Ivaldi et al., 2015). Performance appraisal identifies the individual's contribution to the organizational goals and establishes individual performance standards (Ikramullah et al., 2012). It can become a real job resource (Farndale, 2017). Performance appraisal is a formal assessment of what employees have performed (Snape et al., 1998). Its ultimate purpose is to allow employees to continue to improve their job performance (Selvarajan et al., 2018) and teaching

Edited by:

Caterina Fiorilli, LUMSA University, Italy

#### Reviewed by:

Peizhen Sun, Jiangsu Normal University, China Federica de Cordova, University of Verona, Italy

> \*Correspondence: Laura Dal Corso dalcorso@unipd.it

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 07 June 2019 Accepted: 03 October 2019 Published: 26 November 2019

#### Citation:

Dal Corso L, De Carlo A, Carluccio F, Girardi D and Falco A (2019) An Opportunity to Grow or a Label? Performance Appraisal Justice and Performance Appraisal Satisfaction to Increase Teachers' Well-Being. Front. Psychol. 10:2361. doi: 10.3389/fpsyg.2019.02361

innovation (Benadusi and Giancola, 2016). Performance appraisal also has specific aims, such as accountability, professional development (Delvaux et al., 2013), and organizational growth (Rubel and Kee, 2015).

Many changes have influenced European educational systems since the '70s, in terms of greater school autonomy and human resources management responsibility (Benadusi and Giancola, 2016). An efficient performance appraisal system has important positive effects on teachers' professional development. Moreover, teachers are essential for better education and future workers' growth (Ripamonti et al., 2018; Tripathi et al., 2018). Teaching is a profession with a strong sense of meaning and charged with civic responsibility: the importance of its quality is evident. For example, it allows students to emancipate from their families, internalize norms and values, and be recognized for the goals achieved (Freddano, 2016). Therefore, performance appraisal is a fundamental tool to improve not only teaching quality but also school quality. It identifies teacher's development and training needs, while boosting various outcomes (Obasi and Ohia, 2014). For example, Robinson et al. (2008) stated that teachers' professional development indirectly affects students' outcomes. Teachers' performance appraisal is a delicate process. On the one hand, it greatly influences the well-being of teachers (Borrelli et al., 2014; Benevene et al., 2018b) who can experience anxiety and pressure due to the evaluation (Benevene and Fiorilli, 2015; Girardi et al., 2015, 2018; Falco et al., 2017; Cuevas et al., 2018). On the other, many difficulties may arise, such as lack of time, lack of confidence, or lack of training (Donaldson and Mavrogordato, 2018). Since well-being is conceptualized as a self-realization, as a social integration and as positive orientation toward the task, job satisfaction, life satisfaction and job performance are often used as its indicators (Alonso et al., 2019). Teachers' performance appraisal aims to be an objective system evaluating teachers and teaching through a supervisors' analysis. Principals can make use of other assessments – for example by considering colleagues and students' opinions or the teachers' self-evaluation (Bradford and Braaten, 2018). Thus, the process will not be top-down, but accomodating and participatory, and will be free to make use of scientific methodologies (such as action-research), to establish an example of good practice for the whole school community (Freddano, 2016).

This study focuses on individual perceptions of performance appraisal: although the entire process conveys contextual factors, considering its subjective elements is of utmost importance (Kim, 2016). Reactions, perceptions, and attributions teachers make about the judgments received – in terms of fairness and satisfaction – influence their outcomes. The paper aims to bring the following contributions, based on the little evidence on performance appraisal justice (Rubel and Kee, 2015). First, its mechanism of action – e.g., mediation effects (Gupta and Kumar, 2013) – is not clear. Therefore, we aim to investigate potential mediations in the relations between performance appraisal justice and its outcomes. Second, performance management literature highlights a gap in both organizational and individual performance appraisal positive outcomes (Van De Voorde et al., 2012). Finally, we aim to clarify the relations between performance appraisal perceptions – in terms of justice and satisfaction – and some well-being dimensions, namely job satisfaction, life satisfaction, and job performance.

This study considers positive dimensions and analyzes how to enhance well-being. We draw on positive psychology framework (Seligman and Csikszentmihalyi, 2000). This movement aims to clarify and promote optimal functioning, by amplifying strengths and encouraging global well-being (Shankland and Rosset, 2017). Traditionally, psychology has focused on disease. Positive psychology is not only a research topic; it looks to goodness, both in people and in contexts (Ciarrochi et al., 2016). Emotions and positive feelings lead people to broaden and build themselves (Fredrickson, 2001), whereas their lack – or the presence of negative states – leads to failures and unhappiness. Positive organizational interventions aim to take advantage of strengths and focus on the brighter side of situations (Ghosh and Deb, 2017). Performance appraisal, if perceived as fair, allows teachers to be more flexible and efficient, and to experience the vitality promoted by positive psychology. This process is possible if the organizational interventions follow some values. Among them, Ciarrochi et al. (2016) identified self-challenge and continuous learning. We can connect the enhancement promoted by positive psychology with an important theoretical framework, the conservations of resources (COR) theory (Hobfoll, 1989, 1998). COR theory states that people are motivated to gain, maintain and nurture resources. This resource enrichment positively influences people's well-being, whereas resource loss leads them to distress. Resources are all organizational aspects that can stimulate personal growth, learning and development (Demerouti et al., 2001), so performance appraisal justice and performance appraisal satisfaction are part of them. Therefore, based on COR theory, we wonder whether fair and satisfying performance appraisal processes can originate and enhance positive outcomes.

Among performance appraisal attributions made by workers, fairness is very important, because it substantially directs several outcomes, such as psychological contract (Barbieri et al., 2018) and efficiency (Selvarajan et al., 2018). Clear, rational, and univocal criteria allow conclusion of an efficient performance appraisal (Longenecker, 1997). Moreover, a performance appraisal perceived like a criticism – rather than an instrument for professional development – can determine teachers' attitudes of closure (Lucisano and Corsini, 2015). Fairness regarding performance appraisal is part of organizational justice, an overarching variable formed by various sub-dimensions: distributional justice, procedural justice, and interactional justice; with the latter having been split into interpersonal and informational justice (Greenberg, 1993). According to Colquitt et al. (2001), distributional justice depends on the comparison of efforts made, rewards received, and colleagues' rewards. Procedural justice derives from the evaluation of the processes and policies used in performance appraisal. Interpersonal justice looks to perceived dignity and respect during the feedback. Informational justice regards the information obtained on process management. Performance management quality is positively associated with positive outcomes, such as commitment and intention to remain, and is negatively associated with negative outcomes, like job

stress (Su and Baird, 2017; Falco et al., 2018). In particular, a performance appraisal perceived as partial – e.g., made by biased appraisers – can originate negative consequences for the teacher. Thus, performance appraisal justice is fundamental, because it leads to various positive outcomes, such as engagement (Gupta and Kumar, 2013; Farndale, 2017), motivation to improve job performance (Selvarajan et al., 2018), pay-for-performance effectiveness (Kim, 2016), organizational commitment (Guchait and Cho, 2010), perceived organizational support (Farndale, 2017), and decreased turnover intentions (Rubel and Kee, 2015).

Satisfaction with performance appraisal is another key dimension to analyze subjective responses to performance appraisal. It is a global evaluation of the performance appraisal received, and it involves perceptions regarding one's participation to the evaluation process, the feedback received, and its consequences on rewards distribution. Satisfaction with performance appraisal concerns various facets, such as appraiser's trust and feedback utility (Delvaux et al., 2013). It can influence attitudes and behaviors toward the organization (Hong, 2018). When performance appraisal is congruous with individual efforts, the process is satisfying (Gözükara et al., 2017). Performance appraisal satisfaction stimulates the acknowledgment and the use of the process itself; on the contrary, a lack of performance appraisal satisfaction causes negative consequences, such as intention to quit (Guchait and Cho, 2010), work-family conflict (Ismail and Gali, 2017), and strain (Van Thielen et al., 2018). Fairness perceived on performance appraisal is fundamental to its satisfaction (Naji et al., 2015): some evidence state that performance appraisal justice is an antecedent of performance appraisal satisfaction (Lira et al., 2016; Hong, 2018). Therefore, we hypothesize that:

#### H1: performance appraisal justice is positively associated with performance appraisal satisfaction.

The relation between performance appraisal justice and its possible outcomes deserves an in-depth analysis. For example, the recognition of the individual contribution – also achievable through a fair performance appraisal – contributes to organizational productivity, while inadequate performance management practices lead to great productivity losses (Bloom et al., 2014). The more the performance appraisal is unfair, the less the feedback will be useful to improve job performance (Naji et al., 2015). On the contrary, fair performance appraisal improves job performance (DeNisi and Pritchard, 2006; Gruman and Sacks, 2011). Job performance is considered as a behavior completely under the control of the individual, an act of doing a job and a means to reach a set of goals within a job (Campbell, 1990). The lack of openness, legitimation and integrity, and the perception of favoritisms and biased evaluations lead to a gap between ideal job performance and real job performance (Cunha et al., 2018): employees' resources are depleted by these negative perceptions and are not allocated to job performance anymore (Falco et al., 2013b). It is important to shed light on the mechanism by which performance appraisal justice influences outcomes, with particular reference to mediated relations (Gruman and Sacks, 2011; Gupta and Kumar, 2013; Van Thielen et al., 2018). Considering job performance, Malik and Aslam (2013) pointed out that great performance appraisal satisfaction activates a mechanism that increases job performance. Selvarajan et al. (2018) connected performance appraisal justice to job performance through the motivation to perform better. Drawing on this evidence and considering the positive relation between performance appraisal justice and job performance (Kuvaas, 2006; Aly and El-Shanawany, 2016), we would like to clarify the role of performance appraisal satisfaction in the relation between performance appraisal justice and job performance. Therefore, we assume that:

#### H2a: performance appraisal justice is positively associated with job performance.

#### H2b: performance appraisal justice is positively associated with job performance through performance appraisal satisfaction.

Job satisfaction is a key-factor for organizational success (Mufti et al., 2019). It is positively associated with goals achievement (Malik et al., 2010). This pleasant emotional state derives from individual evaluations about the job. It corresponds to the pleasantness perceived on the job and motivates people to be committed to their job activities (Karimi et al., 2011; Bélanger et al., 2015; Benevene et al., 2018a). Evidence showed the association between performance appraisal justice and job satisfaction (Salleh et al., 2013; Agyare et al., 2016; Van Thielen et al., 2018). Since job satisfaction is a global evaluation of various job features, it can be affected by the performance appraisal system. For example, in the literature abundant evidence have demonstrated the positive relation between performance appraisal satisfaction and job satisfaction (Karimi et al., 2011; Decramer et al., 2015; Van Thielen et al., 2018). Therefore, we assume that:

#### H3a: performance appraisal justice is positively associated with job satisfaction.

#### H3b: performance appraisal justice is positively associated with job satisfaction through performance appraisal satisfaction.

Looking to vocational outcomes without considering personal well-being could be a low-value choice (Van De Voorde et al., 2012). Work context and life domain are not distinct spheres: they are interdependent and could influence each other in many ways (Chummar et al., 2019). Then, we would like to take a further step: to explore the relation between performance appraisal justice and life satisfaction, which is often used as an indicator of well-being (Alonso et al., 2019). Life satisfaction is a cognitive evaluation regarding how satisfying our entire life is (Hart, 1999). Essentially, how much our life quality pleases us. Many studies stated the association between life satisfaction and job satisfaction (Aydintan and Koç, 2016; Goetz et al., 2019; Masdonati et al., 2019), that is associated with performance appraisal justice, as mentioned above. Moreover, organizational justice – the overarching dimension in which performance appraisal justice belongs – is associated with life satisfaction (Tepper, 2000; Lambert et al., 2010). Performance management can influence personal domain, as well. Dissatisfaction with

performance appraisal has negative effects on personal life: for example, it increases work-life conflict (Falco et al., 2013a; Ismail and Gali, 2017; De Carlo et al., 2019). However, to the best of our knowledge, the literature has not examined the relation between performance appraisal system and life satisfaction. We suppose it can follow the performance appraisal satisfaction – job satisfaction relation. Therefore, we would like to explore the role of performance appraisal satisfaction in the relation between performance appraisal justice and life satisfaction, assuming that:

H4a: performance appraisal justice is positively associated with life satisfaction.

H4b: performance appraisal justice is positively associated with life satisfaction through performance appraisal satisfaction.

### MATERIALS AND METHODS

#### Participants and Procedure

Participants were directly contacted and decided to take part in the research on voluntary basis. Consequently, we gathered a convenience sample. One hundred sixty-one Italian teachers filled out the paper-and-pencil self-administered questionnaire (**Table 1**). The sample average age was 46.3 years (SD = 10.11) and gender distribution was 63.4% women and 36.6% men. With regards to educational levels, 89.4% of participants held a degree, 5.6% held a high school diploma, 0.6% completed only middle school, and 4.3% held a further kind. Most of the participants had been working for the same school for more than 4 years (74.5%), 22.4% had been working for a period ranging between 2 and 4 years, and 3.1% had been working for less than 1 year. The


majority of the sample is full-time employed (88.2%), whereas 10.6% is part-time employed (1.2% missing). 90.7% had an openended contract (or apprenticeship) and 9.3% had a fixed-term contract (or replacement). All participants gave their written informed consent before the administration of the questionnaire, in accordance with the Declaration of Helsinki. The study was carried out in accordance the rules of AIP (Associazione Italiana di Psicologia – Italian Association of Psychology), according to which there was no need for previous ethics approval, since it would not deal with animals or vulnerable groups, or would involve risk for the well-being of participants, or use biomedical devices, or invasive investigation tools. Our study was conducted in accordance with the recommendations of the Ethic Committee of Psychology Research of the University of Padua, with the above-mentioned written informed consent from all participants.

#### Measures

Performance appraisal justice was assessed with 17 items (Gupta and Kumar, 2013) adapted from Colquitt et al. (2001). The scale measured four dimensions: distributive justice (e.g., "The outcome of performance appraisal is appropriate for the work I completed"), procedural justice (e.g., "The procedures followed during performance appraisal process are free of bias"), interpersonal justice (e.g., "My supervisor treated me with dignity during the performance appraisal meeting"), and informational justice (e.g., "My supervisor explained the procedures of the performance appraisal process thoroughly"). The 5-point response scale ranged from 1 (strongly disagree) to 5 (strongly agree). The Cronbach's alpha for the scale is 0.95.

Performance appraisal satisfaction was assessed with six items (Kuvaas, 2006): global satisfaction with performance appraisal, satisfaction with the feedback received, and perceived organizational commitment to performance appraisal (e.g., "I am satisfied with the way my organization provides me with feedback," "The feedback I receive on how I do my job is highly relevant," "I think that my organization attempts to conduct performance appraisal the best possible way," respectively). The items were on a 5 point-Likert scale (1 = strongly disagree; 5 = strongly agree). The Cronbach's alpha for the scale is 0.92.

Job performance was assessed with two items rated on a 10 point Likert scale (from 10% to 100%): "We would kindly ask you to specify, using a percentage, how successful you were in reaching your work goals last year" and "How do you rate your job performance during the last year?" The Cronbach's alpha for the scale is 0.85.

Job satisfaction was assessed by asking "How satisfied are you with your working life?" The item was taken from the Qu-Bo test (De Carlo et al., 2008–2011) and was on a 6 point-Likert scale (1 = very dissatisfied; 6 = very satisfied).

Life satisfaction was assessed by asking "How satisfied are you with your overall life?" The item was taken from the Qu-Bo test (De Carlo et al., 2008–2011) and was on a 6 point-Likert scale (1 = very dissatisfied; 6 = very satisfied).

#### Statistical Analyses

We tested the hypotheses by means of structural equation models (SEM) with latent variables, using the Lisrel 8.80 software

(Jöreksog and Sörbom, 2006). Consequently, we estimated another model, fixing all the non-significant paths to zero, to obtain a more parsimonious solution.

We assessed the model fit, starting with the chi-square test (χ 2 ). A model shows a good fit to the data if χ 2 is non-significant. Given that χ 2 is sensitive to sample size, we considered additional fit indices. In particular, we considered the nonnormed fit index (NNFI), the standardized root mean residual (SRMR), and the root mean square error of approximation (RMSEA). Values close to or greater than 0.95 for NNFI, values close to or smaller than 0.10 for SRMR, and values close to or smaller than 0.08 for RMSEA indicate an acceptable fit (Schermelleh-Engel et al., 2003).

We considered 95% asymmetric confidence intervals based on the distribution of the multiplication term, to verify the significance of the indirect effects. The purpose was to manage the non-normality derived from the path a × path b multiplication, as recommended by MacKinnon's procedure (PRODCLIN; MacKinnon et al., 2007). If the confidence interval does not contain zero, the indirect effect is significant (MacKinnon et al., 2012).

Before testing the model, we carried out two procedures. First, we excluded participants with missing values; therefore, the final sample consisted of 154 participants. Second, we tested if common method variance (CMV) was a threat to the study. In fact, as we collected data through self-report measures, the risk of CMV may exist. Therefore, we controlled for the effects of a latent method factor, by using a single-commonmethod-factor approach in a confirmatory factor analysis (CFA) (Podsakoff et al., 2003). We added a new latent variable called "method" on which we loaded all indicators of the five theoretical factors. Consequently, we obtained a six-factor model. Then, we compared χ <sup>2</sup> of the six-factor model to the five-factor model's one. If the p-value associated with 1χ 2 is significant, the effect of the latent method factor exists. To evaluate the impact of this effect, we partitioned the observed variance of the indicators into three component: variance attributable to the theoretical factors, to the method and to the causal error. In particular, Williams et al. (1989) identified the following average partitioning: 50% variance attributable to the theoretical factors, 27% variance attributable to the method, and 23% variance attributable to the causal error. If the variance attributable to the method is up to 27%, the CMV does not lead to inaccurate results.

# RESULTS

#### Descriptive Results

**Table 2** presents means, standard deviations, and correlations. The latter provide initial evidence that all variables could be positively associated to each other.

#### Common Method Variance

The comparison between the χ <sup>2</sup> of the two alternative models suggests the possible existence of a method effect (1χ <sup>2</sup> = 88.45; 1df = 14; p = 0.00). Even though the variance partitioning indicated that the variance attributable to the method exists, this is limited and accounts for 17% of the total observed variance. Therefore, these results suggest that CMV is not a concern in this study.

### Model Testing

We estimated the hypothesized structural model that satisfied the acceptability criteria [χ 2 (69) = 150.83, p = 0.00; NNFI = 0.96; SRMR = 0.05; RMSEA = 0.08]. However, the direct relations between performance appraisal justice and the outcomes – namely job performance, job satisfaction, and life satisfaction – were non-significant (H2a, H3a, H4a rejected). To obtain a more parsimonious solution, we then set all these paths to zero.

This second model tests the total mediation of performance appraisal satisfaction. The model (**Figure 1**) shows an acceptable fit to the data, considering χ 2 (72) = 152.97 (p = 0.00); RMSEA = 0.08; NNFI = 0.97; and SRMR = 0.05. In the model, performance appraisal justice is positively associated with performance appraisal satisfaction (γ = 0.82, p < 0.001). Therefore, H1 is confirmed. Furthermore, performance appraisal satisfaction is positively associated with job performance (β = 0.54, p < 0.001), job satisfaction (β = 0.30, p = 0.001), and life satisfaction (β = 0.32, p < 0.001).

At this point, we verified the significance of the indirect effects. The asymmetric confidence intervals for the relationships between performance appraisal justice and the outcomes, through performance appraisal satisfaction, do not contain zero. We can conclude that performance appraisal satisfaction totally mediates the relationship between performance appraisal justice and the outcomes (H2b, H3b, H4b confirmed). In particular, the unconventional estimate is 0.65, 95% CI [0.41028,0.91924] for the relationship between performance appraisal justice and job performance. The unconventional estimate is 0.34, 95% CI [0.15612,0.55012] for the relationship between performance appraisal justice and job satisfaction. The unconventional estimate is 0.38, 95% CI [0.18804,0.58972] for the relationship between performance appraisal justice and life satisfaction. Consequently, we can conclude that performance appraisal satisfaction totally mediates the relationship between performance appraisal justice and the outcomes.

# DISCUSSION

The study had the following aims. First, we intended to investigate the role of performance appraisal perceptions in a teachers' sample – in terms of justice and satisfaction – in enhancing well-being outcomes. Second, we aimed to clarify its influence mechanisms on the outcomes, particularly mediation.

High-quality teaching allows schools to focus on students' skills and knowledge, to educate future citizens, and promote social inclusion (Barone, 2016). Performance appraisal systems permit not only to regulate educational system, but also to create continuous innovation and improvement (Freddano, 2016). Performance appraisal is a process critical to teaching quality: if perceived as inadequate, it has negative consequences. The performance appraisal should not be a mere red-tape fulfilment: it should be carried out in depth, to actually influence school


N = 161, ∗∗p < 0.01.

functioning (Barone, 2016). According to Cunha et al. (2018) some believe that it would be better to eliminate this appraisal system, because of its weaknesses, such as the excessive focus on the dyadic and unbalanced relation between appraiser and appraisee, and the thorny reference to meritocracy. Furthermore, performance management system is inefficient if openness is lacking and integrity is not perceived. For this reason, perceiving justice in performance appraisal is a key factor to enhance positive outcomes.

Our findings confirm the importance of positive perceptions in performance appraisal. Indeed, performance appraisal justice enhances well-being outcomes, namely job performance, job satisfaction, and life satisfaction. However, this improvement is possible only through satisfaction with performance appraisal, because our findings show that the direct relationships between performance appraisal justice and the outcomes are nonsignificant. Satisfaction with performance appraisal is needed, because it totally mediates the relations between performance appraisal justice and the outcomes. Finally, our study confirms that performance appraisal justice and performance appraisal satisfaction affect job performance (Gruman and Sacks, 2011; Malik and Aslam, 2013) and job satisfaction (Decramer et al., 2015; Agyare et al., 2016). Moreover, we argue that performance appraisal justice and performance appraisal satisfaction have non-vocational outcomes, namely life satisfaction.

Therefore, performance appraisal appropriateness, unbiased procedures, dignity during performance appraisal meetings, and adequate explanation of the procedures used – namely, performance appraisal justice (Gupta and Kumar, 2013) – is not sufficient to increase the positive outcomes. Performance appraisal satisfaction is fundamental to reach this goal. For this reason, performance appraisal has to be perceived as a global, positive process: its perceived justice is important together with its other facets, such as the relevance of the feedback received, the right recognition of the individual performance, and the organizational engagement in providing a constructive feedback – aspects constituting performance appraisal satisfaction (Kuvaas, 2006). Thus, according to our results, all the aspects of performance appraisal satisfaction convey performance appraisal justice.

The study has some limitations. First, its cross-sectional nature does not allow us to determine the direction and the causality of the relations. Even if strong reasons support our results, future research may choose to conduct longitudinal studies. Since we used only one data gathering method, future research could take into account different kinds of evaluations (Falco et al., 2013c; Girardi et al., 2019); for example, they can consider students' and colleagues' perspectives. Drawing on the importance of subjectivity, future research could consider the moderation of some individual characteristics, such as optimism or self-efficacy, or some personality traits (e.g., perfectionism, need for cognitive closure; Falco et al., 2014; Bélanger et al., 2016). Moreover, since contextual variables could explain the results trend, future research could take into account the school level

as an independent variable. Finally, in the future it is worth examining how to promote performance appraisal justice, to make it an integral part of the educational system. It is also essential to reflect on the real purposes of the appraisal system (Maccarini, 2016), that is to adjust to match those of the educational system, and the overall quality-equity binomial (Benadusi and Giancola, 2016).

# CONCLUSION

The study underlines the relationships between performance appraisal justice and some positive individual outcomes (e.g., performance, job satisfaction, life satisfaction). These relationships are totally mediated by performance appraisal satisfaction. The study gives the following contributions. First, in examining the performance appraisal system, it considers its perceptions and its subjective facets, rather than the organizational ones. In doing so, it sheds light on the relationships between these perceptions and the positive outcomes considered. Second, it clarifies the mechanisms of action of performance appraisal justice, which was not clear (Rubel and Kee, 2015). Third, it identifies a possible antecedent of performance appraisal satisfaction (Ismail and Gali, 2017). Fourth, it represents – to the best of our knowledge – a first step in the study of the effects of the performance appraisal system perception on life satisfaction.

### PRACTICAL IMPLICATIONS

Based on our results, justice and satisfaction are basic facets of the performance appraisal system. Therefore, planning training activities would be advisable to strengthen the skills that

# REFERENCES


principals use during these performance appraisal meetings. The interventions could concern communication, soft, and positive, managerial skills, and psychological counseling (Dal Corso et al., 2013; Scaratti and Ivaldi, 2015; De Carlo et al., 2016; Farnese et al., 2017). These competences could help teachers to perceive performance appraisal as a constructive, encouraging process.

# DATA AVAILABILITY STATEMENT

The datasets generated for this study are available on request to the corresponding author.

# ETHICS STATEMENT

Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. The patients/participants provided their written informed consent to participate in this study.

# AUTHOR CONTRIBUTIONS

LD developed the research project, with the contribution of AD, AF, and DG. LD reviewed the literature, with the contribution of AD and AF. FC prepared the data set. DG and FC carried out the data analysis.

# FUNDING

This study was supported by the University of Padua Research Funds.

and subordinates' need for cognitive closure on burnout and stress. Eur. J. Work Organ. Psychol. 25, 287–300. doi: 10.1080/1359432X.2015.1061999




recommended remedies. J. Appl. Psychol. 88, 879–903. doi: 10.1037/0021-9010. 88.5.879


**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The handling Editor declared a shared affiliation, though no other collaboration, with one of the authors FC at the time of review.

Copyright © 2019 Dal Corso, De Carlo, Carluccio, Girardi and Falco. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Reflective Practice: A Method to Improve Teachers' Well-Being. A Longitudinal Training in Early Childhood Education and Care Centers

#### *Ada Cigala1 \*, Elena Venturelli1 and Martina Bassetti1,2*

*1 Department of Humanities, Social Sciences and Cultural Industries, University of Parma, Parma, Italy, 2 PRO.GES. TRENTO s.c.s. Cooperativa Sociale, ONLUS, Trento, Italy*

#### *Edited by:*

*Wong Yau Ho Paul, Tung Wah College, Hong Kong*

#### *Reviewed by:*

*Po Lin Chan, The Education University of Hong Kong, Hong Kong Po Lin Becky Lau, The Education University of Hong Kong, Hong Kong*

> *\*Correspondence: Ada Cigala ada.cigala@unipr.it*

#### *Specialty section:*

*This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology*

*Received: 09 July 2019 Accepted: 30 October 2019 Published: 26 November 2019*

#### *Citation:*

*Cigala A, Venturelli E and Bassetti M (2019) Reflective Practice: A Method to Improve Teachers' Well-Being. A Longitudinal Training in Early Childhood Education and Care Centers. Front. Psychol. 10:2574. doi: 10.3389/fpsyg.2019.02574*

Various studies focused on educational contexts (0–6 years) point out that early childhood multi-age classrooms provide better learning strategies and socio-emotional competences of children, compared to single-grade classrooms. However, these studies have also shed light on the significant role of teachers. The multi-age classroom in particular is an opportunity for child development, provided that teachers consider problem-solving, flexibility, and co-construction as effective education strategies. Starting from these reflections, this study aimed to verify the efficacy of longitudinal training for the purpose of advancing the perceived well-being of early childhood teachers of multi-age groupings (18–54 months). Eight teachers and one pedagogical coordinator of an Italian Early Childhood Education and Care center took part in the study. All the participants were females. The critical aspect identified by the teachers was the multi-age classroom, which was perceived as making teaching and learning very difficult and ineffective for both themselves and for the children. The training lasted 10 months and implied a methodology focused on *observations* of some activities and *reflective practice* in the group that concerned both teachers and the pedagogical coordinator. The training involved the drafting of two types of written protocols: the observational reports of the specific activities observed (20), and the descriptive reports of reflective sessions (6). The content analysis of the reports revealed various and interesting themes regarding the teachers' perceived well-being, in terms of thoughts, behaviors, and feelings. The qualitative and longitudinal analysis of the themes that emerged in these protocols highlighted different processes of change in the teachers' perception, in particular with respect to three specific dimensions of well-being: *sense of belonging*, *self-efficacy*, and *agency*. At the end of the training, the teachers experienced a greater sense of belonging to the group of colleagues, a greater sense of self-efficacy, and an idea of themselves as active and meaningful participants. These results supported different reflections regarding the modalities through which to enhance the perceived well-being of teachers.

Keywords: well-being, preschool teachers, longitudinal training, self-efficacy, sense of belonging, agency, multi-age classroom, reflective practice

# INTRODUCTION

In recent years from many sides, a number of efforts have emerged to increase children's access to high-quality Early Childcare and Education Centers (ECEC), but various authors argue that some attention and study must be given also to the teachers and to their workplaces, with particular focus on their perceived well-being in these contexts (Roffey, 2012; Zinsser and Zinsser, 2016). Indeed, some recent studies have pointed out that many preschool teachers report serious psychological health problems (e.g., Whitaker et al., 2013). It is indeed stressful and tiring work both from a physical and psychological point of view, having to deal with the care and education of very young children, some even just a few months old.

Moreover, the literature on the teachers' well-being in the school context, and in particular in the Early Childcare and Education Centers, shows a rather fragmented framework of data and empirical models (Hall-Kenyon et al., 2014; De Stasio et al., 2017; Benevene et al., 2018). The studies specifically show that the well-being construct has been described in different terms by different authors (Bowling et al., 2010; Hall-Kenyon et al., 2014). It can be noted that most of the research had focused on the individual dimensions that contribute to the teachers' well-being, such as emotional competence, temperament, self-esteem, educational qualifications, and years of professional experience (Hall-Kenyon et al., 2014; De Stasio et al., 2017; Benevene et al., 2018).

Fewer authors have proposed to analyze teachers' well-being addressing the social or systemic dimensions. In the theoretical framework of these studies, teachers' well-being is conceptualized as an emerging quality of a complex of interactions between various factors, such as relationships with children, between colleagues, with managers, and with parents, as well as the interactions with the elements of the space of the school context (Holmes, 2005; Critchley and Gibbs, 2012; Roffey, 2012; Zinsser and Zinsser, 2016).

These studies, albeit not numerous, appear to be very interesting, as they have a high ecological value (Roffey, 2012); central to these contributions is the concept of "*workplace environment*" that includes different aspects (spatial, psychological, and behavioral) interacting reciprocally and taking on several configurations (Zinsser and Zinsser, 2016).

An in-depth analysis of these studies allows us to highlight how the authors in the various studies focused on different aspects of the workplace, proposing specific and different constructs. For example, some studies have shown that there is a close positive relationship between the teachers' well-being and the perception of *internal social support*, defined as the possibility for teachers to experiment with spaces of constructive sharing between colleagues and with the manager that are aimed to identify functional strategies to deal with complex and difficult professional situations (Halbesleben, 2006; Karademas, 2006; Sanglim and Sungeun, 2016).

In other studies, instead, the concept of *psychological safety of the environment* was introduced, understood as the perception of the work group as a context in which there is respect, acceptance, care, and mutual professional trust. On a personal level, a school system perceived as psychologically secure gives teachers the experience of feeling valued, respected, and cared for (Edmondson, 1999; Dollard and Bakker, 2010; Roffey, 2012; Edmondson and Lei, 2014).

According to Zinsser and Zinsser (2016), "One critical component of the supportive workplace climates is the construction of a workplace environment that teachers perceive as psychologically safe and in which they feel capable of engaging in the challenging work of early childhood education" (p. 49).

Still within the research overview on the teachers' well-being, other authors have introduced the concept of *sense of belonging* to the school or, in particular, to the group of colleagues, which has been declined as the perception of having a precise role within the system, a role that is also recognized by other group members. On a personal level, the sense of belonging for an individual corresponds to feeling acknowledged, valued, and included (Rowe et al., 2007; Wike and Fraser, 2009; Skaalvik and Skaalvik, 2010; Roffey, 2012).

This sense of belonging, expressed by some authors as *connectedness* (Rowe et al., 2007) or *sense of community* (McGinty et al., 2008; Hall-Kenyon et al., 2014), is deeply related to the quality of interactions, in particular when the teachers in the group of colleagues feel that trust and reciprocity predominate, that they are positively connected with each other, and that there is a greater chance to perceive well-being.

Finally, still within the relational-systemic approach to the study of the teachers' well-being, the construct of *agency* appears to be extremely relevant. The perceived agency refers to the perception by teachers of the possibility to participate in organizational decision-making (Sanglim and Sungeun, 2016) and to have an active role within the school system, a role that is recognized by the whole group of colleagues (Wilson and Deaney, 2010; Priestley et al., 2013; Hadar and Benish-Weisman, 2019). Specifically, the agentic capacity refers to the concept that teachers do not simply repeat given practices, rather, they have a capacity for pro-active and autonomous actions, that favor important transformations and changes both in themselves and in the whole workplace environment (Hadar and Benish-Weisman, 2019). For Wilson and Deaney (2010), agency can be considered a combination of intention and action that influences experience. Several studies have investigated the relationship between the teachers' perceived agency and the perception of their well-being within the workplace, highlighting interesting connections (Helgeson, 1994; Buchanan and Bardi, 2015; Hadar and Benish-Weisman, 2019).

In all these studies, the authors emphasized the need to orient future research toward teachers' well-being in a more ecological direction, favoring *qualitative*, *situational*, and *longitudinal methodologies,* which allow us to analyze the processual and interactive dimensions of teachers' well-being. These indications address the research toward studies focused on specific situations considered particularly challenging from the point of view of the school system; in this sense, the case studies seem to be a suitable method (Critchley and Gibbs, 2012; Roffey, 2012; Cook et al., 2016; Hadar and Benish-Weisman, 2019; Heyeres et al., 2019).

In accordance with this perspective, the present researchintervention was designed to study the well-being of preschooler teachers, conceptualized, not so much as a construct in general, but rather as well-being perceived in a specific situation considered critical and problematic from the teachers' point of view. In particular, the study involves pre-school teachers engaged in the management of a multi-age group (18–54 months) that include children from two different classes.

It is important to underline that the incentive to create multi-age classes is part of a general orientation in Italian Early Childcare and Education Centers toward the construction of an integrated 0–6 years system. The Italian childcare services are today mostly structured according to a split 0–3/3–6 years system, but recently, there has been a move toward the construction of an integrated system according to a recent law (Law no. 107/2015, art. 1, paragraph 181, letter e; legislative decrees 2017).

The literature on multi-age classrooms in the pre-school age highlights interesting aspects (Maeda, 1994; Veenman, 1996; Dersheid, 1997; McClellan and Kinsey, 1999; Aina, 2001; Gerard, 2005; Logue, 2006; Quann and Wien, 2006; Edwards et al., 2009).

Among the various studies on the topic, the contribution of Edwards et al. (2009) appears to be very interesting; it is a longitudinal study conducted in the preschool educational services in Australia for a period of 11 months. In this study, the teachers' point of view with respect to the multi-age classes was surveyed by analyzing their ideas, their speech, and their practices in the workplace context.

In general, the results of these studies show that multi-age classes foster children's cognitive and social learning. In particular, the multi-age classes are contexts that promote language acquisition, problem-solving skills, active participation in the learning process (Maeda, 1994; Chapman, 1995), the positive management of peer relations, positive conflict resolution, and greater attention to individual differences both by teachers and children (Aina, 2001; Baumgartner and Bombi, 2005; Edwards et al., 2009).

But the most interesting aspect that, in different ways, seems to emerge from the research, is that in the preschool period the multi-age class context fosters child development only if the teachers share some aspects of the educational approach. Specifically, the multi-age class represents a developmental context when the teachers consider: (1) differences between children as a resource for all, (2) problem-solving as an important learning strategy both for themselves and for the children, and (3) flexibility, planning, and continuous communication as essential skills for their educational work (Aina, 2001; Edwards et al., 2009).

Research shows that this educational approach adopted in the multi-age class context helps to improve teachers' perception of self-efficacy and the perception of their well-being in the workplace (Cook et al., 2016).

Based on these reflections and on the training experience of the authors as consultants in the field of the Early Childcare and Education Centers, an interesting research question arises: could the reflective practice in the work group enhance the teachers' perceived well-being?

Specifically, encouraged by the results of some previous studies (Karademas, 2006; Critchley and Gibbs, 2012; Zinsser and Zinsser, 2016; Venturelli and Cigala, 2017), we have identified an important research question: there are some relationships between the systematic reflective practice on some critical aspects of daily education strategies in class and some dimensions of teachers' perceived well-being. These include: the sense of belonging (Roffey, 2012), self-efficacy (Critchley and Gibbs, 2012), and agency (Hadar and Benish-Weisman, 2019). This specific relationship finds significance in the fact that the reflective practice adopted by a work group in a systematic and continuous way allows the teachers, on the one hand, to feel more actively involved with respect to work processes in recognizing their specific role, and on the other hand, the reflection within the group allows the teachers not to perceive themselves as isolated individuals facing difficulties, but rather as a "thinking group" that guarantees protection and safety. Consequently, the reflective practice also enhances a sense of competence in the teachers in identifying solutions and directions for a possible improvement in work practices.

Starting from this research question, this study aimed to verify the efficacy of longitudinal training for the purpose of advancing the perceived well-being of early childhood teachers of multi-age groupings (18–54 months).

In particular, the aim of the present study was to investigate, with a qualitative methodology, the possible changes that occur in the perceived well-being of teachers in the course of a training path focused on reflective practice, which is a method that promotes spaces for discussion and reflection among teachers on work-related practices (Schon and DeSanctis, 2011; Pollard, 2014). More specifically, in this study teachers were offered a training course that hinged mainly on those concrete aspects, inherent to work practices, considered by the teachers themselves to be critical and problematic, with respect to which they felt the need to identify new ways to manage them and to forge new directions of work (Venturelli and Cigala, 2017).

# MATERIALS AND METHODS

#### Participants

Eight teachers and one pedagogical coordinator of an Italian Early Childhood Education and Care Center took part in the study. All the participants were females, of Italian origin, aged between 28 and 55 (average age: 43) with at least 5 years' work experience and belonging to middle socioeconomic status families.

Four teachers belonged to the "young" children class made up of 20 children (from 18 months to 3 years). Four teachers belonged to the "big" children class, made up of 25 children (from 30 months to 5 years). The two classes were adjacent and placed in the same building, a nursery school building hosting a total of four classes. The training course is part of a three-year course which involved the participation of other schools. The training lasted 10 months, from September to June, and was conducted by two trainers, experts in the field of early childhood education.

Prior to the data collection, the parents' and the teachers' informed written consent was obtained, in compliance with the ethical norms defined by the *American Psychological Association.*

#### Training Method: Reflective Practice

The training methodology, already used in other similar paths (see Venturelli and Cigala, 2017), provided for different recursive phases:


*Emergent theme*: all of the teachers and the coordinator identified an emerging theme through a meeting in which a sharing of and a negotiation between the different teachers' points of view takes place. The emergent themes refer to certain aspects that concern the educational practices and that the work group considers to be critical, salient, and significant at that particular time. In a subsequent meeting the identified emergent theme was shared with the trainers that contributed to defining it in detail. These meetings took place in October.

*Observational reports*: all the teachers, the pedagogical coordinator, and one trainer carried out two observations focused on the emergent theme, and subsequently they had to draw up a written protocol (observational report) of each of these observations. The observations took place at two different times: T1 time at the beginning of the training (in November) and T2 time in the second part of training (in March). In each period, 10 observational reports were drafted for a total of 20 reports. The observational reports were shared among all teachers.

*Reflections on actions*: these meetings, which involved all the participants, were aimed at reflecting on the observed practices and at sharing the different points of view and meanings that emerged from the observations. These meetings followed the observation phases. Six reflections in action sessions were conducted, divided into two periods: three meetings were realized in the first part of the training after the first 10 observations (T1 time December–January) and the other three meetings took place in the final part of the training after the second observation period (T2 time April–May). All the meetings were video-recorded and transcribed verbatim by one of the two trainers (*descriptive reports*).

*Re-designing*: in the light of the reflections that emerged in previous meetings, the teachers were invited to meet in the colleague group in order to identify and explore alternative practices and re-design new ways of acting with respect to the topic considered critical (emerging topic). The re-design phases occurred in two periods: T1 time in February and T2 time in June; in T2 time in particular the redesign session was addressed to the future work of the teachers.

## Narrative Material Collected

The narrative material that documents the training process consists of the *observational reports of the specific activities* observed (no. 20) and the *descriptive reports of reflective meetings* (no. 6). As we explained before, these protocols are collected in the course of the training, in particular in two specific periods (T1 and T2).

The observational reports of the specific activities observed were drafted by all eight teachers, by the coordinator, and by a trainer. Each actor made two observations at different times, in November (T1) and in March (T2). The task given to the teachers was to describe what had been observed as objectively as possible; it should be pointed out that these teachers often used observational reports as a working method, so this request was very understandable for them.

The *descriptive reports of reflective meeting* concern the reports of the reflection meetings held in the presence of all the protagonists of the path, coordinator, teachers, and trainers and were carried out in two different periods of time along the training (T1: December–January and T2: in April–May). The second trainer drafted these reports and they consisted of verbatim transcriptions of the video-recording of the reflection meetings.

Two different qualitative analyzes were performed on the material collected:


#### Process Analysis of Reflective Practice Definition of the Emerging Theme

During this moment, the teachers were invited to reflect, and to propose an aspect that they experienced as being particularly critical on which they would have liked to turn their attention, and therefore on which to work within a specific path with the help of the trainers. The critical theme shared by the group was that of the *sleeping and relaxation moment* in the multi-age classroom. That is to say a moment shared by the two classes, small and big children, which provides for the collaboration by the teachers of the two classes and a common management of the space.

In particular, the analysis of the situation shows that, after a meal, both classes are offered a moment of relaxation, which takes place in a dedicated room equipped with mattresses or cots. Not all the children sleep (especially in the 3–6 age group) or at least not all at the same time, so in parallel an educational activity/play is proposed for children who stay awake. In this specific situation, the teachers' working group and the pedagogical coordinator, for about a year, started an inter-class project, focused on the creation of an educational space that can welcome both big and small children in a sharing of spaces and educational resources.

From the exchange of teachers' views, it emerged that this interclass sleeping and relaxation moment was recognized by teachers as a highly complex situation and was perceived as very tiring and difficult, both for the division of roles between teachers and for the organization of space. In the next meeting, the teachers shared with the trainers our analysis of the focused critical aspects; the discussion of different points of view, internal to the educational system (the teachers) and from outside the system (the trainers), enabled a more detailed analysis of the moment. In particular, the most critical aspect that emerged from the discussion was that the teachers in the *interclass sleeping and relaxation moment* having to take into consideration the different needs of children of different ages, who belong to two different classes that work separately during the rest of the day.

Finally, by means of the brainstorming technique, the group identified some questions to be highlighted through the observations, in order to improve their understanding of this moment and to better organize it: *What are the real needs of different children? How do we respond to all the different needs of children at that time? What is the role of the different adults present? How should it be coordinated in a more functional way?*

#### Observation and Reflections on Actions in Time T1

After identifying the emerging theme, the teachers, the pedagogical coordinator, and the trainer carried out the observations focused on the *sleeping and relaxation moment in the multi-age classroom*. In the subsequent reflective meetings, the trainers proposed three *key aspects* that emerged from the observations and that seemed particularly significant in relation to the critical points highlighted by the teachers: *the children's needs*, *the adult's role*, and *the intergroup system.*

#### *The Children's Needs*

From the teachers' observations, different needs emerged that the children expressed in the multi-age classroom at the sleeping and relaxation moment, such as: the need to sleep and relax in different ways, the need to play, and the need to share some stories and some readings with the adult. These needs were very different from each other, with major differences especially identified between small children (18 months-3 years) and the bigger ones (3–5 years).

#### *Role of Adult*

The analysis of observation through the lens of the adult's role allowed us to highlight different aspects. On the one hand, many teachers' movements appeared very dispersive and without a clear objective. Moreover, from the observational reports, it was highlighted that the teachers were an effective reference only for the children of their class, by whom they were recognized and to whom they proposed themselves as such. They did not represent a real point of reference for all the children in the class. Therefore, each teacher represented a resource for the other class only potentially but not at the level of practice. There was not a real availability of teachers for the whole inter-group. Finally, from the analysis of the observational report a good and consolidated coordination among the teachers of each class emerged.

#### *Interclass System*

As regards the lens of the interclass system, the analysis of the observational reports pointed out how the two classes, while occupying the same physical place, did not really seem integrated with each other. There were few contacts between the big and small children of the two classes, the teachers did not seem to perceive themselves as resources for the whole group, and there were many difficulties in fulfilling the different needs of big and small children, such as the need to relax and sleep, and the need to play and share.

A level of practice, and a clear and unique interclass project did not emerge; from the reports, it was noted that the two groups, although sharing the same physical space, seemed to be two different contiguous but unintegrated subsystems, with different times and routines.

Even the spatial organization of the room did not facilitate relationships between children, between adults, and between adults and children from different classes.

#### Re-design

In this phase of the training, the teachers were engaged in the re-design of the sleeping and relaxation moment of multi-age classroom starting from the reflections that emerged during the previous meetings. In particular, the trainers asked the group of teachers to *design*, *implement*, and *document* an interclass project linked to the sleeping and relaxation moment of the multi-age group, that contained a design shared and co-constructed by the teachers of both classes and the documentation of the project implementation.

Starting from this proposed task and from the new awareness acquired following the reflections shared in the previous meetings, the teachers agreed to work together to make the sleeping and relaxation place more familiar and more beautiful for all children and teachers.

The goal of the project was to allow all the teachers and all the children to experience this moment and the environment dedicated to this moment as being more familiar, so that everyone could feel a greater sense of belonging.

To do this, the teachers devised and planned several moments of inter-group activities that would involve all the children and all the teachers.

In particular, the project developed by the teachers included an initial circle time conversation in small groups with the children in order to highlight their experiences, thoughts, feelings about the sleeping and relaxation moment in the multi-age classroom, and their suggestions to improve it.

Two workshops then followed in small multi-age groups proposed to children in order to carry out some proposals to improve the spaces that had emerged in the conversation.

In a workshop, a structured activity was carried out for the construction of airplanes and stars, as proposed by the children themselves in the circle time; in the second workshop, the children were engaged in sky painting and in the graphic reproduction of nocturnal animals.

The children and teachers of both classes were involved in each of the proposed activities.

Finally, the multi-age sleeping area was used and furnished with new materials produced by children.

#### Observation and Reflections on Actions in Time T2

From the observational reports of the second period of the training, in the sleeping and relaxation moment, we highlight multiple relationships between big and small children of the two different classes and several occasions of shared knowledge, just as it was observed that the teachers of the different classes started to coordinate among them.

Moreover, the analysis of the observational reports showed how everyone occupied the space indiscriminately. In working together for explicitly shared common goals, the class of *big* children and the class of *small* children began to integrate within a single system.

The re-design phase at the end of the training involved a general planning of the multi-age group project to be carried out for the following year based on the new methodologies and awareness acquired during the training.

#### RESULTS

#### The Perception of Teachers' Well-Being

From the content analysis of the *descriptive report* of reflective meetings (*Reflections on actions*), the perception of well-being on the part of the teachers was detected. In particular, the aim of this qualitative analysis was to observe the perception of teachers' well-being over time, verifying any changes that occurred in the course of the training.

The conceptual model underlying this qualitative analysis, as previously described, includes the idea that the longitudinal and systematic reflective practice that had engaged the teachers' work group could improve the perceived well-being in some dimensions, such as: *sense of belonging*, *self-efficacy*, and *agency*.

As discussed above in the introduction, the *sense of belonging* refers to the perception of being part of a community/group, of feeling acknowledged, valued and included, and to have a precise role within the system, a role that is recognized by others (Rowe et al., 2007; Wike and Fraser, 2009; Skaalvik and Skaalvik, 2010; Roffey, 2012).

The dimension of *self-efficacy* refers to one's self-perception as being capable of performing one's work well and achieving the set goals, and the feeling of personal satisfaction linked to work (Karademas, 2006; Critchley and Gibbs, 2012; Kim and Yang, 2016).

The dimension of the *agency* refers, instead, to the perception of having a capacity for proactive and autonomous actions. In particular, teachers with agentic capacity feel that they have an active role within their community/work group and that they are able to participate in organizational decision-making (Wilson and Deaney, 2010; Priestley et al., 2013; Sanglim and Sungeun, 2016; Hadar and Benish-Weisman, 2019).

In accordance with some contributions from the literature (Karademas, 2006; Roffey, 2012; Hadar and Benish-Weisman, 2019) for each of these dimensions, some specific indicators of teacher well-being have been identified. Two non-independent scholars who did not know the specific hypothesis of the study, but who were experts in the research field on the topic of well-being, were engaged in order to identify from the descriptive reports some specific indicators for each of the target dimensions of well-being. Through a preliminary content analysis of the descriptive reports, the scholars individuated different indicators, each indicator being specific for a dimension of well-being, and contributed to the description of only one dimension, so that the differentiation of the three dimensions was guaranteed. This analysis consisted of a systematic and repeated reading of the descriptive reports and of a progressive identification of the content aspects that could identify the teachers' agency, sense of belonging, and self-efficacy. The indices individuated through this method, albeit based on previous studies (Karademas, 2006; Roffey, 2012; Sanglim and Sungeun, 2016; Hadar and Benish-Weisman, 2019), appear quite new and original in the research field on teachers' well-being.

The indicators used for each teachers' well-being dimension are presented below.

#### Sense of Belonging


#### Self-Efficacy


#### Agency


Following the identification of the well-being indicators, each of the six descriptive reports was systematically coded by two independent observers, reaching a satisfactory agreement index (*K* mean = 0.85). The judges who proceeded with the coding of the reports were unaware of the aims of the study and did not participate in the path with the teachers. The frequencies of the indicators for each dimension are shown in **Table 1**.

#### Sense of Belonging

Regarding the aspect of the *sense of belonging* during T1 time (at the beginning of the training), the teachers were arranged in the meeting room space separately, on one side were the teachers of the first class and on the other the teachers of the second class, with the coordinator located in the middle. As shown in **Table 1**, the teachers used the singular form referring to both themselves and to their class (no. 25 references). In descriptive reports, no reference was made to colleagues of the other class, except in one case, but in all the remaining situations they referred only to colleagues (no. 15 references) and children in their class (no. 6 references). The teachers reported negative emotions with respect to the inter-class context, especially discomfort and a sense of estrangement (no. 10 references).

*When I'm in the sleeping room I wait for my class colleague… Our children have different habits from those in the other class. When we are in the sleeping room, we feel we are guests of the class of younger children.*

*I do not quite feel I am myself in the multi-age classroom. I do not feel this space to be mine.*

The reference to the whole interclass work group was never present except in the initial statements referring to the general aim of the project. The teachers stated explicitly that there are no moments of encounter between the different classes.

*There is little knowledge between us although we have been working in the same complex for 3 years.*

In time T2, during the last meetings, the dimension of sense of belonging to the work group seemed to be characterized

TABLE 1 | Number of references in teachers' descriptive reports at T1 time and T2 time.


differently. First, the teachers occupied the space of the meeting room independently from it belonging to the first or second class. In regard to the use of the singular/plural form, at this point in the training, the teachers usually adopted the plural form (no. 14 references) and referred to the children (no. 12 references) and to the colleagues (no. 16 references) regardless of belonging to one or the other class. Moreover, the teachers reported positive emotions in regard to the inter-class context, especially spontaneity and ease (no. 9 references).

*As teachers we decided to…*

*We confronted ourselves with colleagues in the other class… We like doing things together. We have learned to do things together. No children of the "big" class or children of the "small" class,* 

*but just children! I felt free as an adult to address all children.*

*We are more spontaneous and relaxed.*

Furthermore, there are several references to the whole work group (no. 8 references) and to the small multi-age groups of children (no. 6 references).

*We have co-constructed a new way of working and approaching each other through mutual knowledge and moments of sharing. A work project has started for adults. The path began with mental availability to each other, fruitful collaboration, and sharing. The first step in the change is the sharing between adults.*

In general, the distribution of the frequencies of indicators showed that at the beginning of the training path the teachers perceived themselves as "individual practitioners"; the work group dimension did not emerge. Moreover, the analysis of the reports in time T1 showed how the interclass moment was perceived as being very difficult to handle and as being connoted by feelings of discomfort and a sense of estrangement.

Over the course of the training, a greater sense of belonging was observed, evidenced by the increased number of references to a whole work group, references to whole multi-age groups of children, and the use of plural form, such as "we." Moreover, the teachers in the last reflective meetings reported several positive emotions in regards to staying in the multi-age classroom.

#### Self-Efficacy

With regard to the dimension of the self-efficacy, in T1 time, during the first meetings, from the descriptive reports, no references emerged to teachers' competence-ability-skills, as well as no references to goals achieved with children.

Specifically, as shown in **Table 1**, the multi-age classroom was viewed with concern and negativity (no. 5 references). Many phrases and terms used by teachers indicated uncertainty (no. 9 references), and failure (no. 6 references) regarding the multi-age group of children. Many uncertainties were related to the risk of losing something positive that had already been built with and for children within the classes.

*We do not want to lose the work that we have done in class with our children.*

*I do not know if we will be able to work in the interclass, the children are very different.*

*I seem unable to pay enough attention to each child.*

Moreover, from the descriptive reports in the T1, references to negative emotional states of the children (no. 5 references) emerged.

*Even the children seem a little confused, disoriented. Children do not feel comfortable.*

Regarding the children, they are described solely within their class, and only one reference to the children's competence was reported. Moreover, the children were not perceived by the teachers as part of a unique interclass group.

In T2 time, during the final meetings, the descriptive reports revealed several references to teacher competence (no. 10 references), autonomy (no. 6 references), and certainty (no. 7 references) concerning the multi-age classroom. Contrariwise, the references to failure, or negative emotional states were absent. In these descriptive reports, the positive emotional states were prevalent (no. 18 references). Finally, in T2 more references to children's positive emotions and children's motivations to learning (no. 6 references) and achieved goals (no. 9 references) were highlighted.

*It was important to recognize that children have different needs, but also to understand that we can respond to these different needs not as single adults, but as a system of adults. Children perceive the harmony among the adults.*

*Children now feel more comfortable in the interclass space, and I also feel more open and available toward my colleagues of the other class.*

*Children are very motivated to learn new things from older partners.*

*Older children are learning to take care of the younger ones. I believe we are offering children new learning opportunities.*

The analysis of frequencies of the indicators revealed an important change in the teachers' perception of themselves in the interclass system. In particular, the teachers, at the end of the training course, expressed a greater sense of competence and certainty in the management of the multi-age class. In the course of the training, the teachers' concerns and reticence decreased and gradually a greater sense of work satisfaction emerged. In parallel, the data also showed a change in the teachers' representation of children, who were perceived more positively, as motivated, competent, and oriented toward achieving their specific learning goals.

#### Agency

As regards the dimension of *agency*, at the T1, from the descriptive reports, no references emerged to the design or co-design of activities (**Table 1**). Furthermore, during the meetings at the initial period of the training, only five teachers out of nine took part in the discussion to express their opinion; in addition, the teachers often said that they agreed with what had already been reported by colleagues in their class, to indicate a strong belonging to one's class group.

Finally, teachers reported that they did not recognize themselves as having a clear role in the management of the mixed group and not knowing how it could be possible to change the usual methods (no. 4 references); just as they believe that they do not think it is their task to introduce changes, they do not think they can suggest changes. (no. 4 references).

*But what can we do? We are only teachers, those who have to decide do not are us.*

*In my opinion, it is difficult to introduce changes in this situation.*

At T2 time, all the teachers took part in the discussions spontaneously (nine out of nine). In the descriptive reports, it was possible to note several expressions that refer to co-design (no. 10 references), and to the re-design (no. 15 references) of space practices and some references to teachers' organizational decision-making (no. 12 references). Moreover, the teachers often referred to the individual and colleagues' role in the work group (no. 16 references) and to potential changes that could improve the management of the multi-age classroom (no. 18 references).

*This is the first time that we have designed something all together.*

*The training structured into repeating phases helped us a lot, making us feel more involved.*

*Even the children played an active role in the activities, they felt they were protagonists!*

*I felt I had do "my bit" for the co-construction of the project. The relaxation room has really changed, now it is more beautiful and functional.*

*I believe that all of us teachers could really be protagonists of this project.*

*This method of working as a group could also serve to design on other aspects in the future.*

The frequencies of the indicators revealed a modification, from the T1 time to the T2, in the teachers' perception of their agentic capacity.

As shown in **Table 1**, in T2 time a larger number of teachers got involved in the training activities, actively contributing to reflection. Moreover, during the training a greater teacher awareness of their own role and the role of colleagues in the management of the interclass emerged. An orientation to the future also emerged: the teachers referred to possibilities of positive changes in the management of the interclass moment, implemented through the co-design and re-design of the practices and of the spaces.

#### DISCUSSION

Several scholars have pointed out that that teachers' well-being has been mostly conceptualized in negative terms (i.e., as stress, psychological and physical health problems, and burnout) and have suggested a reconceptualization of this construct that includes positive indicators of teacher functioning, namely teaching self-efficacy, a sense of belonging to a community, and agency (Roffey, 2012; Renshaw et al., 2015).

Starting from these premises, some interesting studies that share a processual and systemic idea of teachers' well-being have recently been carried out. Some of these aimed to study how, and by means of which processes, directors and teachers can co-construct workplaces in which the individuals can experience professional well-being (Dollard and Bakker, 2010; Zinsser and Zinsser, 2016). Other studies have instead proposed, not so much simply to reduce stress and burnout, but rather to promote positive dimensions of teachers' well-being in some specific school contexts (Critchley and Gibbs, 2012; Roffey, 2012; Cook et al., 2016).

Within this framework, our study aimed to verify the efficacy of longitudinal training with the method of reflective practice for the purpose of advancing the perceived well-being of the early childhood teachers of multi-age groupings (18–54 months).

The results of the study make it possible to highlight which important changes in teachers' perception occurred during the training with respect to some relevant dimensions of their working well-being. In particular, from the analysis of observational and descriptive reports, it emerges how teachers during training increased their perception of self-efficacy in the multi-age classroom. In fact, they reported a greater sense of effectiveness in their educational actions, both in terms of their own practices and the functional and adaptive behaviors of children. In support of this, from the beginning to the end of the training, we noticed a decrease in the uncertainties and concerns of teachers in respect to the multi-age classroom and an increase in positive emotions of satisfaction in the management of this moment (Karademas, 2006; Critchley and Gibbs, 2012; Huang et al., 2019).

During the training, moreover, there is also a development of the sense of belonging in terms of the perception of being part of a community/group; in fact, at the end of training, teachers referred more frequently to teachers' whole group and to multi-age groups of children as compared with the start. Furthermore, in the final part of the training, the teachers tended to use plural forms more in their interventions in the reflective groups, and they reported positive emotional states relating to stay in the multi-age classroom (Rowe et al., 2007; Dollard and Bakker, 2010; Skaalvik and Skaalvik, 2010).

Finally, the perception of the teachers' agentic ability was fostered by the training, in the sense that the teachers in the second part of training often described themselves as capable of proactive and autonomous actions in the management of the multi-age classroom. Also, they felt capable of participating in the organizational decision-making and to co-design the educational practices to manage the new interclass context (Wilson and Deaney, 2010; Buchanan and Bardi, 2015; Hadar and Benish-Weisman, 2019).

We believe these modifications can be attributed to some specific aspects of the proposed training. On the one hand, the method of reflective practice employed in the training allows all the teachers to experiment with an active and proactive role enabling each of them to take part in the decision-making processes in the different phases of the path. Furthermore, the training method allows the teachers to perceive themselves in their educational function, not so much as individuals, but rather as belonging to a group of colleagues, with whom it is necessary to share and negotiate strategies and educational projects.

Another important aspect of the method adopted in the course of training was the observation of the daily practices implemented in the school (Pollard, 2014). We assume that this methodological aspect has promoted in the group a sharing of new meanings and thoughts not so much on generic themes, as often happens in teachers' training, but rather on everyday situations, behaviors, and practices. In this sense, the reflective meetings represent for the workgroup a "place" in which to understand and re-signify the usual and daily practices. (Schon and DeSanctis, 2011).

The training has given the teachers the opportunity to experience having time together, it has created a clear and step-based working context, which has been repeated over time, in which the teachers could trust each other, and in which they felt that their experience and professionalism were recognized. In this sense, we consider that even the aspect of the step-by-step approach in a sufficiently dilated time was a relevant factor for the promotion of a real change in the perception of teachers' well-being.

Finally, in the training, it was possible for teachers to take into account the trainers' point of view, which, as being external to the school system, allows them to broaden the perspective and to recognize new aspects and new meanings concerning the work practices.

We retain that by means of the training, the group of teachers and the work environment in general took on characteristics of greater psychological security for teachers. In fact, in a context that is perceived as psychologically secure, it is possible for teachers to engage in changes in their consolidated work practices, separate from them, and to start exploring new methods. Training of this type has been conducted with different work groups in several ECEC centers and has shown very similar results (Venturelli and Cigala, 2017).

In the light of these results, the present study could contribute to the advancement of knowledge concerning the well-being construct, both from the research point of view and applicative point of view.

With reference to the first point, we argue that the present study, in accordance with other previous research studies (Dollard and Bakker, 2010; Critchley and Gibbs, 2012; Roffey, 2012; Cook et al., 2016; Zinsser and Zinsser, 2016), could provide evidence with respect to the fact that the well-being construct is strongly influenced by systemic-relational factors. The variables that intervene in the definition of well-being cannot be read as rigidly distinguished into "personal variables" and "contextual variables," but rather as complex variables that see the two dimensions (i.e., personal and contextual) interact together in the different daily processes that are played out in the working context. As claimed by several authors, well-being is socially and culturally constructed, rooted in a particular time and place (White, 2010; Atkinson, 2013). Therefore, for a deep knowledge of the nature of well-being, it is not enough to juxtapose descriptive variables of individuals and descriptive variables of the context, but instead we need to find methodologies and tools that allow us to detect the interdependence between these levels and how these complex relationships of interdependence vary over time.

With respect to the applicative relevance of the study, the results seem to highlight the effectiveness of the training, and therefore encourage the implementation in the contexts of the school of training courses that have some specific characteristics: they are longitudinal and systematic, focused on some critical aspects highlighted by the group of teachers and not proposed from the outside, and addressed to a small group of teachers, in order to really be able to create a real space of involvement and participation. The reflective practice, for the reasons illustrated above, appears to be a useful method, especially if it is preceded and followed by the observations in the work context, focused on the emergent critical aspects.

Starting from the results of this study and from the other research studies (Cook et al., 2016; Zinsser and Zinsser, 2016), we argue that for a true promotion and care of the teachers' well-being it would be necessary, not so much for the occasional implementation of pathways, but instead to adopt a reflective working method, like the one proposed in the present study, as a standard practice. We believe that this is the direction in which the school should go in order to make teachers develop an idea of group and teamwork that allows them to feel more confident and competent, and therefore less vulnerable and prone to developing demotivation and burnout (Huang et al., 2019).

Moreover, the present case study has allowed us to develop a system of indicators of some relevant dimensions of perceived well-being that can be applied in descriptive and narrative material. The set of these indicators, rather original within the panorama of studies on the topic, could represent a tool for the analysis of well-being to be used in other studies, as well as a tool to evaluate the effectiveness of interventions aimed at promoting well-being in the school context.

A limit of this study is that it would have been interesting to evaluate the perception of well-being on three investigated dimensions also by means of questionnaire self-reports or semi-structured interviews both at the beginning and at the end of the path; such data could have given the obtained results greater strength. Moreover, it could be interesting to follow the work group and verify the same dimensions of well-being a few months after the end of the training to verify the maintenance of the change.

Finally, it would be useful to replicate the study by proposing training to other groups of teachers, in order to be able to make more accurate and better grounded considerations regarding

the possibility of generalizing the results with respect to the effectiveness of the intervention (Woolcock, 2013).

### DATA AVAILABILITY STATEMENT

All datasets generated for this study are included in the article/ supplementary material.

#### ETHICS STATEMENT

Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. The patients/participants provided their written informed consent to participate in this study.

# REFERENCES


## AUTHOR CONTRIBUTIONS

AC and EV contributed to the design and implementation of the research, to the analysis of the results, and to the writing of the manuscript. MB contributed to the drafting of the final version of the manuscript, to the conception or re-design of the work, and to interpretation of data.

#### FUNDING

Ministero dell'Istruzione, dell'Università e della Ricerca [Award number(s): FFABR 2017, FIL 2018] are funds that the Italian Ministry of University and Research has awarded to researchers to fund their research activities.

childhood education and care. *Aust. J. Early Child.* 34, 55–63. doi: 10.1177/183693910903400408


**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Copyright © 2019 Cigala, Venturelli and Bassetti. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# Teaching Happiness to Teachers - Development and Evaluation of a Training in Subjective Well-Being

Tobias Rahm\* and Elke Heise

Institute of Educational Psychology, Technical University of Braunschweig, Braunschweig, Germany

Teachers' health is a persistent challenge for educational systems all over the world. Moreover, research results – especially in the domain of positive psychology – indicate that high levels of well-being are associated with additional benefits improving teachers' professional performance. Therefore, a training to foster subjective well-being with one training day, two booster sessions, and exercises before, during, and after the meetings was developed. It consisted of about 10 h of face-to-face time and about 3 h for the exercises in total over a 5-week training period. Main contents were conditions and consequences of positive and negative emotions and well-being, emotion regulation, time management, savoring and gratitude and the application of positive psychological interventions (like Three Good Things). Analyses of planned contrasts by means of a waiting control group design with three measurement points (pre, post, and follow-up) showed a significantly higher increase for the training group (n = 42) than for the control group (n = 47) in the frequency of positive emotions, life satisfaction, and flourishing (interaction effects d = 0.44, d = 0.31, and d = 0.32) and a significantly stronger decrease in the frequency of negative emotions, perceived stress, and experiencing emotional exhaustion (interaction effects d = 0.69, d = 0.51, and d = 0.47) from pre to 1-month follow-up. Training effects were also visible up to 5 months, although no control group could be realized for this period due to the field approach.

Keywords: positive psychology, positive education, subjective well-being, teachers' well-being, teacher training, positive interventions, positive emotions

# INTRODUCTION

The overall aim of our educational systems is to provide students with the best possible preparation for their lives. Even if the question of the correct operationalization of this goal and the priorities and methods to be chosen accordingly is answered differently in different societies, one constant remains: for the best possible education, those who deliver it must be able to make the best possible use of their potential. For several decades, the scientific community has been researching the conditions and consequences of the relatively high levels of stress, emotional exhaustion, burnout, and health impairments among teachers (e.g., De Heus and Diekstra, 1999; Gray et al., 2017). Different aspects of the teaching profession are studied to explain these findings. Skaalvik and Skaalvik (2018) summarize that "Teacher stress is typically conceptualized as unpleasant emotions resulting from aspects of the work as a teacher" (p. 1251). Beside other stressors discussed (like student diversity, discipline problems, conflicts with colleagues, lack of administrative support, etc.), time pressure was found by Skaalvik and Skaalvik (2017) to have strong relations to emotional

#### Edited by:

Paula Benevene, Libera Università Maria SS. Assunta, Italy

#### Reviewed by:

Lisa Wagner, University of Zurich, Switzerland Ilaria Buonomo, Libera Università Maria SS. Assunta, Italy

> \*Correspondence: Tobias Rahm t.rahm@tu-braunschweig.de

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 26 June 2019 Accepted: 15 November 2019 Published: 03 December 2019

#### Citation:

Rahm T and Heise E (2019) Teaching Happiness to Teachers - Development and Evaluation of a Training in Subjective Well-Being. Front. Psychol. 10:2703. doi: 10.3389/fpsyg.2019.02703

exhaustion, which in turn is one of the main predictive factors to develop burnout symptoms (Maslach et al., 1996).

It is clear that teachers with health problems are not able to fully exploit the potential of their abilities, which has a negative impact on their professional performance (e.g., instructional performance; Klusmann et al., 2008) and also causes considerable economical costs through medical-psychological treatment and absenteeism. Improving teachers' health therefore remains an important goal. A new perspective on health is introduced by positive psychology, which aims at promoting holistic well-being and individual growth and at contributing to the unfolding of one's full potential. One of the central concepts of positive psychology and the target construct of this study is subjective well-being (SWB), which is regarded as the scientific concept that comes closest to the general term "happiness" (Diener and Scollon, 2014; for an overview of well-being terms see Diener et al., 2018). SWB is composed of life satisfaction and the frequency of positive and negative affect (Diener, 1984; Pavot and Diener, 2013). All three domains of SWB are independent from each other and have distinctive associations with other variables (Diener et al., 2017). Satisfaction with life (SWL) is the cognitive component of SWB and encompasses "People's explicit and conscious evaluations of their lives, often based on factors that the individual deems relevant" (Diener et al., 2018, p. 3). People with high SWL therefore usually have a good balance of what they desire and what they actually have in their lives – especially in persistent domains like health, income or quality of one's work. The affective component of SWB consists of the frequency of positive affect (PA) and negative affect (NA). From a neurobiological point of view, PA is strongly connected to our reward system where it serves as a source of motivation and supports learning processes (Esch, 2012). Following the broaden-and-build theory of positive emotions by Fredrickson (1998, 2013), the "experiences of positive emotions broaden people's momentary thought-action repertoires, which in turn serves to build their enduring personal resources, ranging from physical and intellectual resources to social and psychological resources" (Fredrickson, 2001, p. 218). These additional resources facilitate success, personal growth, and a "successful life" – which in turn leads to more positive emotions and to a self-reinforcing upward spiral. NA, on the other hand, occurs if needs and relevant values and goals are threatened leading to a narrowing of the thought-action repertoires toward fight or flight responses (e.g., Garland et al., 2010) initiated through the release of stress hormones (Costandi, 2015). NA is not negative per se as it yields fast information on situations, helps to avoid threats to people's goals (e.g., Garland et al., 2010), but experiencing too much negativity "leads to health problems in the long run" (Diener et al., 2017, p. 2). In addition to these direct effects of SWL, PA, and NA, research also shows favorable outcomes of high overall SWB. In an extensive meta-analysis, Lyubomirsky et al. (2005) examined 225 empirical studies that tested the effects of SWB (or individual components thereof) on different areas of life. The findings show, among other things, that people with a high level of well-being live longer (Danner et al., 2001; Veenhoven, 2008), have a more efficient immune system (Barak, 2006), are more efficient and successful at work (Achor, 2010), are more creative (Baas et al., 2008), and have more versatile social relationships (Rodríguez-Pose and von Berlepsch, 2012). High SWB is therefore not only highly relevant for people's physical and psychological health but also for their personal growth and their performance in work and life.

# Teachers' Well-Being Matters

The internationally booming field of positive education applies findings from positive psychology to schools and, in addition to academic achievements, focuses on the goal of promoting the holistic well-being of everyone involved in school (e.g., Norrish, 2015; Larsen, 2016). The most important actors in achieving school objectives are teachers (Hattie, 2003). If they are to deliver education with high quality professional performance it would be favorable to secure a good state of health and even more a high level of SWB for them. While Lyubomirsky and Lepper (1999) conclude that high SWB acts as a protective factor against depression and burn-out, Sutton and Wheatley (2003) emphasize in their review, that, for example, positive and negative emotions of teachers significantly influence student outcomes. Of particular importance for the classroom is the effect of "emotional contagion," the transfer of frequent positive emotions from teachers to their students (Frenzel et al., 2009). The frequency of positive emotions also corresponds to frequent use of effective teaching strategies (Moè et al., 2010). Specifically, Kunter (2013) was able to show that affective well-being (here enthusiasm for teaching) of mathematics teachers explains changes in mathematical achievement and the students' enjoyment of mathematics. In addition, positive affect is connected with intrinsic motivation and the ideal image of the "passionate teacher," who is particularly good at stimulating his or her students and motivating them to develop their potential to the greatest possible extent (Kunter and Holzberger, 2014). More recently, Buonomo et al. (2019) found that experiencing positive emotions toward students partially mediates the negative influence of negative emotions toward students on teachers' selfefficacy, which in turn is an important predictor for professional performance of teachers and a protective factor against teachers' ill-being (Zee and Koomen, 2016).

The findings thus provide clear evidence that SWB is linked to many desirable outcomes that are also highly relevant for the education system and the teaching profession. According to general and teacher-specific findings, increasing SWB of teachers should contribute to improving their quality of health and have a positive impact on the teaching-learning processes in schools (see also Gray et al., 2017).

# Training of Well-Being

The question now arising is whether SWB can be improved. In fact, at the end of the 1990s, it was still assumed that an individual set-point existed for SWB, to which it would always return in the long term. Lykken and Tellegen (1996) came to the conclusion that "it may be that trying to be happier is as futile as trying to be taller" (p. 189). Recent research has shown, however, that this assumption was not true. In their remarkable review article, Luhmann and Intelisano (2018) examine studies on hedonic adaptation and come to the conclusion that "together, these studies indicate that SWB is relatively stable but nonetheless changeable" (p. 13). In the following, we describe different psychological interventions or programs aiming at the enhancement of PA and SWL – so called positive psychological interventions (PPIs) – and at the reduction of negative emotions.

#### Positive Psychological Interventions

fpsyg-10-02703 November 30, 2019 Time: 14:38 # 3

With the introduction of positive psychology at the turn of the milennium, more and more studies were carried out with the explicit aim of achieving a sustainable increase in well-being. For this purpose, PPIs (cf. Parks and Biswas-Diener, 2013) were developed and evaluated in various settings. One of the first large randomized placebo-control trials on PPIs was conducted by Seligman et al. (2005). Of the total of five PPIs used in this study, the exercise Three Good Things stood out in particular causing medium to large long-term effects: participants were asked to write down three things that went well each day and to provide a causal explanation for each good thing. The PPI was instructed completely online and should be performed for only 1 week. Participants were able to significantly increase their wellbeing and decrease their depressive symptoms in comparison to the placebo control group, where participants had to write about early memories every night for 1 week. The effect remained significant at a 6-months follow-up and could be replicated several times (Mongrain and Anselmo-Matthews, 2012; Gander et al., 2013). Several meta-analyses (Sin and Lyubomirsky, 2009; Bolier et al., 2013) support the finding that SWB (or components thereof) can be sustainably increased by PPIs. However, studies on individual online interventions predominate so far – probably because they are easier to implement and have the advantage of wider dissemination.

Other publications address more complex intervention programs with multiple exercises and background information on conditions and consequences of well-being. A good example is provided by Feicht et al. (2013), who evaluated a 7-week "happiness training" conducted online. The main focus was on informative video clips and texts, various PPIs (e.g., Three Good Things and Gratitude Letter, see below), and reflection exercises on personal behavior and experience. Questionnaire data show large effects, e.g., for different measures of well-being and stress reduction between training and waiting control group 4 weeks after completion of training.

Presence-based trainings offer the advantages of more direct interchange and sharing reflections and might better reach people who could benefit from a training but would not participate in an online training. University seminars in positive psychology represent a distinct group of presence-based complex interventions, which are usually offered weekly over the course of one semester. In addition to teaching the theoretical contents of positive psychology, part of the workload consists of getting to know PPIs through self-application. Goodmon et al. (2016) report significant improvements in various dimensions of wellbeing (including general happiness, SWL, perceived stress, and depression) in a pre–post design. Participants of a parallel seminar in social psychology served as a control group. Another group of face-to-face multi-component interventions is positive psychotherapy (Seligman et al., 2006). Positive psychotherapy is mainly applied by psychotherapists in clinical settings, containing elements of information and reflection in groups and PPIs as homework between sessions. In a current study in a clinical setting with predominantly depressive participants, Furchtlehner et al. (2019), for example, found better results after 14 group sessions of positive psychotherapy than after conventional cognitive-behavioral therapy.

Basically, it can be assumed that PPIs or more complex programs can improve well-being in the long term by changing the participants' individual experience and behavior. Hendricks et al. (2019) conducted a meta-analysis of multi-component PPIs and conclude that they have "a small effect on SWB and depression, and a small to moderate effect on psychological wellbeing" (p. 1). None of the 50 included studies, however, aimed at the training of teachers. Although more and more programs for positive education aiming at the fostering of students' well-being are being developed internationally (see Norrish, 2015; Larsen, 2016) and most of them contain teacher training in positive psychology, the evaluation of trainings to improve teachers' SWB has so far been neglected. In addition, the aim of increasing teachers' well-being could be well-served by combining PPIs (targeting PA and SWL) with interventions specifically aiming at the reduction of NA.

#### Psychological Interventions on Stress and Negative Emotions

As pointed out above, dealing with stress and NA is a persistent challenge in teachers' occupational life and a threat to various dimensions of health. In the field of emotion regulation, processes to influence occurrence, experience, and expression of emotions are studied. Gross (1998) summarizes five points to regulate emotions: "(a) selection of the situation, (b) modification of the situation, (c) deployment of attention, (d) change of cognitions, and (e) modulation of responses" (Gross, 1998, p. 271). Various intervention programs target different strategies to improve emotion regulation processes, leading to various positive outcomes (for a review see for example Gratz et al., 2015). Elaborated clinical emotion regulation programs like the affect regulation training (ART, Berking and Whitley, 2014) use for example psychoeducation, muscle relaxation, breathing relaxation, non-judgmental perception of emotions, acceptance and tolerance of emotions, compassionate selfsupport, identification of the causes of one's emotional response and active modification of emotions. ART could show its effectiveness in various applications like for example in the reduction of depressive symptoms (e.g., Berking et al., 2013). A special aspect of stress reduction interventions are time management trainings. As time pressure is one of the main stressors leading to teachers' experience of distress (Skaalvik and Skaalvik, 2018), time management interventions are quite often delivered in in-school trainings in Germany. Claessens et al. (2007) reviewed 32 empirical studies and defined time management as "behaviors that aim at achieving an effective use of time while performing certain goal-directed activities" (p. 262). Examples for such behaviors are setting goals, planning tasks, prioritizing activities, striving for self-awareness of one's time use, knowing the limit of one's capabilities etc. The authors concluded that "time management behaviors relate positively to perceived control of time, job satisfaction, and health, and negatively to stress" (p. 255).

#### Aims of the Present Training

fpsyg-10-02703 November 30, 2019 Time: 14:38 # 4

Altogether, research results show that high levels of SWB are beneficial in various areas of life and especially in the teaching profession, and that SWL and the frequency of PA and NA are improvable through (multi-component) PPIs and other psychological interventions like trainings in emotion regulation or time management.

The main aim of the present training therefore is to influence the individual behavior and experience of the participants in such a way that (1) positive emotions are experienced more frequently, (2) negative emotions are experienced less frequently, and (3) satisfaction with one's own life is increased. The methods to achieve these goals are positive psychoeducation (information), exercises, and homework in the sense of cognitive-behavioral therapy as well as reflections on experience and behavior in everyday life and on the experiences gained during the exercises.

Another central concept in positive psychology is flourishing, which can be understood as a more comprehensive construct of well-being describing a state of optimal social and psychological functioning (e.g., Su et al., 2014). As PPIs generally target the promotion of flourishing, we also included this construct in the development and evaluation of the training. Additionally, we were interested whether we could also stimulate and find changes in self-efficacy (Schwarzer and Jerusalem, 1995) and attributional style (Weiner, 2000), which are both variables that could explain the surprisingly high impact of the PPI Three Good Things (which is a core PPI in the training, see below) on well-being.

In order to create an attractive offer for school-internal and cross-school further education, the training should require as little effort as possible for the participating teachers and be easy to integrate into everyday school life.

# Hypotheses

We postulate that the training described below contributes to the improvement of different components of SWB (SWL, PA, and NA), to flourishing, and to emotional exhaustion and perceived stress (as additional measures of NA). In addition to the main aim of increasing SWB, we assume that it might also improve self-efficacy beliefs and attributional styles. Both constructs are not common target variables in (multi-component) PPIs, but as the conception of the training also includes contents that might evoke changes regarding these constructs, we enclosed them as additional hypotheses.

Compared to the waiting control group, the training group reports:


We assume that all differences in changes occur immediately after the training and at 1-month follow-up.

Additional hypotheses: the training group reports a higher increase in general self-efficacy (AH1) and internal locus of control (AH2) and a higher decrease in external locus of control (AH3) than the waiting control group directly and 1 month after the training.

# MATERIALS AND METHODS

#### Schedule of the Training

One elementary component of teacher trainings in Germany are 1-day internal school training courses, which are generally held annually at most schools. As changes in behavior and experience are difficult to achieve through 1-day interventions, the common 1-day format was upgraded by some elements that increase the probability of sustainable changes and can be more easily integrated into teachers' professional and private lives than additional full-day events.

Thus, a 5-week training phase with an entire training day and two 2-h booster sessions as well as small homework tasks before, during and after the training was designed (see **Figure 1**). The contents and exercises mentioned here are explained in more detail in the following two sections.

The training began with an e-mail-instructed Emotion Diary 1 week before the first meeting. Participants were requested to rate how intensely they experienced 20 given emotions and to note one situation with negative and one with positive emotions every evening. On the following 6 h training day, the experiences with the Emotion Diary were reflected upon and a large part of the contents (described in the following sections) was delivered. In the following 2 weeks, the participants carried out the cognitive PPI Three Good Things.

In the first 2-h booster session, participants' experiences were reflected upon and the most important contents were reviewed. Additionally, mindfulness, savoring, and gratitude were introduced as further topics. The participants wrote a Gratitude Letter and thought about a personal Savoring Experience. The task for the next 2 weeks was to carry out a Happiness Day where participants choose to either make a Gratitude Visit or let the planned Savoring Experience become reality.

In the second and final booster session, the experiences with the Happiness Day were reflected upon and important contents from the previous sessions were repeated. The main objective of the last session was to ensure sustainability of the training effects. The participants first wrote a letter to themselves, which was to be opened a month later, reminding them of subjectively important topics, experiences, and intentions. The participants then worked out their personal happiness project with concrete plans for the near future.

An essential message within all sessions was the affirmation that everyone had much more internal control over their SWB than he or she was usually aware of. The total attendance time of the training was approximately 10 h, about 3 h had to be spent

on the exercises at home (Emotion Diary, Three Good Things, and Happiness Day).

# Teaching Methods Used

The fundamental teaching methods included were psychoeducation in the sense of positive psychology, guided reflection, and homework between sessions.

#### Positive Psychoeducation

Psychoeducation in psychotherapy is to be understood as "systematic didactic-psychotherapeutic interventions [. . .] which are suitable to inform patients and their relatives about the disorder and its treatment, to promote the understanding and the self-responsible handling of the disorder and to support them in coping with the disease" (Bäuml and Pitschel-Walz, 2003, p. 3, own translation). Accordingly, through psychoeducation in positive psychology, participants are to become experts for the conditions under which well-being evolves. By means of interactive information transfer they learn to purposefully influence the frequency and intensity of both positive and negative emotions, life satisfaction, stress perception, personal growth, and flourishing. This type of positive psychoeducation was implemented in the training in form of short lectures (20–40 min) on specific topic units, followed by a reflection break (see below).

#### Guided Reflection

Reflection breaks were used to anchor the contents taught and experienced. After every unit, participants had time to reflect on subjectively significant information and insights. The task in the reflection breaks was (1) to review the information received, (2) to write down one to three sentences of personal relevance, and (3) to exchange ideas and thoughts about the topic with one or two other participants. Finally, issues of interest were collected in the plenary session and questions clarified. Reflection stimulating exercises were carried out at several phases during the training. For example, participants should think about what activities reliably generate positive emotions for them, what they would like to do, if time and money did not matter, or in which situations less perfectionism would be desirable. In addition to anchoring the content, the aim was to question current habits and thereby encourage participants to engage in changes in attitudes and behavior. In both booster sessions, sufficient time was given to reflect on the experiences with homework in order to provide support in case of difficulties or to reinforce successes through feedback and encouragement from the peergroup and the trainer.

#### Homework

Exercise is an important mechanism to consolidate changes in behavior and experience. In his extensive second-order metaanalysis, Hattie (2008) reports an overall effect size of only d = 0.29 for homework on learning success for students, but points out that effects are higher if learners are older and more self-regulated and if the homework is about practicing or repeating of learned contents. Also homework takes place in addition to learning effects in sessions and therefore can be considered an important enrichment. Homework also plays a decisive role in cognitive behavioral therapy (CBT) and trainings based on it.

#### Contents of the Training

The training starts with a focus on the concept of SWB as a scientific construct for the everyday term "happiness" (Diener and Scollon, 2014). Benefits of positive and negative emotions and life satisfaction are explained from evolutionary, neurobiological, and psychological perspectives. Favorable conditions for creating SWB are explained on the basis of the three basic psychological needs (competence, autonomy, and relatedness) of self-determination theory (Deci and Ryan, 2000). Information on the consequences of high SWB (see above) is used in particular to create personal relevance and motivation for the participants.

The broaden-and-build theory of positive emotions (Fredrickson, 1998, 2013, see above) plays an important role in the training. It is emphasized to the participants that increasing the personal frequency of PA empowers self-reinforcing upward spirals leading to build enduring personal resources and a more successful, pleasant, and meaningful life. By performing the PPIs described below, participants are animated to experience

positive emotions more consciously, more frequently, and more intensely, thereby setting these upward spirals in motion.

To address the aim of decreasing the frequency of NA, neurobiological, psychological, and physiological conditions and consequences of NA are explained and strategies to regulate NA and stress (like change of cognitions, consciously distracting oneself, active coping and acceptance) are discussed (cf. Gross, 1998). Participants learn about the functions of NA and are thereby empowered to distinguish better between helpful and unhelpful NA. The most important learning objective in this regard is to realize that the frequency of occurrence of unhelpful emotions and even more their persistence is changeable.

Additionally, to reduce stress from time pressure, time management techniques are taught and tried out. In particular, the focus here is on setting priorities that promote wellbeing when choosing and designing tasks, saying "no" to additional assignments, and questioning perfectionist demands. It is emphasized that one has more influence on the selection and order of activities than one is aware of (cf. Claessens et al., 2007). Participants are encouraged to prioritize their own wellbeing (Catalino et al., 2014) and to invest the time gained through time and priority management in activities that promote SWB (for an overview on the interconnection between SWB and time see Mogilner et al., 2018).

Participants are made aware of how SWL develops and that one can deliberately change the personal perspectives of what deems relevant in life (Diener et al., 2018). Also, individual processes of social comparison are addressed. It is explained how these processes sometimes undermine a fair assessment of one's successes and standing in life. Participants are encouraged to think about what might increase their SWL.

In regard to perception processes and their biases, the negativity bias – prominent in positive psychology – receives special attention. It describes the tendency that negative, potentially threatening stimuli have a greater effect on our experience and behavior than positive stimuli (Baumeister et al., 2001; Rozin and Royzman, 2001). This distortion toward an increased attention to negative environmental stimuli is explained to the participants. It is made clear that with conscious attention control, we can influence how many positive things we perceive in our environment (the perception of positive things is particularly practiced with the PPI Three Good Things).

For changes in individual evaluation processes, the training also addresses the topic of causal attribution (Weiner, 2000). People seek causal explanations for their action results. Those who ascribe their success to internal and stable causes promote their self-esteem and brighten their basic mood. The training emphasizes that the individual can gain control over evaluation processes and that it is possible to improve one's own attributional style.

#### Positive Psychological Interventions Used

As Lyubomirsky and Layous (2013) point out, not all interventions have the same effect on all people. Their personactivity-fit model stresses the importance of a fit between interventions and individuals. Also Schueller's (2014) review indicates that fitting activities achieve better effects and are more likely to be maintained. The training takes this into account by selecting and designing exercises that leave as much room as possible for their execution. Additionally, we tried to keep the required effort for them as low as possible.

#### Emotion Diary

The Emotion Diary is a proprietary translation and adaptation of the modified Differential Emotion Scale (mDES) used by Fredrickson (2011) in her broaden-and-build theory studies. Participants were asked to rate the intensity of 10 positive and 10 negative emotions. In addition, each day they were asked to write down one situation that led to positive emotions and one that led to negative emotions. The Emotion Diary should be filled out in the week before the training day and was instructed by e-mail. The aim of the intervention was to activate the participants before the first meeting and to make them perceive the variability of their own emotions. The exercise took about 5 min per day to complete. The collected situations could be worked on during the training.

#### Three Good Things

In the PPI Three Good Things, each evening, the participants were asked to write down three good experiences of the day which had provided positive emotions and to indicate what they themselves had contributed to these experiences. The first part of the exercise aims to increase awareness of positive experiences and thus to improve the frequency of positive emotions. The second part deliberately deviates from the more general original instruction of Seligman et al. (2005) in order to explicitly train a favorable attributional style by ascribing positive experiences to internal causes. Seligman et al. (2005), as well as the replication studies by Mongrain and Anselmo-Matthews (2012) or Gander et al. (2013), were able to show significant improvements in well-being, which built up over a period of 3 months and were still measurable after 6 months. Participants were encouraged to complete the exercise for at least a few days. The exercise booklet contained pages for 14-day practice. Since Gander et al. (2013) found that a 2 week exercise application did not bring any benefit for well-being, participants were informed accordingly and offered possibilities for the variation of the exercise (e.g., to focus on certain domains like work or hobby or to write down three funny things or the like) in order to prevent boredom effects. Also it was emphasized that they should stop after some days if it felt more like a burdensome duty than an enrichment. This exercise took about 5 min per day to complete.

#### Letter of Gratitude

The Letter of Gratitude is another PPI from the studies mentioned above. In the first booster session, participants were asked to write a letter in which they explicitly thanked an important person in their life. One way to increase the positive effect of the exercise (Seligman et al., 2005) is to read the letter to the addressee personally (Gratitude Visit), which could be chosen as one option for the individual design of the Happiness Day (see below).

#### Savoring Experience

fpsyg-10-02703 November 30, 2019 Time: 14:38 # 7

Bryant et al. (2011) assume that "savoring" makes a decisive contribution to experiencing and intensifying positive emotions. In the first booster session, the participants were encouraged to imagine being given 2 h of time, which could only be invested in a Savoring Experience. One of the ideas was to be planned in detail and was the second option for the Happiness Day activity.

#### Happiness Day

The given framework of this mostly self-designable PPI consisted of planning and carrying out an activity of about 2 h that would most likely evoke positive emotions. Therefore, participants could choose between the two PPIs mentioned above. The option of carrying out a Gratitude Visit was very rarely chosen in all groups (in total only about 5 of the participants opted for it). The opportunity to indulge in a Savoring Experience on the other hand, met with great enthusiasm. Examples of the activities carried out included a visit to the sauna, a campfire or musical experiences.

#### Methods to Increase Sustainability

Various elements were integrated into the training to ensure greater sustainability. One measure was to extend the training time by instructing a first exercise (Emotion Diary) by e-mail already 1 week before the meeting. Also, the participants were encouraged to form so-called implementation intentions (Gollwitzer, 1999) in order to increase the probability that the "homework" to be done during the training (Emotion Diary, Three Good Things, and Happiness Day) would actually be carried out. These implementation intentions have the form of an "if, then" sentence and have already been successfully applied in many contexts. In the second booster session, the participants had the opportunity to record important information and insights from the training and to formulate good wishes for their future selves in a letter to themselves. The letter should be opened on a fixed date approximately 1 month after the training. The final exercise of the training was to design a personal happiness project for the future and to specify behavior, life mottos, and further implementation intentions, so that concrete action plans would be available after the training.

#### Measures

In this study, we assessed the frequency of positive and negative emotions (SPANE), general life satisfaction (SWLS), flourishing (BIT), emotional exhaustion (MBI-EE), perceived stress (PSS), general self-efficacy (GSE), and locus of control (IE-4) as well as demographic data (year of birth and gender), and a code for anonymous assignment with an online-questionnaire. In addition, participants answered 10 self-constructed items on their subjective training success.

#### Frequency of Positive and Negative Experiences

The frequency of positive and negative emotions is part of the target construct of SWB and is assessed with the Scale of Positive And Negative Experiences (SPANE; Original: Diener et al., 2010; German version: Rahm et al., 2017). The instrument distinguishes between two subscales, each consisting of six items. Each subscale comprises three more general feelings (e.g., "pleasant" or "unpleasant") and three more specific feelings (e.g., "joyful" or "sad"). Participants are asked how often they experienced the given emotion in the past 4 weeks and answer on a five-point scale from 1 (very rarely or never) to 5 (very often or always).

#### General Life Satisfaction

The general life satisfaction represents the cognitive component of SWB and is measured using the Satisfaction With Life Scale (SWLS; Original: Diener et al., 1985; German version: Glaesmer et al., 2011). The instrument consists of five items on a sevenpoint response scale from 1 (strongly disagree) to 7 (strongly agree). Example items are: "In most ways, my life is close to my ideal" or "If I could live my life over, I would change almost nothing."

#### Flourishing

The construct of flourishing (or thriving) is considered to be a more comprehensive measurement of multidimensional well-being. The Brief Inventory of Thriving (BIT; Original: Su et al., 2014; German version: Hausler et al., 2017) used here is the unidimensional short version of the Comprehensive Inventory of Thriving (ibid.) and contains a total of 10 items on meaning, optimism, emotional state, flow, goal achievement, energy experience, and sense of belonging, which are evaluated on a five-point Likert scale from 1 (strongly disagree) to 5 (strongly agree). Example items are: "What I do in life is valuable and worthwhile" or "In most activities I do, I feel energized."

#### Emotional Exhaustion

The Maslach Burnout Inventory (MBI; Original: Maslach et al., 1996; German version: Büssing and Perrar, 1992) measures the burnout syndrome on three subscales. In this study, we only used the subscale Emotional Exhaustion (MBI-EE), as the corresponding construct is considered to be the main symptom of burnout. Using five items (e.g., "I feel burned out from my work" or "Working all day is really a strain for me"), participants estimate the frequency of the experienced feeling on a sevenpoint scale from 0 (never) to 6 (every day).

#### Perceived Stress

The perceived stress was measured with the Perceived Stress Scale (PSS; Original: Cohen et al., 1983; German version: Klein et al., 2016). The instrument comprises 10 items on a five-point scale from 0 (never) to 4 (very often). Example items are: "In the last month, how often have you been upset because of something that happened unexpectedly?" or "In the last month, how often have you felt that you were unable to control the important things in your life?"

#### General Self-Efficacy

The general self-efficacy describes the confidence in mastering difficult situations and is measured by the General Self-Efficacy Scale (GSE; Original: Schwarzer and Jerusalem, 1999; English version: Schwarzer and Jerusalem, 1995). The scale contains 10 items that are to be rated on a five-point Likert scale from 1 (not at all true) to 4 (exactly true). Example items are: "I can remain calm when facing difficulties because I can rely on my coping abilities" or "I can usually handle whatever comes my way."

#### Internal–External Locus of Control

fpsyg-10-02703 November 30, 2019 Time: 14:38 # 8

Control beliefs were recorded using the German scale Internale-Externale-Kontrollüberzeugungen-4 (IE-4; Original: Kovaleva et al., 2014; English version: Kovaleva, 2012). The scale comprises two subscales, one for the internal locus of control and one for the external locus of control, each consisting of two items. All four items are rated on a five-point Likert scale from 1 (doesn't apply at all) to 5 (applies completely). Example items are: "If I work hard, I will succeed" (internal) or "Fate often gets in the way of my plans" (external).

#### Subjective Training Success

For the assessment of the subjective training success we constructed 10 additional items which should be answered on a seven-point Likert scale from 1 (not applicable at all) to 7 (very applicable). Example items are: "I benefited from the training" or "I can perceive good things in my environment better than before the training."

#### Sample

The training was offered in three institutions. In two German schools, a grammar school (Gymnasium) and a vocational school, participants were recruited via short presentations held at the teachers' council that briefly described aims, contents, and schedule of the training. Interested teachers could voluntarily participate in the training in their working hours. As the third institution, the institute of educational psychology offered the training to staff of the Technische Universität Braunschweig (recruited via the department of vocational training) and to public (recruited via a newsletter for people interested in positive psychology).

For each of the three institutions (grammar school, vocational school, and university) the training was delivered twice. The participants of the first training round acted as the training group. The second training round was necessary in order to set up a waiting control group with similarly interested participants – effects of the second training were therefore not included in the data analyses. Participants were assigned to the groups by taking individual time restrictions into account to realize a sufficiently large sample. **Table 1** shows the distribution of the N = 89 participants among the groups and institutions. The participants were between 24 and 67 years old (M = 46.2; SD = 11.4), 74.2% were female.

TABLE 1 | Distribution of the participants among groups and institutions.


#### Procedure

The training was delivered twice in each institution by the first author. **Figure 2** shows the sequence of the measurement points in accordance to the training periods for each institution. All participants (training and control group) from one institution (grammar school, vocational school, or university) answered the questionnaires at the same time, while the exact training dates and measurement points differed between the institutions. The first training period for the grammar school started in April, for the vocational school in Mid-August and for the university at the end of August, all in 2017.

At measurement time point 1 (t1: pre), all participants from the respective institution received an e-mail asking them to participate in an online survey. For the participants of the training group of this institution, this e-mail also contained the instruction for the first exercise (Emotion Diary). One week after t1, the training group of this institution had their training day, 2 weeks later booster session 1, and another 2 weeks later booster session 2. One day after the last booster session, all participants (training and control group) from this institution received an e-mail asking them to take the second online survey at t2 (post). The waiting control group had to be trained in the same school year. Due to organizational reasons, the second round of the training period (for the waiting control groups) had to start at different times in the different institutions. Therefore, the follow-up measurement t3 took place with different time intervals (29 days for the vocational school, 36 for the university course and 70 for the grammar school) after t2.

To investigate long term effects, participants were asked to complete two more questionnaires. The second follow-up (t4) took place 9 weeks (vocational school) to 15 weeks (grammar school) after the end of training, the third follow-up (t5) took place 19 to 25 weeks after the end of training. As the waiting control group had already received the training at t4 and t5, the analyses of long term effects could only be conducted without a control group.

For each measurement point, depending on the response rate, one or two reminder e-mail(s) were sent a few days after the first request. The online-surveys were created with the software Questback and answered anonymously within 1 week after the according request per e-mail. All participants were informed in detail about the voluntary nature of their participation, the measures taken to maintain anonymity, and the purpose of the study and gave their written informed consent for further data processing. The study was approved by the Ethics Committee of the Faculty of Life Sciences, Technische Universität Braunschweig (FV-2019-07).

#### Data Analysis

First, a t-test was used to check whether training and control group differed in their initial values. The changes from t1 to the other measurement points within the groups were tested with t-tests for dependent samples. Cohen's d was calculated as an effect measure using the standard deviation of the mean differences in accordance to Field (2014). The differences in changes between groups were tested by planned contrasts. For

each of the scales used, the respective mean differences between the measurement points (post: t2 minus t1; follow-up: t3 minus t1) were compared between training and control group using t-tests. Cohen's d was calculated using the pooled standard deviation at t1 and the bias correction suggested by Morris (2008). In case of specific hypotheses, planned contrasts provide stronger statistical power than ANOVA (Furr, 2008). Power analyses indicate that with the given sample size and α = 0.05, effect sizes of d ≥ 0.50 could be detected with 1 – β = 0.80.

To compare the same participants at all measurement points, only cases with complete data at t1, t2, and t3 were included in the analyses. Initially 106 participants answered the survey at t1. 17 participants (9 from the training and 8 from the control group) did not take part in t2 or t3 and were therefore excluded from analyses, leading to a total sample of N = 89 participants. A ttest of mean differences between the N = 89 included and the n = 17 excluded participants showed no significant differences on five of the six target variables – only the SWLS indicated a significantly higher mean for the included cases. As Questback enforced complete answers to the questionnaires, there are no missing values in single items of the questionnaires.

To test long-term effects, changes from t1 to t4 and from t1 to t5 in the training group (n = 42) were tested for significance using t-tests. As some of the participants did not take part in these measurement points, these analyses had to be carried out with reduced sample sizes (t4: n = 30; t5: n = 35). Again Cohens d was calculated. (The results of these analyses, however, are to be interpreted cautiously as a control group could not be realized here.) All calculations were carried out with SPSS 25.

#### RESULTS

**Table 2** shows the mean values, standard deviations, and internal consistencies of the measures used at t1 for all participants. With the exception of the subscale for external locus of control (α = 0.58), all scales achieved at least good values of internal consistencies (α > 0.80). The initial values (t1) did not show significant mean differences between the training and the control group on any scale.

**Table 3** shows the results of the t-tests for mean differences between the measurement points t1 (pre) and t2 (post) withingroups. With the exception of the variables general self-efficacy and internal and external locus of control, all variables in the training group (n = 42) improved significantly from t1 to t2. The waiting control group (n = 47) showed no significant changes from t1 to t2, except for the decrease of the frequency of positive emotions. Analyses of the interaction contrasts are shown in **Table 4**. The between-group effects show a significantly higher increase in frequency of positive emotions (H1), life satisfaction (H2), and flourishing (H3) in the training group than in the control group (interaction effects d = 0.65, d = 0.40, and d = 0.43) and a significantly stronger decrease in frequency of negative emotions (H4), emotional exhaustion (H5), and perceived stress (H6) with interaction effects of d = 0.53, d = 0.42, and d = 0.49. Selfefficacy and internal and external locus of control were not significantly influenced.

**Table 5** shows the results of the t-tests for mean differences between the measurement points t1 (pre) and t3 (follow-up1) within groups. With the exception of the variable external locus of control, all variables in the training group (n = 42) improved significantly from t1 to t3. In the waiting control group (n = 47), we found a significant improvement in general self-efficacy. Analyses of the interaction contrasts are shown in **Table 6**. The between-group differences of within-group changes show a significantly higher increase in frequency of positive emotions (H1), life satisfaction (H2), and flourishing (H3) in the training group than in the control group (interaction effects d = 0.44, d = 0.31, and d = 0.32) and a significantly stronger decrease in frequency of negative emotions (H4), emotional exhaustion (H5), and perceived stress (H6) with interaction effects of d = 0.69, d = 0.47, and d = 0.51). Additional to the results at t2, an interaction effect on internal locus of control could be found (d = 0.38). External locus of control and general self-efficacy were not significantly influenced.

#### TABLE 2 | Scale characteristics (pre).

fpsyg-10-02703 November 30, 2019 Time: 14:38 # 10


M = mean; SD = standard deviation; N = 89.

#### TABLE 3 | Within-group training effects 1 day after training (pre–post).


Mx = mean value at measurement time X; SDx = standard deviation at measurement time X; d = Cohen's d; <sup>∗</sup>p < 0.05; ∗∗p < 0.01.


TABLE 4 | Between-group differences of within-group differences between t2 and t1 (pre–post).

M1post = mean difference between measurement points t2 and t1; SD1post = standard deviation of the difference; T = T-Value; df = degrees of freedom in t-test or Welch-test; d = Cohen's d; <sup>∗</sup>p < 0.05; ∗∗p < 0.01.

In summary, all main hypotheses (H1 to H6) were confirmed, the additional hypotheses only for internal locus of control and only at t3 (AH2).

Outside the waiting control group design, two additional measurement points were implemented for the training group round 1. The first was about 3 months (t4: 9–15 weeks) after the training and the second about 5 months (t5: 19– 25 weeks). As can be seen in **Table 7**, most instruments also revealed a significant improvement compared to t1. The improvements in the frequency of positive and negative emotions were not significant at t5. It is important to emphasize that no comparative data from control groups are available for this period, as the waiting control group had already received the training at this time.

In addition to the psychological constructs mentioned, participants were asked to report their subjective training success. **Table 8** shows the items in wording with mean values and standard deviations at t5. In particular, the participants reported a

#### TABLE 5 | Within-group training effects 1 month after training (pre – follow-up1).


Mx = mean value at measurement time X; SDx = standard deviation at measurement time X; d = Cohen's d; <sup>∗</sup>p < 0.05; ∗∗p < 0.01.

TABLE 6 | Between-group differences of within-group differences between t3 and t1 (pre – follow-up1).


M1fu1 = mean difference between measurement points t3 and t1; SD1fu1 = standard deviation of the difference; T = T-Value; df = degrees of freedom in t-test or Welch-test; d = Cohen's d; <sup>∗</sup>p < 0.05; ∗∗p < 0.01.

TABLE 7 | Long-term changes in the training group – without control group.


Mx = mean at measurement time X; SDx = standard deviation at measurement time X; d = Cohen's d; <sup>∗</sup>p < 0.05; ∗∗p < 0.01.

high personal benefit (M = 5.68, SD = 1.04 on a seven-point Likert scale, where 7 indicates the highest agreement) and a subjective increase in knowledge (M = 5.44; SD = 1.08). The values of all items are above the average value of the scale (4.00).

At each measurement point, the training participants also had the opportunity to leave anonymous comments on particular strengths and improvement ideas, which was frequently used. The positive comments often related to gratitude for the training and for the impulses for personal life improvement or referred to concrete personal successes attributed to the training. A number of participants also expressed that some or most of the contents had already been known before, but that it was good to be assured of "doing it right." Suggestions for improvement were, among other things, often related to individual exercises (e.g., the Letter of Gratitude was viewed critically) or the timing (more time for or between booster sessions). Quite often the wish for an additional


n = 34; M = mean; SD = standard deviation; range: 1 (strongly disagree) – 7 (strongly agree).

booster session after about half a year was expressed. All in all, the open statements showed a positive picture and a high level of satisfaction with the training.

#### DISCUSSION

In general, it could be shown that the developed training was able to meet the goal of promoting SWB. Immediately after the training (t2: post), the participants in the training group showed significantly stronger increases in frequency of positive emotions, life satisfaction, and flourishing than the participants in the control group and a significantly stronger decrease in frequency of negative emotions, emotional exhaustion, and perceived stress, accordingly. The effects persisted about 1 month (t3: follow-up 1) after the end of the training. Thus, all main hypotheses could be confirmed. Therefore, the present study adds another effective multi-component intervention to the field of positive psychology. However, because of the methodological limitations (see below), further studies should be carried out to validate the results.

Up to date, there are three meta-analyses on the effectiveness on (multi-component) PPIs (Sin and Lyubomirsky, 2009; Bolier et al., 2013; Hendricks et al., 2019), revealing small to medium effect sizes for the enhancement of well-being and the reduction of depressive symptoms. Results of the recent study are therefore in line with other findings. In these meta-analyses, different measures for well-being were used. It is worth mentioning that the present study seems to be one of few taking the measurement of NA into account which seems surprising, as the construction of SWB with its three composites (SWL, PA, and NA) is quite common (Diener, 1984; Pavot and Diener, 2013) and most of the studies included in these meta-analyses claim to increase SWB. We deem it remarkable that we found the strongest effects at t3 in the reduction of the frequency of negative emotions, although this was not trained directly in the exercises between the training day and the booster sessions. The influence on this variable thus seems to be achieved mainly through positive psychoeducation and general reflections or as side effects of the PPIs. Therefore, we recommend including the measurement of NA in future intervention studies (see below).

A main learning objective of the present training in total and especially of the PPI Three Good Things was to strengthen the belief of the participants to have more influence on their SWB than they are usually aware of and to motivate them to make use of this influence by changing behavior. Therefore, our additional hypotheses expected increases in general self-efficacy (AH1) and internal locus of control (AH2) and decreases in external locus of control (AH3). This was only partly confirmed by a significant increase in internal locus of control and only between t1 and t3 (AH2). In regard of general self-efficacy, it seems plausible that the training has no significant effect on this broad assessment of beliefs about mastery or problem-solving expectations and more specific self-efficacy constructs should be considered (see below). For locus of control it might also be possible that the short scales used in the present study with only two items for each subscale were not suitable to capture all relevant aspects of the constructs. Additionally, the subscale for external locus of control only revealed a Cronbach's alpha of 0.58. Therefore, we recommend using a more comprehensive scale in future studies.

Compared to the first time of measurement (t1: pre), significant improvements for the training groups were also observed after approximately two and a half (t4: follow-up 2) and approximately 5 months (t5: follow-up 3) respectively – however, due to the lack of a control group, these results could also be attributed to other circumstances so that they can only be interpreted cautiously as a first indication of long-term effects. The frequency of positive and negative emotions did not significantly change from t1 to the follow-up after approximately 5 months (t5). Since there was no control group at t5, it cannot be tested whether the training possibly prevented a decline in emotions caused by external influences such as more stressful phases at school or if this result is due to hedonic adaptation processes or due to the small sample size.

In addition to its overall effectiveness, the training has the advantage of a manageable amount of time, which could facilitate (active) participation: the personal contact time was only 10 h in total, divided into one training day and two booster sessions, and the self-structured exercises lasted only about 3 h in total over a period of 5 weeks. The feedback received during the training, the results of the subjectively assessed effects of the training, and the open comments showed that the offer was well-received by the target group. The integration of coping strategies for aversive emotions and time management techniques probably also contributed to this. Both contents are known to be important skills for work performance and health of teachers (Skaalvik and Skaalvik, 2018) but have rarely been considered in multi-component interventions claiming to increase SWB.

#### Limitations and Further Research

Due to the field research approach, this study has two obvious limitations: (1) For organizational reasons in the schools, the follow-up measurement could not be carried out in all three

groups after exactly 30 days so that there was a notable difference in the amount of time between the end of the training and the first follow-up measurement (t3) for the three institutions, with the grammar school being assessed more than 1 month later (70 days in comparison to 29 and 36 days). (2) The first two training groups consisted of teachers from one school each, while the third group also included some teachers from different schools but mostly members of other professions. In our opinion, however, the fact that the training has proven effective even under these suboptimal conditions indicates the robustness of the effects. Nevertheless, in subsequent studies these restrictions should be avoided.

As mentioned above, 17 participants who did not take part in all three measurement points were excluded from the analyses. These excluded cases revealed a significantly lower life satisfaction at t1. Therefore, it seems that less contented people might also be less motivated to complete the follow-up questionnaires. To avoid biases in this regard, one could consider working with incentives in future studies.

Another issue are the shortcomings of the evaluation design. As Hendricks et al. (2019) found an effect of the study quality in their meta-analysis (expressed, e.g., in randomization and placebo-control) showing lower effects in high-quality studies, it may be that the effects of the present study are overestimated. In a larger project with more resources for recruiting schools and teachers, it should be possible to secure a larger sample size which would be especially favorable for two different approaches: (1) A cross-school training with randomized assignment of the participants to the groups including an active alternativetreatment-control group would avoid biases concerning the selection procedure (which was based on available time of the participants in this study) and could ensure that changes are due to the training and not due to social interaction or other confounded variables. Additionally, as more research about the precise mechanisms of PPIs and multi-component interventions is still needed (Mongrain and Anselmo-Matthews, 2012; Chow, 2018), it would be even more valuable to also carry out a component evaluation to check which individual components are responsible for the effects. Therefore, one could develop several independent training modules (e.g., positive emotions, emotion regulation, or time management) and apply them in different order to several groups to find out which modules affect which facets of well-being. (2) In-school trainings with several participating schools would allow to investigate the influence of school-specific variables such as openness to change, peer support, experience of stress, or general school climate on the sustainability of training successes using multi-level analyses. For example, Ebersold et al. (2019) were able to show that teachers' well-being was significantly influenced by the perceived autonomy support by their principals, mediated by satisfaction of the basic psychological needs for autonomy, competence, and relatedness (Deci and Ryan, 2000). It can be assumed that effects of well-being interventions will last longer in a supportive environment.

As the research on the broaden-and-build theory revealed favorable outcomes of increasing the frequency of PA, such as higher resilience, social connection, and functioning at optimal levels (Fredrickson, 2013), one can hope that the present training will also contribute to build such enduring resources not targeted directly. This, of course, needs to be examined in additional studies preferably with long-term follow-ups 6 or more months after intervention. Regarding the inclusion of interventions to decrease the frequency of NA, it would also be of interest if such combined interventions targeting all components of SWB could be even more effective than focusing PPIs only.

In addition to the constructs used here for evaluation, the construct of self-efficacy seems especially promising, as it is associated both with well-being (e.g., Luszczynska et al., 2005) and with intention building and behavior (e.g., physical activity, Rovniak et al., 2002). Since general self-efficacy was not affected in this study, domain-specific variants should be used in further investigations. In the school context, Schmitz and Schwarzer (2000) were already able to show correlations between teacher-specific self-efficacy and job satisfaction or burn-out. Zee and Koomen (2016) synthesized 165 articles on teacher self-efficacy (TSE) from over 40 years of research concluding: "Results suggest that TSE shows positive links with students' academic adjustment, patterns of teacher behavior and practices related to classroom quality, and factors underlying teachers' psychological wellbeing, including personal accomplishment, job satisfaction, and commitment" (p. 981). Furthermore, in the field of positive psychology, a scale for measuring well-being-related selfefficacy with seven factors (positive focus, emotion regulation, engagement, connectedness, meaning, goal achievement, and time sovereignty) is currently being constructed and validated (Rahm et al., in preparation).

In order to better classify results from interventions such as well-being trainings, but also applications of short PPIs or even more complex changes such as school development processes, it would be helpful to obtain more precise information about the fluctuations in well-being over the course of the (school) year. Findings about such general variations might also contribute to the explanation of effects of PPIs and multi-component interventions. In addition, knowledge about periods regarding well-being or ill-being, such as more relaxed or more stressful times, could be used to determine the right timing for wellbeing interventions. Furthermore, findings in this area could also make valuable contributions to the current discussion on hedonic adaptation and set point theory (Brown and Rohrer, 2019; Luhmann and Intelisano, 2018).

Another promising field for further research is the influence of improved teacher well-being on students. What happens to school climate and class climate, what happens to pedagogical relationships, and what happens to the well-being and academic performance of the students if the teaching staff is continuously trained and supported in the sense of positive psychology?

# CONCLUSION

In summary, it can be concluded that the training was wellreceived by the participants and increased their SWB for at least 1 month after the training. Due to the low organizational and

time expenditure, it can be easily integrated into the training systems for teachers. In the context of the growing field of positive education, the training could prove to be a motivating starting point for school development processes to promote wellbeing and flourishing for teachers and students if the results found here can also be replicated in methodologically stricter research designs.

#### DATA AVAILABILITY STATEMENT

The datasets generated for this study are available on request to the corresponding author.

#### ETHICS STATEMENT

The studies involving human participants were reviewed and approved by Ethical committee of the Faculty 2, Technische Universität Braunschweig (FV-2019-07). The patients/participants provided their written informed consent to participate in this study.

#### REFERENCES


#### AUTHOR CONTRIBUTIONS

TR developed and delivered the training, organized the data collection, prepared the data set for the analyses, conducted the analyses, and wrote the first draft of the manuscript. EH scientifically supervised each step of the study, contributed to the development and the evaluation design of the training and was substantially involved in the final version of the manuscript.

#### FUNDING

The Publication Fund of the Technische Universität Braunschweig funded the open access publication fees.

#### ACKNOWLEDGMENTS

We would like to thank the management and organizers of the two participating schools for their support and all participating teachers for their commitment.


and negative feelings. Soc. Indic. Res. 97, 143–156. doi: 10.1007/s11205-009- 9493-y



eds J. Weinman, S. Wright, and M. Johnston, (Windsor: NFER-NELSON), 35–37.


**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Rahm and Heise. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Teachers' Burnout: The Role of Trait Emotional Intelligence and Social Support

Caterina Fiorilli<sup>1</sup> \*, Paula Benevene<sup>1</sup> , Simona De Stasio<sup>1</sup> , Ilaria Buonomo<sup>1</sup> , Luciano Romano<sup>1</sup> , Alessandro Pepe<sup>2</sup> and Loredana Addimando<sup>3</sup>

<sup>1</sup> Department of Human Sciences, LUMSA University, Rome, Italy, <sup>2</sup> Department of Human Science "R. Massa", University of Milano-Bicocca, Milan, Italy, <sup>3</sup> Department of Teaching and Learning, University of Applied Science and Arts of Southern Switzerland, Locarno, Switzerland

The current study investigates the relations among teachers' trait emotional intelligence, internal and external social support, and their levels of burnout. We hypothesized that both emotional intelligence and teachers' perceived social support were associated with low level of teachers' burnout. We further expected that internal and external support mediated the relationship between trait emotional intelligence and burnout scores. Participants were 318 in-service Italian teachers. The structural equation modeling analysis supports the idea that teachers' trait emotional intelligence is strongly and directly associated with their burnout. Furthermore, internal social support (from the teachers' workplace relationships) was more effective on burnout than support forthcoming from their external context. On the contrary, the mediation hypothesis was partially supported by the empirical data. These findings shed light on the relationship between teachers' emotional competence and their burnout experience at school.

#### Edited by:

Renato Pisanti, Niccolò Cusano University, Italy

#### Reviewed by:

Chris Verhoeven, Leiden University, Netherlands Fernando Doménech-Betoret, Jaume I University, Spain Kia Gluschkoff, University of Helsinki, Finland

> \*Correspondence: Caterina Fiorilli fiorilli@lumsa.it

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 20 June 2019 Accepted: 20 November 2019 Published: 10 December 2019

#### Citation:

Fiorilli C, Benevene P, De Stasio S, Buonomo I, Romano L, Pepe A and Addimando L (2019) Teachers' Burnout: The Role of Trait Emotional Intelligence and Social Support. Front. Psychol. 10:2743. doi: 10.3389/fpsyg.2019.02743 Keywords: burnout, trait emotional intelligence, social support, Italian teachers, family

# INTRODUCTION

Trait Emotional Intelligence (trait EI) is a constellation of self-perceived, emotion-related abilities that enable individuals to recognize, process, and use emotional information (Sevdalis et al., 2007; Petrides, 2010). According to Petrides et al. (2016) trait EI refers to how people perceive their own emotional and social effectiveness, and represents a comprehensive dimension of the affective aspects of personality. It is generally operationalized in four components: general well-being (e.g., self-esteem, happiness, and optimism), self-control (e.g., stress management, and emotional control), emotionality (e.g., emotional perception and expression), and sociability (e.g., social awareness, emotional management, assertiveness, adaptability, and self-motivation).

Numerous studies have shown a strong association between trait EI and well-being in the workplace (e.g., Di Fabio and Kenny, 2016; Di Fabio, 2017). Individuals with high EI are more likely to see themselves as efficient, experience more positive than negative emotions, forge more positive relationships with others, and perceive everyday challenges in a way that promotes wellbeing, engagement, and job satisfaction (Maslach et al., 2001; Furnham and Petrides, 2003; Zeidner et al., 2004, 2012; Avsec et al., 2009; Brackett et al., 2010, 2011; Santisi et al., 2014; Benevene et al., 2018a,b). The literature on teaching has amply documented the role of trait EI in preventing adverse outcomes, such as burnout (O'Boyle et al., 2011; Mérida-López et al., 2017).

**229**

There is a wide agreement to consider burnout syndrome as a negative work-related outcome due to the long-term effect of strain resulting from repeated exposure to stressful events (e.g., Demerouti et al., 2002). According to several scholars the main core of burnout is feeling emotionally exhausted, which is characterized by physical and psychological fatigue (e.g., Travers, 2017) in three different social contexts, namely: private life, workplace, and relationship with clients (in this study with students) (Kristensen et al., 2005; Avanzi et al., 2014; Schonfeld et al., 2017). Furthermore, previous studies indicated that teachers' burnout is also related to gender (i.e., females are more at risk than men) and positively associated with years of experience (i.e., high risk of burnout among less experienced teachers) (Leiter et al., 2014).

Teachers with more advanced emotional competencies are better equipped to handle the relative strain and emotional burden and to make sense of their reactions to sources of stress (e.g., Chan, 2006; Pishghadam and Sahebjam, 2012; Borrelli et al., 2014; Ramaci et al., 2016; Buonomo et al., 2017; De Stasio et al., 2017; Fiorilli et al., 2017a).

Moreover, teachers' social support (internal and external to the workplace) may constitute a further resource when facing stressful events. Internal support comes from within the work setting itself (i.e., support from colleagues, supervisors, school leaders); whereas external support comes from the teachers' private life (i.e., support from friends, family members, partners) (Schaufeli and Bakker, 2004; Gavish and Friedman, 2010; Carlson et al., 2014; McNall et al., 2015; Pérez-Fuentes et al., 2019). Internal support is generally associated with teachers' sense of belonging, their commitment, and general well-being (e.g., Collie et al., 2017; Travers, 2017). On the other hand, external support has positive effects on job satisfaction and performance (Halbesleben, 2006; Skaalvik and Skaalvik, 2009; Carlson et al., 2014; McNall et al., 2015). Such findings come from a variety of cultural settings including Spain, Turkey, Korea, Jordan, Israel, Italy, and Palestine (Aycan and Eskin, 2005; Veronese et al., 2018; Fiorilli et al., 2019).

Overall, teachers' social support seem to have a strong relationship with both teachers' trait emotional intelligence, on one hand, and their burnout level, on the other. Due to the relevant implications of the trait EI in maintaining good social interactions (e.g., ability to interpret social cues, self-regulation of emotions, supportive social interactions), it is expected that teachers with high trait EI are more likely to positively perceive existing resources from both inside and outside their workplace (Gallagher and Vella-Brodrick, 2008).

Nevertheless, while the relationships among the aforementioned variables are well known and supported by substantial findings, the mediating role of teachers' perception of their social support (both internal and external to the workplace) is less known. In one of the few studies that has investigated the mediating role of social support between trait emotional intelligence and burnout risk, the authors address the internal social support without any comparison with the external one (Ju et al., 2015).

Overall, an in-depth comparison of the two kinds of social support (internal versus external) is required to build up a more comprehensive picture of teachers' burnout risk in relation to their personal characteristic named trait emotional intelligence.

# The Current Study: The Protective Role of Social Support

The aim current study was to model within a single structural equation model, the cumulative network of relationships among teachers' trait emotional intelligence, perceived internal (e.g., colleagues and supervisor) and external social support (e.g., family and friends), and levels of professional burnout (see **Figure 1**). In line with evidences gathered in previous studies, the current study tested the following specific directional hypotheses: (1) teachers' trait emotional intelligence would be negatively correlated with burnout scores as well as positively associated with perceived external and internal support. In turn, we also expected that both dimensions of social support (e.g., external and internal) were negatively associated with burnout scores; (2) the relationship between emotional trait intelligence and burnout scores would be mediate by perception of social support (e.g., external and internal). In order to test the network of associations among the variables of interest, structural equation modeling with the decomposition of total effects and analysis of regression coefficients was adopted (Byrne, 2013).

# MATERIALS AND METHODS

#### Participants

The sample was composed of 318 teachers (94.7% female) aged 27–65 years (M = 47.1, SD = 8.55) who were mostly employed in primary schools (71.2%) and public nursery schools (13.1%) in Northern (8.8%), Central (66.1%), and Southern Italy (25.1%). Most of the participants were tenured teachers (72.3%); about 63% held a master's degree or other postgraduate qualifications, and 37% held a high school diploma. Their years of teaching experience ranged from 1 to 41 (M = 18.1, SD = 11.1).

#### Procedure

The data was collected in late 2018 in seven schools in Northern, Central, and Southern Italy. Having obtained approval from the school principals, a researcher contacted the teachers informing them about the study and inviting them to participate. Paper and pencil questionnaires were used for data collection. All participants signed informed consent forms and were informed about the anonymity and confidentiality of their answers. The research protocol was approved by the Ethics Committee of the LUMSA University, Rome.

#### Instruments

#### Trait Emotional Intelligence Questionnaire (TEIQ)

Trait Emotional Intelligence was assessed via the Italianlanguage version of the 30-item Trait Emotional Intelligence Questionnaire-Short Form for Adults (TEIQue-ASF; Petrides, 2009; Italian validation by Di Fabio, 2013). The TEIQue-ASF is a self-report questionnaire whose items (e.g., "Expressing my emotions with words is not a problem for me"; "I often find it

difficult to see things from another person's viewpoint") are rated on a 7-point Likert scale, ranging from 1 (completely disagree) to 7 (completely agree). The Italian version of the TEIQue-ASF measures four latent factors: Emotionality (eight items), Sociability (five items), Self-control (seven items), and Well-being (10 items). In the current study, the alpha coefficients ranged from 0.716 to 0.658, whereas the overall reliability for the measure was equal to 0.868.

#### Internal Social Support (HSE)

We used the published Italian version of the Health and Safety Executive Stress Indicator Tool (Balducci et al., 2015) regarding teachers' perception of social support coming from the school network to deal with stressful events. In the current study we used two subscales, namely: Supervisor Support (five items) and Peer Support (four items). Examples of items in these subscales included the following: "I am given supportive feedback on the work I do" (for Supervisor Support), and "If work gets difficult, my colleagues will help me" (for Peer Support). All items were answered by using a Likert-Type response scale ranging from 1 (Never) to 5 (Always). In the current study the alpha reliability coefficients were 0.893 (peer support) to 0.904 (supervisor support).

#### External Social Support (MSPSS)

The Multidimensional Scale of Perceived Social Support (Zimet et al., 1988; Di Fabio and Busoni, 2008) is an Italian 12-item self-report questionnaire focused on teachers' perception of social support coming from their private life and able to support them in school-related stressful events. A few examples of the items included the following: Significant Other (four items) (e.g., "There is a special person who is around when I am in need"), Family (four items) (e.g., "My family really tries to help me"), and Friends (four items) (e.g., "I have friends with whom I can share my joys and sorrows"). Items were scored by using a Likert-Type response scale ranging from 1 (Very Strongly disagree) to 7 (Very strongly Agree). In the current study the alpha coefficients ranged from 0.930 (family) to 0.870 (significant others).

#### Burnout

We used the published Italian version of the Copenhagen Burnout Inventory (Kristensen et al., 2005; Fiorilli et al., 2015). This selfreport questionnaire is composed of three subscales with a total of 19 items, each rated on a 5-point Likert scale. The three subscales are: Personal Burnout (six items) (e.g., "How often do you feel tired?"), Work-related Burnout (seven items) (e.g., "Do you feel worn out at the end of the working day?"), and Student-related Burnout (6 items) (e.g., "Do you find it hard to work with students?"). Items were scored by using a Likert-Type response scale ranging from 1 (Never) to 5 (Always). In the current study, the alpha coefficients were as follows: personal burnout (0.886), work-related burnout (0.875), and studentrelated burnout (0.846).

# Analytic Strategy and Data Modeling

Data analytic strategy consisted of two sequential stages. First, main descriptive statistics, as well as zero-order correlations, were computed. The presence of multivariate outliers was assessed by computing Mahalanobis' distance (p < 0.001) for all variables. No multivariate extreme values were found and consequently removed from the sample. The distribution of empirical indicators was then explored. None of the variables under study reported kurtosis or skewness values falling outside

the recommended thresholds of +2 and −2 (George and Mallery, 2010), and they resembled a normal distribution.

As required in the case of structural equation modeling strategies (Kline, 2011; Pepe et al., 2017, 2018; Veronese et al., 2017), the goodness of fit indexes were analyzed to estimate the overlap between the observed matrix of covariances (S) and the reproduced matrix of covariances (6). Thresholds for good model fit were: RMSEA < 0.07 (Schermelleh-Engel et al., 2003), NFI > 0.95, NNFI > 0.95 (Marsh et al., 2004; Nagengast et al., 2014), CFI > 0.95 (Hu and Bentler, 1999). The maximum likelihood method (Kline, 2011) was used to estimate the parameters for the structural models. All analyses were conducted by using SPSS AMOS 23 (Arbuckle, 2014).

#### RESULTS

The results are presented in two sections covering general descriptive statistics/zero-correlations and the results of the structural model. Descriptive statistics and zero-order correlations are summarized in **Table 1**.

Trait emotional intelligence and burnout scores showed negative medium-large and statistically significant correlations. In particular, well-being scores were negatively correlated to work-related burnout (r = −0.460, p < 0.001) and personalrelated burnout (r = −0.435, p < 0.001). Similarly, negative correlations were found between student-related burnout and both sociability (r = −0.342, p < 0.001), and self-control (r = −0.353, p < 0.001). Besides, work-related burnout scores were generally correlated negatively with all TEIQ subscales. Moreover, trait emotional intelligence was positively associated with internal as well as external support that teachers perceived by their sources. With regard to the associations between burnout and social support, we found that, as expected, negative correlations with all subscales, with the exception of the studentrelated burnout that showed no association with supervisor support. The results of the structural equation model are reported in **Figure 2**.

The analysis of goodness of fit indexes supported the acceptance of the proposed conceptual model (χ 2 (58) = 120.3, p < 0.01; NC = 2.07; RMSEA = 0.058, CI 90% [0.043;0.073], NFI = 0.928, NNFI = 0.961, CFI = 0.961), suggesting that the effects among the modeled variables were both conceptually and statistically significant. The main paths of the structural model were then assessed by decomposing the total standardized effects in direct and indirect effects (standardized values are reported in **Figure 2**), in order to identify the directions and the sizes of the associations. With regard to the second hypothesis, which focused on the direct effects among studied variables, we found a negative total effect of teachers' trait EI on burnout levels (β = −0.63, p = 0.010). Whereas trait EI resulted in two positive direct effects, similar in both size and direction, with perceptions of external support (β = 0.28, p = 0.008) and internal support (β = 0.34, p = 0.010). Finally, two others direct negative associations were estimated in relation to burnout scores, with both internal (β = −0.17, p = 0.007) and external (β = −0.09, p = 0.029) support reporting negative direct effects, meaning that


TABLE 2 | Summary of direct, indirect, and total standardized effect.


N/A, not applicable; 95th CI = confidence intervals, <sup>∗</sup>p < 0.05, ∗∗p < 0.01. Indirect effect of trait emotional intelligence on burnout (−0.083) was the sum of standardized indirect effects via emotional support (−0.024<sup>∗</sup> , 95th CI [−1.139 to −0.043]) and internal support (−0.057<sup>∗</sup> , 95th CI [−1.358 to −0.059]).

the higher the perceptions of being supported (i.e., internally and externally) the lower the burnout levels (see **Table 2** for details).

Regarding the indirect effects of social support perceived by teachers, we found a very minimal mediating role of teachers' social support (both internal and external) on the relationship between trait EI and burnout levels. A closer look into the composition of the total effect (β = −0.63, p = 0.010) in direct (β = −0.55, p = 0.016) and indirect (β = −0.08, p = 0.028) components revealed that the effect of emotional intelligence trait on burnout (via different type of social support) was only minimal.

#### DISCUSSION

In this study, we investigated the role played by perceived social support in the relationship between teachers' trait emotional intelligence and levels of burnout. As predicted, all the key research variables were significantly associated showing that high level of teachers' trait emotional intelligence (i.e., emotionality, sociability, well-being, and self-control) was associated with low burnout level (i.e., personal, work-related, and student-related). As expected, and in line with previous research (e.g., O'Boyle et al., 2011; Schutte and Loi, 2014), teachers' trait EI gives them emotional resources to face school-related stressful events (i.e., work-related burnout and student-related burnout) as well as stressful events in their own private life (i.e., personal burnout). Coherently with the personality-based dimension approach (Petrides et al., 2016), trait EI is a dimension able to predict a wide range of people's well-being (e.g., high level of happiness, wellbeing, and satisfactory interpersonal relationships) by having an impact on several life contexts (e.g., Vergara et al., 2015). In the same direction we found that the more social support (from both inside and outside the school setting) the teachers perceived, the less they experienced burnout. According to previous research (Fiorilli et al., 2017a), when teachers show high satisfaction with support perceived as available to them their burnout risk decreases. Surprisingly, no association was found between student-related burnout and support coming from teachers' supervisors. Previous findings have shown that supervisors are one of the most important support mechanisms for teachers who may feel more effective in their work with

students by having a greater sense of autonomy. However, the role played by supervisors is related, among others variables, with teachers' years of experience that, in the current study, has shown a very weak association. This leads us to believe that further investigations with a larger sample of teachers may be needed on this association (e.g., Van Droogenbroeck et al., 2014).

Moreover, in line with previous findings (Halbesleben, 2006; Pomaki et al., 2010; Pietarinen et al., 2013), we found a strong association between teachers' personal resources (in this case, trait EI) and social support. In practice, teachers with high trait EI perceived as effective the support coming from internal and external sources. This was an expected finding, given the existing literature addressing the relationships between peoples' trait EI and their positive attitude toward coping strategies, such as seeking support to face difficult situations (Andrei and Petrides, 2013). Furthermore, no associations were found between all subscales of TEIQ and the additionally external social support coming from significant others. It is an unexpected result that may lead to reconsider the informative value of the subscale labeled "significant others" within the same instrument where family and friends dimensions may exhaustively explore peoples' relations. In this regard, it has to be stressed that the Multidimensional Scale of Perceived Social Support was administered in the Italian context only among students (Di Fabio, 2015). Whereas in the current study we addressed workplace life by asking teachers about the sources available to support them in school-related stressful events. In this regard, friends' support subscale may overlap with the significant other subscale.

With regard to the direct and indirect relationship among studied variables, the findings partially supported our hypotheses. First, teachers' trait emotional intelligence may buffer individuals from burnout and makes them more inclined toward receiving social support from internal as well as external sources. Second, comparison of the direct and indirect effects of trait emotional intelligence on levels of burnout in our sample of teachers suggests that the indirect effect of trait EI, via social support, is trivial if compared to its direct effect. Looking at the trait EI dimensions we can observe that high level in sociability, emotionality, well-being, and self-control gives people a set of several competences to face with emotional events where it is presumable that perceived social support plays an important role, but not essential, as our findings seem to support. The direct impact of trait EI on teacher burnout identified in this study is characterized by an extensive body of research suggesting that burnout is a subjective phenomenon that varies as a function of the interaction between personal resources and events experienced (Lambert et al., 2009; Platsidou, 2010; Fiorilli et al., 2017b). The components of trait EI include general wellbeing, sociability and self-control, all of which may reasonably be expected to help teachers adopt a positive perspective on critical school events (Miao et al., 2016). This appears to be particularly true in light of the weak indirect effect – via social support – of trait EI on burnout in our sample. In other words, teachers' trait emotional intelligence gives them sufficient resources in terms of stress management, empathy, and social awareness that may assure them adaptability to workplace challenges (Petrides et al., 2016). More specifically, teachers' trait EI leads teachers to be less vulnerable to burnout by giving them more sensible and effective ways to use emotional information coming from their workplace context (Brackett et al., 2010), as well as their personal life.

Furthermore, we found that internal social support reported a higher effect on burnout levels than support received from the teachers' private life. More specifically, teachers with high trait EI are more likely to positively perceive the support coming from their supervisor and colleagues than from their private life. A perspective focuses on workplace well-being leads us to positively evaluate teachers' confidence in internal social support. Effectively, colleagues and supervisors may be more likely to support workers by helping them in solving problems and implementing new professional strategies and practices. Even though these results require further investigation, by taking into account the role played by all subscales of external support in explaining its role (see the not significant association found in relation to significant others subscale which may partially account for the weaker role of external support compared with the internal one), it is important to highlight the extensive literature supporting the relevant role of the supervisor and colleagues in helping teachers with their school-life challenges (Day and Leiter, 2014).

# CONCLUSION, LIMITATIONS, AND FUTURE RESEARCH DIRECTIONS

Theoretical perspectives on trait EI suggest routes to enhancing teachers' personal resources and thereby their ability to cope with stressful events. More specifically our findings suggest that personal competence in the emotional domain may significantly contribute in teachers' well-being rather than social support itself. Importantly, trait EI has proved susceptible to improvement through training programs (Vesely et al., 2014), and improved EI – as the present findings confirm – can help to prevent teacher burnout (Brackett and Katulak, 2006). Our findings lead us to take into account the implications for teachers' professional development. Despite the fact that trait EI is a personal variable, it is not unalterable one. On the contrary, it is susceptible to improvement through training programs (Brackett and Katulak, 2006; Vesely et al., 2014). Additionally, in line with previous research (e.g., Demerouti, 2014), our results should seriously lead us to consider individual strategies as key variables for reducing burnout as well as enhancing personal resources. In the current study, self-control is the most relevant dimension within teachers' trait EI that may support the idea that their ability to manage emotions in the social interactions could be a guideline for the teachers' professional development programs.

The current study displayed some limitations that should be taken into account in future research. First, longitudinal studies are required to establish whether internal and external support available to teachers result and/or impact their trait emotional intelligence. Second, future research should include an analysis of contextual factors such as leadership from the school principal and school climate (e.g., McCarthy et al., 2017) in order to further understand how teachers' evaluation of internal support may

depend on the quality of workplace resources. Third, in future research, both the quality and quantity of teachers' relationships with other teachers should be assessed in greater depth and using a range of measures, rather than exclusively relying on self-report instruments as in the present study. More specifically, using a qualitative method for collecting data (e.g., interviews and observations) may shed a light on the quality of each social support (i.e., whether it offers emotional support, instrumental or informative support).

#### DATA AVAILABILITY STATEMENT

The datasets for this article are not publicly available because of local legal and privacy restrictions (Italian Data Protection Code – Legislative Decree No. 196/2003).

# REFERENCES


#### ETHICS STATEMENT

The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation. All participants gave written informed consent in accordance with the Declaration of Helsinki.

#### AUTHOR CONTRIBUTIONS

CF, IB, and LR designed and carried out the study and contributed to the analysis of the results and writing of the manuscript. AP collected data and contributed to the analysis of the results and writing of the manuscript. PB, SD, and LA supervised the study design and the manuscript draft.

of originating and receiving domains. J. Manag. 40, 845–865. doi: 10.1177/ 0149206311414429




**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Fiorilli, Benevene, De Stasio, Buonomo, Romano, Pepe and Addimando. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fpsyg-10-02822 December 12, 2019 Time: 14:58 # 1

# Teaching Beliefs on Developmentally Appropriate Practice Among Chinese Preschool Teachers: The Role of Personality

#### Paul Yau-Ho Wong\*

School of Arts and Humanities, Tung Wah College, Mongkok, Hong Kong

Research has shown that teachers' personality is related to their teaching practices. However, few such research has been conducted in the preschool settings. This research adopted a cross-sectional quantitative approach to examine how preschool teachers' personality related to their beliefs of Developmentally Appropriate Practice (DAP). 544 preschool teachers were randomly selected to complete Teacher Beliefs and Practices Survey and the Myer–Briggs Type Indicator (MBTI). Findings showed that the personality type profiles of preschool teachers are characterized predominantly by Sensing-Feeling-Judging. Among all personality types, extraverted and intuitive teachers tended to show higher scores in DAP beliefs and practice than that of the introverted teachers. Analyses from regression findings also indicated that the Extroversion-Introversion dimension predicted significantly teachers' beliefs and practices of DAP. In addition, teachers' teaching experiences in association with personality predicted their DAP practices, suggesting that there was an combined effect of teachers' personality and work experience in adopting DAP. Implication for school managers to provide professional development to support teachers in understanding and practicing DAP in preschools are discussed.

#### Edited by:

Xiaomin Sun, Beijing Normal University, China

#### Reviewed by:

Sanja Tatalovic Vorkapi ´ c,´ University of Rijeka, Croatia Michail Kalogiannakis, University of Crete, Greece Stamatis Papadakis, University of Crete, Greece

#### \*Correspondence: Paul Yau-Ho Wong

paulwong@twc.edu.hk

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 01 June 2019 Accepted: 29 November 2019 Published: 13 December 2019

#### Citation:

Wong PY (2019) Teaching Beliefs on Developmentally Appropriate Practice Among Chinese Preschool Teachers: The Role of Personality. Front. Psychol. 10:2822. doi: 10.3389/fpsyg.2019.02822 Keywords: Developmentally Appropriate Practice, personality, preschool teachers, teaching beliefs, teaching practices

# INTRODUCTION

Quality education is conducive to young children's later development (Early et al., 2007) and learning motivation in later schooling (Magnuson et al., 2007). According to the National Association for the Education of Young Children [NAEYC] (2013), quality early childhood education builds on the Developmentally Appropriate Practice (DAP) that "involves teachers meeting young children where they are (by stage of development), both as individuals and as part of a group; and helping each child meet challenging and achievable learning goals." Evidences have shown that the DAP approach has positive impact on young children's learning and development. As such, children found their learning more meaningful in schools that implemented the DAP approach (Lee and Lin, 2013) and displayed better psychological well-being outcomes in terms of self-competency (Jambunathan, 2012) and self-efficacy and social skills (Marcon, 2002). In turn, teachers tend to be more job satisfied upon observations of students' authentic learning (Shoshani and Eldor, 2016). Taken together, the DAP approach supports the fact that it has positive impact on both young children and teachers.

# THEORETICAL FRAMEWORK

fpsyg-10-02822 December 12, 2019 Time: 14:58 # 2

According to Bronfenbrenner's (1979) ecological theory, teachers' beliefs and attitudes are affected by multi-factors, including external factors such as ethnic cultural beliefs and education systems and internal factors such as personality and personal beliefs of an individual. Moreover, dynamic inter-connections exist among these systems. Before the handover of sovereignty to China, Hong Kong was a British colony characterized by a culture of "East meets West." After the handover, this cultural diversity has remained despite the fact that 92% of the population are ethnic Chinese (Census and Statistics Department, 2016). Although the Hong Kong government has intended to cultivate child-centeredness in preschool education since the past two decades, preschools tend to adopt traditional Chinese cultural values such as demands for student compliance and memorization which are regarded as barrier against childcenteredness (Rao et al., 2016). Comparable to child-centered pedagogy such as the Project approach and High Scope approach, the DAP addresses children's "social and cultural contexts" (Copple and Bredekamp, 2009) that is argued to allow more contextual flexibility for teachers to meet children's learning needs and thus believed to mismatch, at least to some extent, with the local cultural values.

Setting aside the influence of cultural values, internal personality has been found to be specifically attuned to teaching beliefs and practices among pre-service teachers. Pre-service teachers are referred to those student teachers who undergo full-time training at a teacher education institution in order to attain the Qualified Teacher Status from the Education Bureau (2015), whereas in-service teacher are those qualified teachers who works currently in preschools. Ripski et al. (2011) found that the Extroversion trait of pre-service teachers negatively predicted student–teacher interaction quality, whereas Agreeableness and Openness were found to be related to teachers' beliefs about students' diversity (Garmon, 2004; Unruh and McCord, 2010). Similarly, Decker and Rimm-Kaufman (2008) indicated that open and liberal teachers tended to encourage more students' creativity and diverse thinking in classrooms. In preschool settings, McMullen (1999) found that preschool teachers characterized by an internal locus of control tended to prefer the DAP approach. Unfortunately, it is imperative to note that the majority of research that has been studied was mainly conducted among elementary and secondary teachers in the West; few such research indeed has been conducted in preschool settings, especially in the Chinese context.

This research adopts the Myers–Briggs Type Indicator (MBTI, Myers and McCaulley, 1985) to assess teachers' personality types. While most personality inventories such as the five-factor model of personality (FFM, McCrae and Costa, 1999) lack a theoretical framework, the MBTI adopts Jung's theory of psychological types (Jung, 1921/1971) as its theoretical background. The MBTI measures four pairs of personality dimensions, including the Extroversion-Introversion, Sensing-Intuition, Feeling-Thinking, and Judging-Perceiving dimensions (see Wong, 2012 for details). Jung (1921/1971) advocated that personality types are analogous to a person's preferences for left-handedness or right-handedness. In other words, within the same personality dimension, one personality type can be dominant over the other personality types. Although the MBTI has been criticized of its operationalization of dichotomized type, it is a popular personality instrument (Furnham et al., 2003). Among all MBTI personality type dimensions, the Extroversion-Introversion, Sensing-Intuition and Feeling-Thinking dimensions are believed to be related to openness and sensitivity to children's learning and development needs. Research has shown that Extroversion, openness, agreeableness and conscientiousness of the FFM were positively related to extroversion, intuition, feeling and judging personality types of the MBTI, respectively (McCrae and Costa, 1989; Furnham et al., 2003). In terms of vocational personality, Wong and Zhang (2013) found that the personality types of preschool teachers are predominantly characterized by the Extroversion (60%), sensing (91%), feeling (65%), and Judging (79%) types. While Wong's study mainly examined in-service teachers, this research aimed to examine both pre-service and inservice teachers' personality types. Apart from personality, other teachers' personal characteristics such as age and field experiences (Kim, 2011), learning experiences from past schooling (Saban, 2003), school type and policy (Friedrich and Hron, 2011), and teaching specialty (Llurda and Lasagabaster, 2010) have been shown to impact on teachers' teaching beliefs. These variables are anticipated to contribute to teachers' teaching beliefs.

Three research questions and corresponding hypotheses are proposed:


Hypothesis 1: Teachers of the extraverted types would show higher scores of DAP than teachers of the introverted types, respectively.

Hypothesis 2: Teachers of the intuitive types would show higher scores of DAP than teachers of the sensing types.

(3) Do teachers' demographic characteristics contribute to beliefs and practice of DAP?

# METHOD

#### Design

This research adopted a cross-sectional quantitative approach.

#### Sample

A random sample of 1180 questionnaires package were distributed and eventually 544 returned (response rate: 47.2%). Though the response rate was moderately low, it was comparable to that of the average figure reported by Baruch (1999). Of all 544 participants, 530 (97.4%) and 14 (2.6%) were female and male, 156 and 401 were Bachelor of Education (B. Ed, Early childhood education) degree students (in-service) and Higher Diploma (HD, Early childhood education) (pre-service), respectively. All participants had prior knowledge about DAP from their learning curriculum at The Education University of Hong Kong. Their mean age and mean total teaching experience were 24.18 years (age range from 18 to 53, median age = 23) and 2.77 years, respectively.

### Measures

#### The Teacher Beliefs and Practices Survey (Burts et al., 2000; Leung, 2012)

fpsyg-10-02822 December 12, 2019 Time: 14:58 # 3

The Teacher Beliefs and Practices Survey (3–5 years old), originally designed by Burts et al. (2000) and then modified by Leung (2012), was adopted to measure preschool teachers' beliefs and practices. The modified versions were back translated to Chinese and then reviewed after a pilot trial by the researcher.

#### **The teacher beliefs survey (TBS, Leung, 2012)**

The TBS is a 19-item scale that measures teachers' beliefs about DAP, with satisfactory reliability and validity. Leung demonstrated that the TBS comprised three factors, namely the Developmentally Appropriate Belief (DAPB, 9 items), the Developmentally inappropriate Practice Belief (DIPB, 6 reverse items), and the Family, Culture and Inclusion (FCI, 4 items) that matched with the original versions. After reverse scoring of the DIPB, participants that have higher scores show stronger beliefs in DAP. Kim (2005)reported that the Cronbach's alphas of DAPB, DIPB, and FCI were 0.85, 0.82, and 0.81, respectively, whereas in the present research were 0.67, 0.66, and 0.69. All items are written in a 5-point Likert format: 1 = Not at all important, 2 = Not very important, 3 = Fairly important, 4 = Very important, and 5 = Extremely important. Sample items for each subscale are:

For DAPB: It is ( ) to read stories daily to children, individually and/or on a group basis.

For DIPB: It is ( ) for teachers to regularly use punishments and/or reprimands when children aren't participating.

For FCI: It is ( ) for parents/guardians to be involved in ways hat are comfortable for them.

#### **The instructional activities scale (IAS, Leung, 2012)**

The IAS is a 19-item scale that assesses teachers' practices about DAP, with satisfactory reliability and validity. Leung demonstrated that the IAS comprised four factors namely, the Developmentally Appropriate Practices Principles (DAPP, 8 items), Developmentally Appropriate Practices Activities (DAPA, 5 items), Developmentally Inappropriate Practices Activities (DIPA, 4 reverse items), and Developmentally Inappropriate Classroom Practices (DICP, 2 reverse items) that matched with the original versions. After reverse scoring, participants that have higher scores show stronger beliefs in related appropriate practices. In addition, Kim (2005) reported that the Cronbach's alphas of DAPP, DAPA, DIPA, and DICP were 0.82, 0.76, 0.73, and 0.59, whereas in the present research were 0.69, 0.64, 0.65, and 0.67. All items are written in a 5-point Likert format, 1 = Almost never (less than monthly), 2 = Rarely (monthly), 3 = Sometimes (weekly), 4 = regularly (2–4 times a week), and 5 = Often (daily). Sample items for each subscale are:

For DAPP: Solve real math problems using real objects. . .

For DAPA: Draw, paint, work with clay,. . . For DIPA: Experiment with writing. . . For DICP: Participate in whole-class, teacherdirected instruction.

#### The MBTI—Form G (Myers and McCaulley, 1985; Myers et al., 2003)

The MBTI is a self-report and forced-choice inventory that comprises four personality dimensions, including the Extroversion-Introversion (21 items), Sensing-Intuition (26 items), Thinking-Feeling (23 items), and Judging-Perceiving (24 items). For each item, the respondents are allowed to choose either one or two options that best reflected their feelings and attitudes in daily life. The scoring instruction of MBTI in this research adopts the continuous scoring method (Myers and McCaulley, 1985). For each personality dimension (e.g., Extroversion-Introversion), a lower score indicates the dominance of the first personality type (i.e., Extroversion) over the second personality type (i.e., Introversion). The MBTI has been widely used (Furnham et al., 2003) and shows acceptable reliability (Capraro and Capraro, 2002) and validity (Carlson, 1985). In Hong Kong, Wong and Zhang (2013) reported that the Cronbach's alphas for the four personality dimensions of Extroversion-introversion, sensing-intuition, thinkingfeeling, and judging-perceiving were 0.74, 0.66, 0.65, and 0.69, respectively, whereas in the present research were 0.76, 0.62, 0.72, and 0.67. Sample items for each dimension are:

For extraversion-introversion: Are you usually (A) a "good mixer," or (B) rather quiet and reserved?

For sensing-intuition: Do you usually get along better with (A) imaginative people, or (B) realistic people?

For thinking-feeling: Are you more careful about (A) people's feelings, or (B) their rights?

For judging-perceiving: When you go somewhere for the day, would you rather (A) plan what you will do and when, or (B) just go?

# Data Collection

Prior to data collection, ethical approval was obtained from the Research Ethical Committee at The Education University of Hong Kong. A random class list of students from B. Ed (Parttime) and HD (Full-time) was generated from the computer. For each class, concerned lecturers were approached by the research assistant and explained the purpose of research. A student list that showed those randomly selected students for that class was provided to the lecturer for dispatching the questionnaires package. Each package of questionnaires included the consent letter, the demographic sheet, the TBS, the IAS, the MBTI, and a returned envelope. All inventories would take 40 min to finish. Students were asked to complete the questionnaires at home. After completing the questionnaires, students were requested to post back to the researcher using the returned envelope. Reminder emails were sent to selected students 2 days and 4 days after receiving the questionnaires package.

#### Data Analyses

fpsyg-10-02822 December 12, 2019 Time: 14:58 # 4

For research question 1, the frequency distributions (in percent) was used to show and compare the dominance of personality types between teachers from the two Programmes. Thus, the higher the percentage, the larger the dominance. For research question 2, independent t-test was used to examine difference(s) in the scores of teaching beliefs and practices between personality types. For research question 3, the linear regression was used that included a two-step single-level analysis initially entering teachers' demographic characteristics, followed by the addition of personality dimensions factors in Step 2. A negative beta value indicates the dominance of the first personality type over the second one and vice versa. The rationale was to examine the differential contributions made by teachers' demographics such as year of working experience and age, and four dimensions of personality types on teachers' DAP scores.

#### RESULTS

As shown in **Table 1**, findings indicated that the personality profiles of pre-service (HD) and in-service (B. Ed) teachers were similar in each personality dimension and characterized, respectively, by the dominance of Extroversion (52.6%)- Sensing (78.4%)-Feeling (62.0%)-Judging (82.6%) and Extroversion (51.4%)-Sensing (86.7%)-Feeling (64.0%)-Judging (83.4%), respectively.

Because both in-service and pre-service teachers indicated very similar patterns of personality types, the data were combined for further analyses. As shown in **Table 2**, Extraverted teachers showed higher scores of TBS (t = 2.50, p < 0.01), DAPB (t = 2.64, p < 0.01), DAPP (t = 1.99, p < 0.05) and DAPA (t = 2.59, p < 0.01) than that of the Introverted teachers. Intuitive teachers displayed higher scores of TBS (t = −2.87, p < 0.01), DIPB (t = −2.16, p < 0.05), and DICP (t = −2.89, p < 0.01) than that of the Sensing teachers. Teachers of the Perceiving types showed higher scores of DAPB (t = −1.99, p < 0.05) than teachers of the Judging types.

The regression findings as shown in **Table 3**, indicated that the Extroversion-Introversion dimension (model 2: β = −0.11, t = −2.46, p < 0.05) predicted DAPB as a single predictor,

TABLE 1 | Distribution of kindergarten teachers' personality types from B. Ed and HD programmes.


B. Ed, Bachelor of Education (Early childhood education) Part-time programme, HD (ECE), Higher Diploma (Early childhood education) Full-time programme. Personality types in italics are predominant.

whereas the Sensing-Intuition dimension (model 2: β = 0.20, t = 1.94, p = 0.05) was significant predictor of TBS total. For the IAS, teaching experience predicted significantly DAPP (model 1: β = 0.29, t = 4.29, p < 0.001), DAPA (model 1: β = 0.16, t = 2.21, p < 0.05), DIPA (model 1: β = −0.17, t = −2.44, p < 0.05) and IAS total (model 1: β = 0.17, t = 2.39, p < 0.05). The Extroversion-Introversion was a significant predictor of DAPP (model 2: β = −0.13, t = −3.13, p < 0.001), DAPA (model 2: β = −0.14, t = −3.14, p < 0.001), DIPA (model 2: β = 0.09, t = 2.15, p < 0.01), and IAS total (model 2: β = −0.09, t = −2.03, p < 0.05) when teaching experience remained significant. The Sensing-Intuition dimension (model 2: β = 0.16, t = 3.42, p < 0.001), and the Judging-Perceiving dimension (model 2: β = −0.10, t = −1.98, p < 0.05) predicted DICP, respectively.

# DISCUSSION

This research aimed to examine how preschool teachers' personality related to their beliefs of DAP. Findings showed that the personality type profiles of preschool teachers were characterized predominantly by Sensing-Feeling-Judging. Among all personality types, extroverted and intuitive teachers tended to show higher scores in DAP beliefs and practice. In addition, teaching experiences associated with personality predicted DAP practices. The followings discuss current findings and its implications to the field:

# Robustness of Preschool Teachers' Personality Types

The frequencies of teachers' personality types from the two different teacher training programmes are similar to a large extent, showing that preschool teachers are characterized predominantly by the personality types of Sensing-Feeling-Judging. The findings are comparable to that of the previous study by Wong and Zhang (2013). Taken together, consistent evidences have provided strong support to the robustness of preschool teachers' personality types. According to Myers and McCaulley (1985), the personality types of Sensing-Feeling-Judging are characterized by preferences of following instructions in a structured work environment and caring for others but are not comfortable with applying DAP that taps on liberal and flexible thinking. Thus, this robustness of personality types that indicates the relative weakness of openness and flexibility among preschool teachers is argued to act as a barrier to implement the DAP in schools. Wong and Zhang (2013) argued that intuitive teachers show leadership capability and incline to be more comfortable with planning and implementing creative curriculum in preschool settings. Furthermore, it is interesting to note that the frequency of intuitive pre-service teachers (21.6%) was higher than that of in-service teachers (13.3%), suggesting that there is greater propensity of the former to adopt the DAP beliefs, whereas more professional training programmes in DAP are necessary for in-service teachers. As such, teacher training institutions are recommended to strengthen pre-service teachers' understanding of DAP and enhance self-reflection with respect to their personality characteristics.

fpsyg-10-02822 December 12, 2019 Time: 14:58 # 5


TBS, teacher beliefs survey, DAPB, Developmentally Appropriate Belief; DIPB, Developmentally inappropriate Belief; FCI, Family, Culture and Inclusion; IAS, Instructional Activities Scale; DAPP, Developmentally Appropriate Practices Principles; DAPA, Developmentally Appropriate Practices Activities; DIPA, Developmentally Inappropriate Practices Activities; DICP, Developmentally Inappropriate Classroom Practices, E, Extroversion; I, introversion; S, Sensing; N, intuition; T, thinking; F, feeling; J, judging; P, perceiving. ∗∗p < 0.01; <sup>∗</sup>p < 0.05.

TABLE 3 | Linear regression of predictors of teaching beliefs and practices.


TBS, teacher beliefs survey, DAPB, Developmentally Appropriate Belief; DIPB, Developmentally inappropriate Belief; FCI, Family, Culture and Inclusion; IAS, Instructional Activities Scale; DAPP, Developmentally Appropriate Practices Principles; DAPA, Developmentally Appropriate Practices Activities; DIPA, Developmentally Inappropriate Practices Activities; DICP, Developmentally Inappropriate Classroom Practices; E-I, Extroversion-Introversion; S-N, Sensing-Intuition; T-F, Thinking-Feeling; J-P, Judging-Perceiving; t\_exp, teaching experience. ∗∗∗p < 0.001, ∗∗p < 0.01, <sup>∗</sup>p < 0.05.

# Extraverted and Intuitive Teachers Tend to Show Stronger Beliefs in DAP

As expected, extraverted teachers and intuitive teachers tended to exhibit stronger beliefs in DAP than that of the introverted teachers and sensing teachers, respectively (a lower score indicates the dominance of the first personality type (i.e., Extroversion) over the second personality type (i.e., Introversion) and vice versa). Therefore, both hypotheses are accepted. These findings support that notion that teachers' personality that are characterized by openness and flexibility are more willing to adopt and practice the DAP. In preschools, young children tend to find teachers who are communicative and flexible more approachable (Saracho, 2003). It follows that extraverted and intuitive teachers are more inclined and able than that of their introverted and sensing counterparts to implement DAP in classrooms. As such, it may be desirable for school principals to identify teachers of extraverted and intuitive personality that match with the characteristics of the DAP for recruitment. In addition, school managers are recommended to provide individualized professional development programs that focus on understanding and practicing DAP to all teachers in their schools.

# Combined Effect of Teaching Experience and Personality on DAP Beliefs

Current findings also indicated that there was a joint effect of personality (i.e., extraverted and intuitive personality types) and teaching experiences that contributed to teachers' beliefs and practice in DAP. It is imperative to note that though the majority of preschool teachers are of the sensing types that tend not to prefer the beliefs and practices of DAP, their teaching experiences are believed to minimize this limitation to a certain extent. In this sense, positive experiences about DAP are believed to be able to enhance teachers' belief and subsequent practices (Murphy et al., 2015; Smith, 2015). The implication is that school principals and teacher training institutions can provide ongoing professional development training programs for teachers to construct their own successful teaching experiences of DAP such that teachers are able to develop substantial amount of positive experiences fpsyg-10-02822 December 12, 2019 Time: 14:58 # 6

to cultivate and maintain their DAP beliefs. In addition, school managers are suggested to maintain the stability of teaching team so that experienced and intuitive teachers (Rushton et al., 2007) are able to lead the development of DAP in schools.

# LIMITATIONS, RECOMMENDATIONS, AND CONCLUSION

This research adopted a cross-sectional design such that there was no causal relationship among variables to be generated. Future research is recommended to use qualitative approach such as semi-structured interviewing. Though random sampling was adopted, the sample was based on only one tertiary institution and female dominated such that any generalization of findings should be cautious. Future studies are recommended to adopt a longitudinal research design that has the strength of making inferences from cause-effect relationships among research variables. In addition, multiple measures of teachers' responses by employing physiological indicators such as cortisol, heart rate, and body temperature, as objective measurements of participants' levels of subjective well-being should be considered.

In conclusion, preschool teachers are characterized by a robustness of Sensing-Feeling-Judging personality types. In particular, extraverted and intuitive teachers tend to show stronger beliefs in DAP. Nonetheless, teaching experience is a compensatory factor in enhancing the DAP beliefs. School principals and teacher education institutions are recommended

#### REFERENCES


to provide appropriate professional development programmes to promote the readiness of preschool teachers in adopting the DAP approach in preschools.

#### DATA AVAILABILITY STATEMENT

All datasets generated for this study are included in the article/supplementary material.

# ETHICS STATEMENT

The studies involving human participants were reviewed and approved by The Education University of Hong Kong. The patients/participants provided their written informed consent to participate in this study.

#### AUTHOR CONTRIBUTIONS

The author confirms being the sole contributor of this work and has approved it for publication.

# FUNDING

This research was funded by The Education University of Hong Kong.


**Conflict of Interest:** The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2019 Wong. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fpsyg-10-02822 December 12, 2019 Time: 14:58 # 7

# Association Between Chronic Health Conditions and Quality of Life in Rural Teachers

#### Pablo A. Lizana<sup>1</sup> \*, Gustavo Vega-Fernandez<sup>1</sup> and Lydia Lera<sup>2</sup>

<sup>1</sup> Laboratory of Morphological Sciences, Instituto de Biología, Pontificia Universidad Católica de Valparaíso, Valparaíso, Chile, <sup>2</sup> Public Nutrition Unit, Institute of Nutrition and Food Technology, University of Chile, Santiago, Chile

Aim: The school teacher profession has been reported to be associated with an increased rate of health problems that can affect the quality of life (QoL) of teachers. However, there is little information about rural teachers.

Objective: To investigate the associations of obesity, abdominal obesity, and hypertension with the perception of QoL in rural teachers.

#### Edited by:

Wong Yau Ho Paul, Tung Wah College, Hong Kong

#### Reviewed by:

Manpreet Kaur Bagga, Partap College of Education, India Izabella Uchmanowicz, Wrocław Medical University, Poland

> \*Correspondence: Pablo A. Lizana pablo.lizana@pucv.cl

#### Specialty section:

This article was submitted to Educational Psychology, a section of the journal Frontiers in Psychology

Received: 23 July 2019 Accepted: 06 December 2019 Published: 09 January 2020

#### Citation:

Lizana PA, Vega-Fernandez G and Lera L (2020) Association Between Chronic Health Conditions and Quality of Life in Rural Teachers. Front. Psychol. 10:2898. doi: 10.3389/fpsyg.2019.02898 Materials and Methods: This cross-sectional study included a representative sample of teachers from eight rural schools in the Valparaíso Region of Chile. Obesity was evaluated by the percentages of fat mass (% FM) and abdominal obesity, and hypertension was recorded. The Short Form-36 Health Survey was administered to subjects to evaluate QoL. A logistic regression adjusted for age and gender was used to associate QoL with chronic health conditions.

Results: Eighty-five percent of teachers presented abdominal obesity, 56% presented obesity evaluated by % FM and 33% presented hypertension. Thirty percent of teachers in the ≥45-year-old age group presented all three chronic conditions simultaneously. Teachers presenting abdominal obesity in the ≤44-year-old age group had lower scores in the mental health component (MCS) (p < 0.05) of the survey than the teachers in the ≥45-year-old age group. In addition, teachers in the ≤44-year-old age group who presented obesity scored lower in the MCS than those in the same age group without obesity (p = 0.004). The occurrence simultaneous of two and three chronic health conditions increased the risk of reduced MCS significantly (p = 0.015, OR 17.0, 95% CI: 1.741–165.90 and p = 0.003, OR 82.6, 95% CI: 4.58–1490.70, respectively). In addition, ages under 45 years old were associated with low score of MCS (OR 8.8, 95% CI: 1.565–49.698).

Conclusion: A combination of chronic conditions affects the mental health (MH) of rural teachers. Although the association does not imply causation, these findings support the notion that teachers present early MH problems. This suggests that resources must be made available to detect early mental and chronic health conditions of rural teachers.

Keywords: mental health, physical health, quality of life, school teachers, obesity, hypertension

# INTRODUCTION

fpsyg-10-02898 January 10, 2020 Time: 16:10 # 2

According to the international labor organization (ILO) and the world health organization (WHO), a good occupational health (OH) and quality of life (QoL) standard is achieved with the promotion and maintenance of mental, physical and social wellbeing in all workplaces (OIT, 2003; Bagtasos, 2011). In this sense, one of the professions associated with a large decrease in OH worldwide is the school teacher (Travers and Cooper, 1993; Johnson et al., 2005; Innstrand et al., 2011); teaching is associated with significant increases in the deterioration of mental health (MH) and physical discomfort during the practice of the profession (Porto et al., 2006; Bauer et al., 2007; Fernandes and Rocha, 2009; Jurado et al., 2019), as well as subsequent psychosocial deterioration (Bringi and Ranbhare, 2015) and QoL disorders due to occupational stress (Yang et al., 2009). It should be noted that one of the problems that contributes to further deterioration of the QoL is the considerable workload outside of the corresponding schedule (OIT, 2003; Bravo, 2005). The United Nations educational, scientific, and cultural organization (UNESCO) has reported similar conditions in Latin American countries such as Mexico, Chile, Peru, Argentina, Ecuador and Uruguay; Chile is ranked first for teachers who work more than 40 h per week (Cuenca et al., 2005).

Some results in Chile reflect the global situation, where the teaching workload is related to high psychosocial and physical risks and causes a considerable deterioration of the QoL (Cabrera et al., 2003; Cuenca et al., 2005). It is also important to point out that these professionals have a higher disease recurrence rate than the country average (Chávez, 2009), an aspect that has also been reported in other countries (Unterbrink et al., 2008; Bogaert et al., 2014).

The population of teachers in Chile belongs almost entirely to the age ranges of 25–44 and 45–64 years old, which are two of the four age categories studied by the National Health Survey of Chile (Ministerio de Salud de Chile, 2010a). The majority of the working population, particularly those that present diverse health risks due to the prevalence of some chronic noncommunicable diseases (CNCDs) related to nutritional status and body composition, is concentrated in these two age ranges. For example, obesity according to body mass index (BMI) (≥25) has a considerable prevalence of 72.26% in both age categories, and a prevalence of 43.8% for high blood pressure has been reported for those 45–64 years old (Ministerio de Salud de Chile, 2010a). In this sense, teachers ≥45 years old have a lower working capacity than teachers ≤45 years old (Freude et al., 2005). Therefore, identifying factors that influence health by age category is relevant for the development of public policies.

In Chile, the rates of overweight, obesity, and abdominal obesity are higher in women from rural areas than in women from urban areas (Ministerio de Salud de Chile, 2010a), which is an important factor since females are mainly employed in the teaching profession (Cabrera et al., 2003; Cuenca et al., 2005). In addition, chronic conditions significantly affect the perception of QoL (Vinaccia and Orozco, 2005; Trevisol et al., 2011; Pimenta et al., 2015), and therefore a decreased QoL may also be present in the population of teaching professionals. However, among the studies on Chile's workforce, there have been no reports of research on CNCDs in teachers and the impact of these conditions on the perception of QoL in rural teachers.

# MATERIALS AND METHODS

# Participants

The target population was all teachers working in eight rural primary schools (N = 97) in rural establishments in the communities of Calera (32◦ 47'0"S, 71◦ 130 0 <sup>00</sup>W) and Hijuelas (32◦ 480 0 <sup>00</sup>S, 71◦ 10'0"W), Valparaíso, Chile (MINEDUC, 2016). To calculate the sample size, the variable with the greatest variance for this group was selected according to the literature published at the time of planning of this study. The sample was determined with the variable BMI (Ministerio de Salud de Chile, 2010a) for adults of the Valparaíso region. The sample was calculated with 95% confidence and a 5% error. The minimum sample size was calculated as 69 teachers.

Sampling was conducted between April and July 2016. Eighty teachers agreed to participate in the study. The final sample consisted of 70 teachers; 10 teachers were excluded for the following reasons: incomplete data (5), pregnancy (1), and sick leave at the time of the evaluation (4). The teachers presented ages ranging from 24 to 64 years, with an average age of 40.78 ± 12.57 years; 68.6% of the teachers were women. All teachers were evaluated in the same educational establishments during morning sessions.

#### Instruments

#### Anthropometry, Body Composition, and Cardiovascular Risk

The weight and height of the subjects were evaluated to determine the BMI (height in meters/kilograms squared), which was categorized as underweight (BMI < 18.5 kg/m<sup>2</sup> ), normal weight (18.5 kg/m<sup>2</sup> ≤ BMI < 25 kg/m<sup>2</sup> ), overweight (25 kg/m<sup>2</sup> ≤ BMI < 30 kg/m<sup>2</sup> ), or obese (BMI ≥ 30 kg/m<sup>2</sup> ). To evaluate obesity based on the percentage of fat mass (% FM), a bioelectrical impedance device was used (TANITA BC 420 SMA, Tanita, Tokyo, Japan); the subjects were asked not to carry metallic objects, drink alcohol 48 h before the evaluation, perform intense exercise 12 h before the evaluation, eat or drink, especially caffeine or diuretics, 4 h before the evaluation, and were asked to urinate before the evaluation. To classify obesity by % FM, the recommendation by the "American Society of Endocrinologists" was adopted; the cutoff points were >35% in women and >25% in men (Flegal et al., 2012).

The waist circumference (abdominal fat marker) was measured abdominally at the level of the iliac crest and after a normal exhalation while the subject was in a standing position. The tape measure (LufkinTM) was adjusted without compressing the structures. The protocol was outlined in the Clinical Guidelines for Obesity of the United States National Institute of Health (NIH) and was replicated in the National Health Survey from Chile 2003 and 2009–2010 (Ministerio de Salud de Chile, 2010a). The suggested cutoff points were ≥102 cm for men and ≥88 cm for women.

Blood pressure was measured using automated sphygmomanometers (OMRON HEM-705CPINT; Omron Co., Kyoto, Japan). The evaluation protocol involved placing the subject in a sitting and relaxed position for at least 5 min with the left upper limb supported on a table at the level of the heart; the cuff was then placed and adjusted to the perimeter of the subject's arm. A classification of high blood pressure was a systolic pressure ≥ 140 mmHg and a diastolic pressure ≥ 90 mmHg, according to the clinical guidelines of arterial hypertension outlined by the Ministry of Health of Chile (Ministerio de Salud de Chile, 2010b). For subjects who presented high blood pressure, the procedure was repeated after a ten-minute rest, and the average of three readings performed at an interval of 2 min was considered, similar to the procedure performed in the National Health Survey 2009–2010 (Ministerio de Salud de Chile, 2010a).

#### Quality of Life Questionnaire

The QoL of the teachers was evaluated using the Short-Form 36 Health Survey (SF-36) questionnaire, which is an instrument developed in the United States to evaluate QoL-related health in adults (Ware and Sherbourne, 1992); the SF-36 has been adapted syntactically and semantically for the Chilean population (Olivares, 2006). The questionnaire includes 36 items measured on a Likert scale. The 36 items are grouped into 8 health topics: physical function (FF), physical role (PR), body pain (BP), general health perception (GH), vitality (VT), social function (SF), emotional role (ER), and MH. The 8 topics can be categorized into two components, the physical component summary (PCS) and the mental component summary (MCS). For each scale, T scores, with an average of 50 and a standard deviation of 10, were calculated (McDowell and McDowell, 2006).

#### Procedure

Procedures for anthropometric measurements, body composition analysis using bioimpedance, and the SF-36 questionnaire were orally explained to the teachers. In addition, the subjects signed an informed consent form to indicate that they understood the scope of the investigation. Participation in the study was voluntary and the anonymity of teachers was guaranteed. This study was approved by the Ethics Committee of the Pontificia Universidad Católica de Valparaíso in accordance with the Declaration of Helsinki (The World Medical Association, 2009).

#### Data Analysis

Statistical analyses were performed using STATA 15 software (2017, Stata Corp. LLC, College Station, TX, United States). In this cross-sectional study, descriptive analyses are presented as the mean, standard deviation (mean ± SD) and percent according to age category (≤44 years old and ≥45 years old), cutoff scores for age were obtained from the National Health Survey of Chile of Chile (Ministerio de Salud de Chile, 2010a). Comparisons of anthropometric measurements, body compositions and each item of the QoL questionnaire between the groups were carried out using specifics test (t-test or its non-parametric equivalent according to Shapiro Wilk's normality tests). The chi-square (χ2) test was used to evaluate the associations between obesity by BMI, obesity by % FM, abdominal obesity and blood pressure by age category, as well as to evaluate their associations with chronic conditions (obesity by % FM, abdominal obesity and arterial hypertension). The 25th percentiles (25p) of each summary factor (MCS and PCS) were used as a cut-point to dichotomize the data. The teachers were classified in low (below 25p) or fair/good (above 25p) categories. Logistic regression analysis was performed to estimate the association between low scores of MCS and PCS factors with chronic conditions adjusted by age and gender. To verify the accuracy of fit of the model, the Hosmer-Lemeshow test was applied.

#### RESULTS

#### Characteristics of the Subjects

The total number of subjects participating in this study, their sociodemographic characteristics, and their health characteristics, including body composition, hypertension status, and chronic health conditions, are shown in **Table 1**. The total sample of the evaluated teachers was 70 individuals (68.57%, women), with an average age of 40.79 ± 12.58 years; the age categories of ≤44 years old and ≥45 years old represented 61.42% and 38.57% of the sample population, respectively, resulting in an average age of 31.62 ± 4.94 years for the first group and 55.37 ± 4.53 years for the second group.

# Body Composition and Chronic Conditions

The study sample did not present significant differences between the two age categories in terms of weight (p = 0.209) and height (p = 0.132). Regarding BMI, both age groups showed significant differences (p = 0.022), with the second group having the highest BMI (28.55 ± 3.64 vs. 26.32 ± 4.00 kg/m<sup>2</sup> ). In turn, the association between age and BMI was also significant (p = 0.028); both age groups had a high prevalence of overweight, at 41.86% (≤44 years old) and 51.85% (≥45 years old), with an evident tendency for BMI to increase with age. Additionally, it was observed that the highest prevalence of obesity was found in the older group, at 18.60% vs. 37.03% in the younger group (**Table 1**).

Waist circumference was significantly greater in the older age group than in the younger age group (p = 0.010), and a majority (85.71%) of the total sample presented abdominal obesity. In addition, waist circumference associated with abdominal obesity was more prevalent in subjects ≥45 years old (79.6% in the younger age range vs. 96.29% in the older age range, p = 0.045).

Fat mass (kg) was not significantly different between the age categories (p = 0.055), nor was the variable of fat-free mass (FFM, kg) (p = 0.542). However, the results of% FM showed a significant difference (p = 0.022) between the age two ranges; % FM was higher in the older age group than in the younger age group (**Table 1**). In addition, the results suggested a significant association (p < 0.001) between the age of the subjects and the


TABLE 1 | Sociodemographic and health characteristics of rural teachers in Valparaiso, Chile stratified by age group.

BMI, body mass index; SD, standard deviation; FFM, free fat mass (kg & %). (≤44–≥45), age categories (years). P < 0.05. <sup>a</sup> t-test. <sup>b</sup>Mann-Whitney-Test. <sup>c</sup>Chi– squared. <sup>c</sup> ∗ Fisher's exact test. <sup>d</sup>Body components evaluated by bioimpedance (TANITA BC 420SMA). <sup>e</sup>Data are expressed as frequency (percentage). <sup>f</sup>Chronic conditions: hypertension, central obesity, and obesity.

presence of obesity, with the prevalence of obesity increasing from 39.53% in the younger group to 81.48% in the older group. The % FFM was significantly different between the age groups and was higher in subjects ≤44 years old than in subjects ≥45 years old (68.89 ± 9.05 vs. 63.63 ± 8.18, p = 0.017).

Systolic blood pressure was significantly higher (p < 0.01) in the older age group. Diastolic pressure was not significantly different between the two age groups (p = 0.302); however, these age ranges had a significant association (p = 0.008) with hypertension, which tends to occur with age. The older age group had a hypertension prevalence rate of more than 50%.

In the combined analysis of obesity by % FM, abdominal obesity and hypertension in relation to the presence of one, two, or all three chronic conditions in the respective age groups, significant associations were observed with the presence of all three chronic conditions in the older age category (4.56% vs. 29.62%, p = 0.010) (**Table 1**).

### Perception of Quality of Life

**Table 2** shows a comparison of the average scores of each of the eight scales and the two summary measures of the SF-36 survey; results are stratified by age and the presence and absence of each chronic condition. For the subjects in the younger age group, there were significant associations between the presence and absence of abdominal obesity and the QoL scores for both the MCS measures (p = 0.001) and four of the eight survey scales (PF, RE, BP, and GH; p = 0.046, p = 0.018, p = 0.001, and p = 0.042, respectively), while for the older age group, there were no significant differences. In the overall sample of teachers, significant differences were also shown; Abdominal obesity was associated with a deterioration in the QoL in four of the eight scales of the survey (PF, RE, BP, and GH; p = 0.014, p = 0.027, p < 0.001, and p = 0.034, respectively) as well as the MCS (p = 0.012). Regarding hypertension, there were no significant differences between age groups for conditions associated with QoL scales. For the chronic condition of obesity in the younger age group, significant differences were found between the presence and absence of obesity for one of the eight QoL scales (RE; p = 0.026), as well as for the MCS measure (p = 0.004). No significant differences were observed for the older age group. Nonetheless, in the total sample, there were significant differences in obesity for three scales of the QoL survey (PF, RE, and MH; p = 0.035, p = 0.044, and p = 0.036, respectively); however, none were observed for the summary measures.

**Table 3** shows that% FM and abdominal obesity have significant associations with a poor perception of QoL in the MCS in the younger age group (p < 0.005). In addition, abdominal obesity also presented a significant association with the PCS in the ≤44-year-old age group (p = 0.037).

The results for the association of 25th percentiles of MCS and PCS with chronic health conditions variables adjusted by age and gender are given in **Table 4**. The association between chronic health conditions and PCS was not significant. However, the occurrence simultaneous of two and three chronic health conditions increased the risk of reduced MCS significantly (p = 0.015, OR 17.0, 95% CI: 1.741–165.90 and p = 0.003, OR 82.6, 95% CI: 4.58–1490.70, respectively). In addition, ages under 45 years old were associated with low MCS (p = 0.014, OR 8.8, 95% CI: 1.565–49.698), but the association with gender was insignificant.

### DISCUSSION

This study found that teachers from rural locations had a high prevalence of CNCDs and that each of the studied CNCDs increased in subjects ≥45 years old. The prevalence of overweight at ages 25–44 and 45–64 years, as reported in the National Health Survey of Chile (Ministerio de Salud de Chile, 2010a), decreased slightly from 44% to 41%. However, we observed an increase in overweight due to BMI (51.85% of the subjects) in the older age group. Regarding obesity measured using BMI, this research showing that the older age group had the highest prevalence (present study 37.03%, 35.8%) were similar to the National Health Survey of Chile (Ministerio de Salud de Chile,

TABLE 2 | Comparison of the scores on the eight scales of the SF-36 QoL questionnaire in subjects with obesity, abdominal obesity, and hypertension stratified by age group.


QoL, quality of life; PF, physical function; RP, role limitations due to physical problems; RE, role limitations due to emotional problems; BP, bodily pain; SF, social functioning; MH, mental health; VT, vitality; GH, general health perception; P < 0.05. <sup>∗</sup> t- test. <sup>a</sup>≤44–≥45, age categories (years).

2010a). A high prevalence of overweight and obesity evaluated by BMI in teachers was previously been reported for Chilean urban teachers (Cabrera et al., 2003; Kain et al., 2010), but the results were not stratified by age category. The usefulness of obesity evaluated by BMI is limited because it does not consider differences in corporal components (Prado et al., 2015). For this reason, in the present study, it was decided to define obesity according to % FM. When obesity was measured by % FM, the obesity prevalence in rural teachers in the older age group 81.48%, whereas when obesity was evaluated by BMI, the prevalence in the older age group was 37.03%; therefore, there could be an underestimation of obesity in working-age population studies that use BMI.

Teachers belonging to the second age group (≥45 years old) showed a higher prevalence (51.85%) of high blood pressure than teachers in the same age group at the country level (43.8%) in the National Health Survey of Chile (Ministerio de Salud de Chile, 2010a). This coincides with Cabrera et al. (2003), who also reported a higher prevalence of hypertension in teachers than that in the reference group. Regarding abdominal obesity and its categories stratified by age group in the National Health Survey of Chile (Ministerio de Salud de Chile, 2010a), a prevalence of 62.4% was observed for the group aged 25–44 years and 80.6% for the group aged 45–64 years. These values were far less than the values reported in the present study, with a prevalence of 79.6% in the younger group and 96.29% in the older group. These results indicate that cardiometabolic risk and metabolic syndrome are important factors (Despres et al., 2008).

Chronic conditions in rural teachers were mainly associated with the older age group, but the QoL MCS was affected by age. In Chile, teachers have a high workload that includes lessons and administration, as well as a series of working hours outside the classroom, which may include test review, planning, and educational material preparation. Additionally, tasks such as interacting with parents, solving school problems (inside and outside the classroom), and building relationships with peers, among others, may require an important social or emotional component (Cuenca et al., 2005). Therefore, this work overload may affect QoL, especially the MCS, of teachers, as observed in this study. Bauer et al. (2007) applied a general health

TABLE 3 | Comparison of physical component and mental component summary measures for subjects with three chronic health conditions stratified according to age group.


PCS, physical component summary; MCS, mental component summary; SD, standard deviation. <sup>a</sup>≤44–≥45, age categories (years). <sup>∗</sup>Mann-Whitney-Test. P < 0.05.

TABLE 4 | Logistic regressions for the association of 25th percentiles of MCS and PCS with chronic health conditions adjusted by age and gender.


PCS, physical component summary; MCS, mental component summary. <sup>a</sup>Odds Ratios [Confidence interval]. P < 0.05.

questionnaire and found that the MCS was affected in 30% of German urban school teachers. Therefore, it is important to generate mitigation strategies, such as social networks inside and outside the classroom, to promote MH in teachers.

In addition, in this study, a low score in the MCS of the QoL survey was associated with the simultaneous presence of two and three chronic health conditions in rural teachers, even after adjustment for gender. We adjusted for gender because the teachers who participated in this study were mostly women (68.57%), an aspect that coincides with previous studies in Chilean school teachers (Cabrera et al., 2003; Cuenca et al., 2005). The significant associations between the MCS and the chronic conditions indicate that teachers should have ways to prevent and identify MH issues and strategies to improve their health, thus improving their QoL. Interestingly, teachers perceive MCS deterioration but not PCS deterioration, which can be related to the varied high-stress tasks that the teaching profession entails in Chile. Accordingly, Chile has one of the highest medical leave rates in South America (Cuenca et al., 2005), and these medical leave rates are often associated with emotional fatigue, which corresponds to the results of our work, as lower MCS scores were associated with chronic conditions.

The present study has several limitations. The first is typical of cross-sectional research; one of the instruments used, the QoL survey, was based on self-reports, so no causality can be derived from the reported associations. Nevertheless, we obtained a high response rate from teachers. Additionally, although we investigated a representative sample of rural teachers, we worked in only two rural areas in a region of Chile, which may limit the generalization of the study results to other regions. Further studies should include teachers in urban areas who may have other conditions that affect their QoL. Although there are important studies on the QoL of urban teachers (Fernandes and Rocha, 2009; Yang et al., 2009; Pizolato et al., 2013; Nusseck et al., 2018), there are few studies focusing on the health of rural teachers; therefore, the present study has the strength of being a pioneer study involving rural teachers, health and QoL.

Currently, occupational health in Chile confers safety and prevention to workers in general but not specifically to school teachers. Therefore, this research has an important implication in enhance the background on health in rural school teachers, providing a clear picture to establish possible improvements and implementation of public policies in the prevention of mental and physical health of school teachers. In this sense,

Arvidsson et al. (2016) suggests a series of measures for teachers at the social, organizational and individual levels. At the organizational and individual level that improve leadership and strengthen school teacher self-efficacy. They also suggest improving coordination between the various stakeholders, the rationalization of administrative tasks and better ergonomic conditions during work, aspects that point to the mental and physical health components. In addition, and given the lack of time that has been reported in Chilean school teachers (Cuenca et al., 2005) should include policies for balancing work life and leisure time, in order to generate healthy living spaces and reduce the prevalence of CNCDs in school teachers.

# CONCLUSION

In conclusion, chronic health conditions mainly affect the QoL MCS rather than the PCS in rural teachers. This suggests that sufficient resources must be available to detect and manage early chronic conditions in teachers in rural areas since chronic conditions are associated with the MCS. Teaching is an occupation that requires significant mental work, and chronic conditions affecting the MCS could affect work performance and thus affect the students.

# DATA AVAILABILITY STATEMENT

The datasets generated for this study are available on request to the corresponding author.

# REFERENCES


# ETHICS STATEMENT

The studies involving human participants were reviewed and approved by the Pontificia Universidad Católica de Valparaíso. The patients/participants provided their written informed consent to participate in this study.

# AUTHOR CONTRIBUTIONS

PL designed the study. PL and GV-F performed the measurements, processed the data, drafted the manuscript, and designed the tables. PL and LL performed the statistical analysis. All authors discussed the results and commented on the manuscript.

# FUNDING

This research was supported by the FONDECYT Grant No. 11170716 (FONDECYT-CONICYT, Chile) and the Vicerrectoría de Investigación y Estudios Avanzados of the Pontificia Universidad Católica de Valparaíso through project 039.346/2016.

# ACKNOWLEDGMENTS

The authors thank the directors and teachers of all the rural schools in La Calera and Hijuelas for their willingness to participate in this study.

en Argentina, Chile, Ecuador, México, Perú y Uruguay. Santiago de Chile: UNESCO.



**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2020 Lizana, Vega-Fernandez and Lera. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

fpsyg-10-02933 January 9, 2020 Time: 18:25 # 1

# Spirituality and Prayer on Teacher Stress and Burnout in an Italian Cohort: A Pilot, Before-After Controlled Study

Francesco Chirico<sup>1</sup> , Manoj Sharma2,3,4, Salvatore Zaffina1,5 \* and Nicola Magnavita1,6

<sup>1</sup> Post-graduate School in Occupational Health, Catholic University of the Sacred Heart, Rome, Italy, <sup>2</sup> Department of Behavioral and Environmental Health, School of Public Health, Jackson State University, Jackson, MS, United States, <sup>3</sup> School of Health Sciences, Walden University, Minneapolis, MN, United States, <sup>4</sup> Health for All, Omaha, NE, United States, <sup>5</sup> Occupational Health Service, Pediatric Hospital Bambino Gesù, IRCCS, Rome, Italy, <sup>6</sup> Department of Woman and Child Health and Public Health, Agostino Gemelli University Policlinic, IRCCS, Rome, Italy

Introduction: Teaching is a stressful profession that exposes workers to the risk of burnout. Techniques involving higher mental functions, such as transcendental meditation and prayer, have been used in stress and burnout prevention programs. In this study, we report the results of an experience conducted in a group of teachers of a religious institute, in which prayer was used as a technique to prevent burnout.

#### Edited by:

Paula Benevene, Libera Università Maria SS. Assunta, Italy

#### Reviewed by:

Alessandro De Carlo, Giustino Fortunato University, Italy Laura Dal Corso, University of Padua, Italy

> \*Correspondence: Salvatore Zaffina salvatore.zaffina@opbg.net

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 29 August 2019 Accepted: 11 December 2019 Published: 21 January 2020

#### Citation:

Chirico F, Sharma M, Zaffina S and Magnavita N (2020) Spirituality and Prayer on Teacher Stress and Burnout in an Italian Cohort: A Pilot, Before-After Controlled Study. Front. Psychol. 10:2933. doi: 10.3389/fpsyg.2019.02933 Methods: Fifty teachers and support staff employed at a Catholic school of a Congregation of nuns volunteered for this study. They were randomized into two groups: prayer treatment (n = 25) or control group (n = 25). The treatment protocol was based on the combination of individual Christian prayer and a focus group of prayer-reflection. The participants received two 30 min training sessions a week over 2 months. Job satisfaction, well-being, and burnout symptoms (emotional exhaustion and depersonalization sub-scales) were measured at baseline and at followup (4 months) with the Italian versions of the Maslach Burnout Inventory validated for teaching and education sector, the General Health Questionnaire, and the Warr, Cook, and Wall's Job Satisfaction Scale.

Results: At follow-up, a significant improvement of all outcome measures was observed. Emotional exhaustion (16.80–4.92, p < 0.001), depersonalization (3.72– 0.60, p < 0.001) levels, and psychological impairment (10.08–2.04, p < 0.001) were significantly decreased, and job satisfaction (45.96–77.00, p < 0.001) was increased. The effect sizes (Glass' 1) of the therapeutic interventions ranged from 0.53 (satisfaction level) to 2.87 (psychological health), suggesting moderate to large effects.

Discussion: Prayer could be effective, no less than meditation and other spiritual or mind-body techniques, in contrasting the negative effects of occupational stress and preventing burnout among teachers and possibly other human service professionals.

Keywords: clinical trial, job burnout, job satisfaction, mental health, meditation, occupational health, prayer, teachers

# INTRODUCTION

fpsyg-10-02933 January 9, 2020 Time: 18:25 # 2

Teaching is a high-stress profession, and many teachers are exposed to high levels of emotional distress at the workplace (Borrelli et al., 2014; Fiorilli et al., 2015; De Stasio et al., 2017; Herman et al., 2018; Chirico et al., 2019b). Teacher stress has been defined as the experience of unpleasant negative emotions such as anger, frustration, anxiety, depression, and nervousness, resulting from some aspect of work (Kyriacou, 2001). Teacher burnout has been defined as a psychological condition that leads to exhaustion, depersonalization, and decreased teacher achievement and self-worth (Evers, 2011). Recently, burnout syndrome (BOS) has been included in the 11th Revision of the International Classification of Diseases as an occupational phenomenon, resulting from chronic workplace stress that has not been successfully managed. BOS is characterized by three dimensions: (1) feelings of energy depletion or exhaustion; (2) increased mental distance from one's job, or feelings of negativism or cynicism related to one's job; and (3) reduced professional efficacy (Chirico, 2017b; World Health Organization, 2019). However, BOS and work-related stressstrain should be considered as two different constructs and therefore as two different psychosocial risk factors to be both specifically addressed by employers at the workplace (Chirico, 2015, 2017a,d). According to the British Health and Safety Executive, the consequence of work-related stress is "the adverse reaction people have to excessive pressures or other types of demand placed on them at work" (HSE- Health and Safety Executive, 2019, p. 42). This negative consequence is generally called strain, or distress. In other terms, occupational distress is not an illness, yet if excessive and prolonged, it can result in several physical and mental illnesses including BOS, among others. BOS therefore may be also considered as a consequence of long-term occupational stress-strain associated with negative health outcomes at individual and organizational levels, including mental health issues like anxiety and depression, low job satisfaction, low performance and student care, and high absenteeism and turnover rates (Chirico, 2016a).

Spirituality is a very broad concept, which includes the search for meaning in life and can involve a sense of connection to something greater than ourselves. From a Christian perspective, it has been defined as a way a person lives his/her everyday life in view of his/her relationship to the gods/the "spirit world" and can be considered as that which animates a person's life of faith and moves a person's faith to greater depths and perfection (McGrawth, 1999). In the literature, there are also other definitions, that depend on the different opinions of the researchers. Despite the methodological difficulties arising from the different constructs and definitions of spirituality (Martsolf and Mickley, 1998; Buck, 2006; Weathers et al., 2016), there is agreement that the concept of spirituality is not limited to religiosity, and that it includes all forms of meditation (Steinhorn et al., 2017) and mindfulness. Meditation has long been proposed as a prevention strategy for stress-strain and BOS. In the literature, there are several systematic reviews on the efficacy of meditative interventions to reduce physician burnout, yet with inconsistent findings. Indeed, in a systematic review of randomized clinical trials by Dharmawardene et al. (2016), meditative interventions provided a small to moderate benefit for informal caregivers and health professionals for stress reduction, but more research had been advocated by authors to establish effects on burnout. Conversely, a systematic research review by Busireddy reported decreases in depersonalization scores and reduced emotional exhaustion, through self-care workshops and meditation interventions, respectively (Busireddy et al., 2017). In a 4-month study intervention, Elder et al. (2014) showed that the Transcendental Meditation program was effective in reducing psychological distress and burnout even in teachers and support staff working employed at a therapeutic school for students with behavioral problems. Mindfulness can be defined as "the awareness that emerges through paying attention, on purpose, and non-judgmentally to the unfolding of experience moment by moment" (Kabat-Zinn, 2003, p. 145). There is strong evidence that mindfulness practice can reduce job burnout among health care professionals and teachers (Luken and Sammons, 2016).

Religiousness is a particular form of spirituality that proved to be effective in patients with cardiovascular disease or cancer and their caregivers (Movafagh et al., 2017; Abu et al., 2018; Faccio et al., 2018), as well as in depression and mental health problems (Weber and Pargament, 2014; AbdAleati et al., 2016; Braam and Koenig, 2019). In the literature, there are also some studies showing the effectiveness of prayer to tackle occupational stress and burnout. The prayer has been considered as a powerful spiritual coping mechanism in illness (Baldacchino and Draper, 2001), and perception of the transcendent was reported to have a positive influence for stress-related impairment of health in German pastorals (Frick et al., 2016). However, few empirical studies have been conducted on the health effects of Judeo-Christian contemplative prayer practices (Ferguson et al., 2010), especially as a means of coping to address occupational stress and burnout. A research study has documented the use of prayer as an effective coping strategy against work-related stress in a sample of 916 Christian educators, showing a significant relationship between frequency of prayer and job satisfaction (LaBarbera and Hetzel, 2016), and a review showed the essential role of spiritual intelligence in job burnout among teachers (Mirshahi and Barani, 2016). However, although scholars believe that prayer is a valuable tool to be used in nursing and medicine, there is still much controversy on this tool (Sciarra, 2013). Indeed, it is difficult to have a full understanding and unique definition of prayer. Paloma and Pendleton identified four different types of prayer: (i) petitionary prayer, that is, specific requests for oneself or others; (ii) colloquial prayer, that is, a conversational style of prayer in which people may ask for personal guidance, forgiveness, or general blessings; (iii) ritual prayer, that is, memorized prayers or prayers from books; and (iv) meditative prayer, that is, prayer involving reflection upon, and adoration of the divine. According to the authors, each of the well-being measures considered in their studies was influenced by only one type of prayer (Paloma and Pendleton, 1991). According to Jors et al. (2015), prayer could be defined in many ways, based on individual beliefs and religious traditions, but personal prayer involves "the raising up of one's mind to God" (Damascene, 1864), whereas personal meditation is often considered as fpsyg-10-02933 January 9, 2020 Time: 18:25 # 3

a mental exercise or state of being involving reflection or contemplation or mindfulness without being directed toward a higher being (Jors et al., 2015).

The literature on this type of interventions is very scarce, and to our knowledge, there are neither systematic reviews nor randomized controlled trials to support its use among teachers. Therefore, in this study, we wanted to conduct a BOS prevention intervention in the teachers of a religious institute and verify the results of the intervention on the psychological quantities associated with BOS, job satisfaction, and psychological well-being.

# MATERIALS AND METHODS

# Participants and Setting

Workers from a Catholic school of a Congregation of nuns were invited to participate in a workplace health promotion campaign. This population was chosen because the workers had been selected by the employer based on their religious beliefs and adherence to religious practices. They all had experience of Catholic prayer. Under Italian law, all workers were subjected to health surveillance in the workplace, due to the fact that they were occupationally exposed to biological risk or other occupational risks. The promotion program was conducted by the occupational doctor in charge of their surveillance (FC), who had access to their personal medical records. The study population consisted of teachers: (i) working in the institution for over 6 months with a minimum of a 20-h-per-week; (ii) being a lay teacher employed at a primary, preschool, or kindergarten school of the Congregation; (iii) having no abuse of alcohol or drugs consumption with neurological effects; (iv) having no history of psychiatric disorders; and (v) being not unfit for work. Since 62 out of 66 teachers were females, males were excluded from the study.

From among 62 potentially eligible employees, 50 teachers volunteered for this study. Employees were randomized to two groups: prayer treatment (PT) (n = 25) and control group (n = 25). Recruitment, intervention, and follow-up took place between September 2017 and February 2018. The study was conducted according to the Declaration of Helsinki. Written and verbal informed consent was obtained from each participant following verbal description of all experimental details, with this obtained priority to any experimental data collection.

#### Interventions and Randomization

In a Christian prospective, prayer can take different forms, among which are conversational prayer, meditative prayer, ritual prayer, and intercessory prayer. In this study, we used meditative prayer, which consists of contemplation of spiritual themes and the relationship of the divine with the mankind (Jantos and Kiat, 2007).

Intervention protocol consisted of 16 training sessions that occurred over 8 consecutive weeks, with a maximum of two training sessions per week and a minimum of 1-day rest between training sessions. The participants received two ∼30 min training sessions a week for 2 months before the conclusion of the experiment, delivered by the same expert in religion psychology for all participants. The protocol required the combination of an individualized Christian prayer and a focus group of prayerreflection. In this study, we used for the intervention group a combination of ritual Christian prayers with the meditative adoration of the divine.

Before the treatment, participants attended two didactic lectures, followed by an individual interview with the instructor. Participants were advised to practice the prayer once daily for 10 min at home before sleeping. Participants in the control group continued with their usual schedule and were not instructed in PT until after the 2-month intervention study was concluded.

Simple randomization procedures were used to assign participants to groups. The schedule of treatment group allocations was concealed by the study statistician, with individual treatment group assignments revealed to the project manager only when study participants completed baseline testing and were ready to commence treatment. All measures were drawn from self-administered questionnaires, to avoid any interviewer bias.

# Outcome Measures

Participants were administered a battery of tests at baseline, before instruction in the PT program. The baseline testing took place at the end of September 2017. Participants were then administered the same battery of tests approximately 4 months later, at the beginning of February 2018. The primary outcomes measures were job satisfaction, psychological wellbeing, and burnout-related variables (emotional exhaustion and depersonalization). According to the authors who first described the syndrome, burnout is not defined by the sum of the three components, and the third, reduced efficacy, correlates scarcely with the other two (Leiter and Maslach, 2016); consequently, many researchers have chosen to measure the first two components, emotional exhaustion and depersonalization, or only one of the two (West et al., 2009). Consequently, we measured EE and DP.

For the assessment of BOS, the Italian version of the Maslach Burnout Inventory (MBI) validated for teaching and education sector (Sirigatti and Stafanile, 1993) was used. Participants were asked to rate from 0 (never) to 6 (daily) how often they experienced feelings described in each of the 14 items. The questionnaire consisted of the two "core" dimensions of BOS: emotional exhaustion (nine items), e.g., "I feel frustrated by my job," and depersonalization (five items), e.g., "I don't really care what happens to some students." Emotional exhaustion (EE) values ranged from 0 to 54, whereas depersonalization (DP) values ranged from 0 to 30. Cronbach's α coefficient was 0.85 for exhaustion and 0.77 for depersonalization.

Psychological status was evaluated using the General Health Questionnaire (Goldberg and Williams, 1988) in its version consisting of 30 items (Lattanzi et al., 1988). The General Health Questionnaire scoring traditional method entails that each item is scored 0 for the answers "much worse than usual" or "worse than usual," and 1 for the answers "better than usual" or "same as usual." The sum of each item score gives an overall total score (range, 0–30). The score obtained is indicative of the level of psychological malaise. Cronbach's α in this study was 0.81.

Job satisfaction level was obtained by using the Warr, Cook, and Wall's Job Satisfaction Scale (Warr et al., 1979), in its Italian version (Cronbach's alpha = 0.94) (Magnavita et al., 2007). The questionnaire is a 7-point Likert scale composed of 15 items (score ranges, 15–105), which gives a total score, indicating the job satisfaction level. Cronbach's α in this study was 0.84.

Compliance with the home practice of PT was measured by each participant's self-report at post-testing.

#### Statistical Analysis

fpsyg-10-02933 January 9, 2020 Time: 18:25 # 4

Descriptive analyses were conducted for demographic variables (age groupings, marital status, religious belief, teaching class, length of service), and differences between PT and control group were evaluated using the chi-square test. Unpaired and paired t-tests were used for comparisons. The effect size of the treatment, that is, whether the results are clinically relevant or not, was calculated using Cohen's D. The data provided by participants were statistically analyzed using the data-analysis software R, package Rcmdr version 2.5-1 (Fox, 2017).

### RESULTS

**Table 1** shows the baseline data of participants. Mean ages were 35.56 (SD 6.8) years in the intervention and 37.50 (8.23) years in the control group, respectively. At baseline, differences between groups in age, educational background, religious belief, teaching class, marital status, and years of experience were not statistically significant.

The analysis of the results at the end of the follow-up is shown in **Table 2**. In the prayer group, a statistically significant improvement of all outcome measures was observed. Emotional exhaustion level decreased from 16.80 to 4.92 (p < 0.001); depersonalization, from 3.72 to 0.60 (p < 0.001); and psychological impairment, from 10.08 to 2.04 (p < 0.001), whereas job satisfaction level increased from 45.96 to 77.00 (p < 0.001). The effect sizes (Glass' 1) of the therapeutic intervention ranged from 0.53 (for job satisfaction) to 2.87 (for psychological health), suggesting moderate to huge effects. The compliance with the practice of the prayer technique was high; 100% of the participants prayed at least once a day at home. Participants reported no unexpected, study-related serious adverse events.

# DISCUSSION

In this experience, we observed a significant increase in job satisfaction and mental well-being, as well as a reduction in the levels of depersonalization and emotional exhaustion, in the prayer group, if compared with the control. It is important to note that the two groups shared the same religious belief and were both used to Christian prayer. The results of this research indicate that the prayer training experiment, which was developed in the framework of a voluntary workplace

#### TABLE 1 | Baseline data.




<sup>∗</sup>Unpaired t-test between groups and paired t-test within treatment group. ∗∗Cohen's D.

health program, was effective in BOS prevention in a sample of female lay teachers employed in a Catholic school. At the end of the intervention period, workers of the prayer group reported significant improvements compared to controls for all examined outcome variables, i.e., job satisfaction, mental wellbeing, and emotional exhaustion and depersonalizations that are considered in the literature as the "core" dimensions of BOS. The intervention could be considered of moderate to large effect size, with the largest effect measured on psychological stress and subscales of BOS.

The results of this study are not unexpected, based on the literature. Prayer, indeed, has been considered a Western form of transcendental meditation, leading to a psychological and physical status of well-being (Jantos and Kiat, 2007). Numerous studies have shown that Christian and non-Christian prayer can have beneficial effects. An Indian research study revealed that workplace spirituality moderates the negative relationship of stress and health, and positively correlates with health (Kumar and Kumar, 2014). Benevene and Fiorilli (2015) comparing burnout of Catholic consecrated teachers with lay teachers from Italian public and Catholic schools pointed out that teachers from public schools obtained higher means on the emotional exhaustion and depersonalization dimensions (Benevene and Fiorilli, 2015). Zhang et al. (2019) showed that enhancing the spiritual climate among nurses can enhance job satisfaction and reduce nursing burnout and turnover intention. Chandler (2010) has underlined the role that prayer time plays in preventing burnout in ministry. Our study is also in agreement with findings of past research showing how religiousness is associated with low levels of burnout in helping professions like medical students (Wachholtz and Rogoff, 2013), end-of-life care setting operators (Holland and Neimeyer, 2005), hospital employees fpsyg-10-02933 January 9, 2020 Time: 18:25 # 5

(Chirico et al., 2018; Carneiro et al., 2019), nurses (Kovacs and Kèdzy, 2008; Chirico et al., 2018), nuns serving as nurses (Kovacs, 2009), and lay and consecrated school teachers (Benevene and Callea, 2011; Chirico, 2017c). According to Frederick, Christian human services workers who practice three spiritual exercises, i.e., the Jesus Prayer, the daily examination, and the prayer of consideration, by reconnecting with the empowering, living spirit of God, may effectively prevent and cope with burnout (Frederick et al., 2017). In past experimental studies, the impact of group sessions and individual practice of Centering Prayer two times daily was hypothesized to decrease stress and increase "collaborative relationship with God" in Catholic congregants (Pargament et al., 1988; Ferguson et al., 2010).

Not only prayer but also other forms of Meditation produce desirable physiological changes, by means of the so-called "relaxation response" (Benson, 1975). In the literature, there are many studies about the effectiveness of mindfulness techniques and training programs to alleviate and reduce job burnout in nurses and health professionals (Smith, 2014; White, 2014). They have been found to increase personal presence and empathy and combat stress, burnout, and anxiety among human service professionals as well (McCollum, 2015). Other research showed that higher levels of spirituality reduce levels of burnout among individuals (Captari, 2010) and dimensions of spirituality can buffer the negative effects of burnout (Amen, 2006; Bade and Cook, 2008; Bänziger et al., 2008). For instance, Frederick describes as mindfulness, differentiation of self, and Christian spiritual practices may play a role in preventing and coping with burnout (Frederick et al., 2017). Based on all these studies, spirituality has been considered as a useful coping strategy to address BOS (Doolittle et al., 2013), to be included in the WHO's health definition with a holistic view of this concept (Chirico, 2016b; Chirico and Magnavita, 2019).

Our study confirms also that the results of a quasiexperimental (pre/posttest) pilot study showing a significant improvement of mental health outcomes occurred in a convenience sample of 27 teachers and professional staff working in special education with the use of mindfulness and prayer (Sharp Donahoo et al., 2018). Both meditation and prayer are mind-body-spirit interventions and just differ in the conditioning of the mind. If one is conditioned from birth to believe in the external form of support, prayer seems to be effective, while if someone is conditioned from birth to believe in the internal form of support, meditation seems to be beneficial (Sharma, 2018). Past studies showed that the practice of meditation can lower stressstrain levels by a decreased sympathetic nervous system and hypothalamic-pituitary-adrenal axis and reductions in elevated cortisol (stress hormone) levels (Walton et al., 2004). Therefore, based on our findings, the "meditative" prayer would have the potential to do likewise. In the literature, prayers, meditation, or similar rituals like yoga have all been considered as emotionfocused coping strategies capable of reducing and managing the negative and distress emotions experienced by healthcare professionals at the workplace (Giacalone and Jurkiewicz, 2003). However, despite a growing body of research connecting mindfulness to Christian spiritual practices (Frederick et al., 2017) and emerging evidence of the importance of these spiritual practices when used as coping against stress, worry, and burnout, they are underestimated.

Strengths of this work were a randomized control design, with the inclusion of all study participants employed at schools of the same Congregation. The health surveillance carried out at the workplace has prompted the full participation of teachers who were all post-tested. Furthermore, this was one of the few studies focusing on the beneficial effects of prayer on mental health and burnout levels of teachers. Study limitations included reduced generalizability of a study conducted in a single private school to other types of school or other types of religion. Other studies, conducted on subjects of other religions and which include people not proficient in prayer, may be useful to deepen the results. Furthermore, participants could not be blinded to their treatment assignment. Finally, the use of self-report instruments to measure outcome variables introduced the possibility of bias. However, this was an original pilot study that by identifying a new key area requires further research in this field to analyze the beneficial effects of prayer in other fields and professions.

# Practical Implications

This research has important implications because it is one of the few studies showing that prayer could be effective, especially in religious contexts, no less than meditation and other spiritual or mind-body techniques, to contrast the negative effects of stress-strain and burnout at workplace among teachers (Luken and Sammons, 2016). Practical implications of the research refer to the usage of prayer alternatively or in common with meditation in the framework of workplace health programs, to improve mental well-being, especially in occupational settings like helping professions and human service professionals in which risks arising from work-related stress and occupational burnout hazards can be especially important. In the European Union, interventions to address all the psychosocial hazards at the workplace are mandatory for all employers (Chirico et al., 2019a). However, evidence-based prevention and control strategies in the public health field should be evaluated in terms of cost-effectiveness. The role of religion and spirituality in the promotion of health has been discussed in numerous medical publications (Kirschner, 2003). Workplace health promotion may give several benefits to employees and organizations, due to a safe and healthy work environment (WHO).

# CONCLUSION

Our study therefore confirms that spirituality could be used in the framework of workplace health promotion programs to improve employees' performances and organizational effectiveness, especially in professions, e.g., the teaching, where emotional demands are very high (Karakas, 2010). Prayer and other spiritual techniques could increase the individual resources of workers. As shown by Demerouti et al. (2001) in the Job-demand resources model, more spirituality-related individual resources may buffer the negative effect of job demands on job strain, including burnout, and influence motivation, by promoting the work engagement, when job demands are high.

#### DATA AVAILABILITY STATEMENT

The datasets generated for this study are available on request to the corresponding author.

#### ETHICS STATEMENT

fpsyg-10-02933 January 9, 2020 Time: 18:25 # 6

Ethical review and approval was not required for the study on human participants in accordance with the

#### REFERENCES


local legislation and institutional requirements. The patients/participants provided their written informed consent to participate in this study.

#### AUTHOR CONTRIBUTIONS

FC drafted the manuscript. NM and MS revised the manuscript. FC and SZ performed statistical analyses.


fpsyg-10-02933 January 9, 2020 Time: 18:25 # 7

Frederick, T. V., Dunbar, S., and Thai, Y. (2017). Burnout in Christian perspective. Pastoral Psychol. 67, 267–276. doi: 10.1007/s11089-017-0799-4


**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2020 Chirico, Sharma, Zaffina and Magnavita. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Is Teaching Less Challenging for Career Switchers? First and Second Career Teachers' Appraisal of Professional Challenges and Their Intention to Leave Teaching

#### Larissa Maria Troesch\* † and Catherine Eve Bauer†

Teachers Biography and Professionalization, Institute for Research, Development and Evaluation, Berne University of Teacher Education, Bern, Switzerland

#### Edited by:

Caterina Fiorilli, LUMSA University, Italy

#### Reviewed by:

Katerina Georganta, University of Macedonia, Greece Laura Dal Corso, University of Padua, Italy

\*Correspondence: Larissa Maria Troesch larissa.troesch@phbern.ch

†These authors share first authorship

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 06 June 2019 Accepted: 26 December 2019 Published: 22 January 2020

#### Citation:

Troesch LM and Bauer CE (2020) Is Teaching Less Challenging for Career Switchers? First and Second Career Teachers' Appraisal of Professional Challenges and Their Intention to Leave Teaching. Front. Psychol. 10:3067. doi: 10.3389/fpsyg.2019.03067 Teacher attrition is a major problem in many countries. One possible and widely spread counter measure is to recruit persons from other occupational fields to become teachers. Although the existent literature suggests that second career teachers (SCT) have additional resources compared to first career teachers (FCT), empirical data are still scarce on whether SCT are able to transfer prior skills and knowledge into teaching, how this affects the way they deal with professional demands, as well as their intentions to stay in their new profession. On this basis, the present study explores whether FCT and SCT differ in how challenged they feel by typical professional demands associated with teaching, and in what way their challenge appraisals are relevant for the intention to leave the profession. A questionnaire survey was conducted with a sample of 297 teachers, comprising 193 FCT and 104 SCT. Both groups had had regular teacher training, resulting in a full teacher diploma. Professional demands associated with student learning and assessment were rated as most challenging, whereas classroom management, establishing a professional role and cooperation with colleagues were perceived as less challenging. A group difference was found in professional demands concerning student learning and assessment, where SCT felt less challenged than FCT. Also, SCT were more intent to stay in the teaching profession. Further analyses showed that both group differences are mainly attributable to the higher proportion of male teachers among SCT, as well as to their higher general self-efficacy beliefs. Both career background and the degree to which the teachers felt challenged in their job played a subordinate role for the intention to leave the teaching profession. The findings indicate that SCTs' background as career switchers might be less important for coping with specific professional demands than the existent research literature implies. On the other hand, they indicate that SCT feel nearly as challenged when starting to teach as traditional teachers, and may have the same needs for good teacher induction.

Keywords: second career teachers, professional challenges, job demands, teacher attrition, intention to leave the profession

# INTRODUCTION

fpsyg-10-03067 January 27, 2020 Time: 15:22 # 2

Teaching is considered a highly demanding occupation, as it involves a broad range of job demands that call for multifaceted skills and a high degree of flexibility (e.g., Frey, 2014; Keller-Schneider et al., 2018). While student learning and achievement is often seen as a teacher's core business, being a teacher involves many other sometimes underestimated tasks such as school development, cooperation within and outside the school, or administrative tasks (Danielson, 2013; Frey, 2014). These professional tasks and related professional demands are potential stressors as defined by the transactional stress theory (Lazarus and Folkman, 1984). Teacher stress is widely recognized as a primary factor in causing low job wellbeing (Skaalvik and Skaalvik, 2015) as well as turnover and attrition (McCarthy et al., 2016). General protective and risk factors for teacher attrition have been well documented (cf. Kyriacou, 2001; Montgomery and Rupp, 2005; Borman and Dowling, 2008; Schaefer, 2013), and in the last decade, research on the association between the perception of selected job demands and the intention to quit teaching has been intensified as well (Skaalvik and Skaalvik, 2011, 2018). These questions are of particular interest also for second career teachers (SCT). In contrast to first career teachers (FCT), who pursue teaching as a first career after completing their high school diploma, SCT have completed at least one professional training prior to becoming teachers. In many countries, considerable resources are being invested in the recruitment and training of SCT, often as a measure to compensate for teacher shortages; yet, there is only scarce information concerning SCTs' ability to cope with the demands of their new profession, and their long-term engagement in teaching. The existent literature implies that SCT bring skills and resources into the teaching profession that are highly relevant for coping with teachers' professional demands (e.g., Tigchelaar et al., 2010). But do SCT actually appraise professional demands differently from FCT, and in what way are their appraisals connected to their intentions to stay in the profession or leave teaching again? The present article explores these questions.

# THEORETICAL FRAMEWORK

# Professional Demands as Challenges

Job demands are defined as the physical, social, or organizational aspects of an individual's work activities that require effort and are thus associated with physiological and/or psychological costs (Demerouti and Bakker, 2011). This definition includes both general demands that are not bound to a particular profession, i.e., demands that can occur in every job, such as work overload or interpersonal conflict, as well as profession-specific demands (hereafter referred to as "professional demands"), i.e., the requirement to cope with the profession's typical work tasks, calling for specific resources in the form of professional knowledge and skills. Teaching-related professional demands require teachers to activate a broad range of adequate resources; this can lead to feelings of stress, especially in early career teachers (Voss and Kunter, 2019). On the other hand, professional demands can be important sources of professional development, as the gap between situational requirements and an individual's resources, if not too wide, fosters the acquisition of new skills and resources (Keller-Schneider et al., 2018). Other than the exposure to general job demands such as work conflict, the mastery of professional demands that are specific to the teaching profession – demands related to classroom management or parent conferences, for instance – will inevitably contribute to the expansion or consolidation of a teacher's professional competencies. In this sense, teachingrelated professional demands always imply a potential gain in the form of professional growth and personal achievement, making them challenges in the sense of the transactional stress model by Lazarus and Folkman (1984) or, as LePine et al. (2005) suggest, challenge stressors as opposed to hindrance stressors.

The transactional stress model (Lazarus and Folkman, 1984) puts appraisals at the heart of the stress process and considers the individual's personal resources. The model explains feelings of stress as the product of a complex and highly subjective transaction between situational demands and individual resources that involves two major appraisal processes: Primary appraisal, evaluating a given situation as either irrelevant, positive or stressful, and secondary appraisal, evaluating coping options and resources. The primary appraisal process leads to stressful appraisals if the situation's demands tax or exceed the individual's resources. A stressful appraisal can occur in the form of harm or loss, if some sort of physical or emotional damage has already occurred; in the form of threat, if the situation exceeds the person's resources, and bears the potential of future damage; or in the form of challenge, if the situation, although taxing the person's resources, bears the potential of success and an expansion of resources. The boundaries between the three forms of stressors are fluid: e.g., a challenge, albeit potentially positive, can turn into a negative stressor if the second appraisal exposes the available resources as insufficient.

Of the three suggested stressors, challenge has probably been the least studied so far. As Smith and Kirby (2011) point out, the transactional model's definition of stress has often been interpreted rather restrictively, limiting its occurrence to situations in which the demands clearly exceed the resources, precluding challenge-related forms of stress. This interpretation disregards the crucial role of appraisal in the stress process, and the fact that an individual's assessment of a situation as either threatening or challenging is context-dependent. Indeed, current stress research – particularly the widely used job demands-resources framework – regards job demands primarily as negative stressors, and considers appraisal processes rather rarely (e.g., Bakker and Demerouti, 2017). Yet, some authors have chosen a different approach. As mentioned above, LePine and colleagues suggest a differentiation between hindrance and challenge job demands (LePine et al., 2005), defining hindrance job demands as negative stressors that interfere with an individual's goals, and challenge job demands as positive stressors that can promote growth and achievement. Empirical findings corroborate this approach: As predicted by the transactional model, challenge appraisals mediate the relationship between

stressors and outcomes such as job satisfaction or turnover intentions (Webster et al., 2011; Searle and Auton, 2014), but are fluid in the sense that stressors can be simultaneously appraised as both a challenge and hindrance (Webster et al., 2011). These studies provide evidence that challenge appraisals might be an important link in the relationship between job demands and work outcomes such as feelings of stress or turnover intention. In order to look at teachers' challenge appraisals and their role for turnover intentions, it is necessary to first establish what professional demands are specific to the teaching profession.

## Teachers' Professional Demands

Many international studies have identified a wide range of teacher job demands over the past decades (e.g., Kyriacou, 2001; Gu and Day, 2007; Keller-Schneider, 2014). Among the professionspecific demands that are perceived as most challenging are discipline problems, low student motivation, ill-defined roles between members of the teaching staff, difficulties in teamwork and great diversity of the student body (Hakanen et al., 2006; Keller-Schneider, 2010; Skaalvik and Skaalvik, 2018). In the first few years of their careers in particular, mastering job-related tasks places high demands on teachers (Keller-Schneider, 2014). In order to be able to examine the ways how teachers appraise specific professional demands and the way they match their professional skills, Keller-Schneider developed an empirically grounded model of the perception of professional demands (Keller-Schneider, 2008; Keller-Schneider et al., 2018). Based on interviews with school teachers and supervisors, and pursuing a factor analytic approach, Keller-Schneider found that the professional demands teachers experience in their daily work can be allocated to four major domains (Keller-Schneider et al., 2018):


Following the transactional stress model outlined above, Keller-Schneider's approach suggests that it is not a professional demand per se, but the affected person's subjective appraisal that is crucial for the individual stress response. An empirical evaluation of the model showed that of these professional demands, cooperation was appraised as the least challenging by beginning teachers, whereas teaching, professional role and identity and adaptive classroom management were rated as more challenging (Keller-Schneider et al., 2018). To what extent teachers felt challenged varied with their career phase: Beginning teachers felt more challenged by the demand to establish their professional role and identity as a teacher, but less challenged by cooperation tasks than more experienced teachers, suggesting that they perceived cooperation with their colleagues more as a resource than a challenge.

# The Role of Professional Demands for Teacher Attrition

Teachers often do not remain in the teaching profession for their entire working life. International findings on teacher attrition and retention are difficult to compare because they often measure different things. However, there is agreement that many countries are struggling to train and retain a sufficient number of qualified teachers, and that the rate of leaving is particularly high among young teachers (e.g., OECD, 2005; Clandinin et al., 2015).

The significance of work-related demands for teacher attrition becomes clear in surveys on the reasons for leaving the profession: Among the most frequently mentioned reasons for attrition are high demands and job stress in general, work overload, the desire for better career options, family responsibilities and lack of administrative support (Kersaint et al., 2007; Druschke and Seibt, 2016). Skaalvik and Skaalvik have shown that workload and time pressure are among the job demands that are most strongly related to the intention to leave teaching, mediated by low job well-being (Skaalvik and Skaalvik, 2011, 2018). These findings highlight the impact of general job demands for teacher attrition. Yet, they explain little whether the many professional demands that are specific to the teaching profession contribute differently to the intention to leave teaching again. Is stress related to classroom management more influential than stress related to cooperation, for instance?

This is a complex question since, as explained above, workrelated demands can be hindrance stressors and challenge stressors at the same time (Webster et al., 2011). The teaching profession indeed has the reputation of being both highly demanding and highly rewarding (Johnson and Birkeland, 2003; Skaalvik and Skaalvik, 2015). Teachers' challenge appraisals might be a crucial factor to explain this supposed discrepancy, and how it is related to teacher attrition, as professional challenges imply potential rewards on the one hand, but are on the other hand a form of stress in the sense of the transactional model (LePine et al., 2005). Teachers' appraisals of workrelated demands are influenced by personal resources such as professional skills, self-efficacy beliefs, goals and expectations (Kyriacou and Sutcliffe, 1978; Rudow, 1999; Troesch and Bauer, 2017a), and in turn have an influence on job-wellbeing and eventually the intention to stay in the profession (Skaalvik and Skaalvik, 2011). The relevance of personal resources

for the appraisal process indicates that teachers' individual biographies might have a substantial impact on how they perceive professional challenges, and whether they intend to stay in teaching. A newer discourse, framing teacher attrition as an identity-making process, points into a similar direction (Schaefer et al., 2012). Emphasizing the significance of personal goals and experiences, these authors argue that leaving the profession can be a deliberate life-making process if teachers are strongly challenged by their professional role and identity, and experience a frustrating discrepancy between their original images of what kind of a teacher they wanted to be, and their actual experience as teachers.

Against this background, we suggest that a focus on teachers' biographies is important to gain a deeper understanding of how challenging specific professional demands are for different groups of teachers, and how challenge appraisals are connected to the intention to stay in the profession, or to leave teaching again.

# Second Career Teachers (SCT): Characteristics and Resources

In the last decades, the development of alternative routes into teaching has expanded considerably in many countries, often combining academic curricula with work-place learning in order to efficiently qualify professionals for the teaching profession (e.g., Tigchelaar et al., 2010; Marinell and Johnson, 2014). As a measure to compensate for teacher shortages, these programs offer a way to expand the workforce by tapping a pool of professionals from other fields. As conditions, standards and training modalities for career changers vary widely between countries and even universities, SCT are a very heterogeneous group of teachers in terms of age, prior educational and professional background, and duration as well as content of teacher training. Yet, there are few common characteristics that will be addressed in the following paragraphs.

#### Life Experience and Professional Expertise

Having completed at least one professional qualification prior to switching to teaching, SCT are older than FCT, with the consequence that they find themselves in a different phase of their lives, and are more likely to have children of their own (e.g., Tigchelaar et al., 2010; Troesch and Bauer, 2017a). From a career development perspective, SCT have at least once experienced the novice phase of career entry already, and reached a certain level of expertise in another domain, making them "expert novices" when entering the teaching profession (Williams, 2013). Yet, despite the widespread practice of recruiting professionals to become teachers, it is not clear whether SCT are actually able to benefit from their prior training and work experience when having to deal with teacher professional demands. Findings of expertise research generally emphasize that expertise is domain-specific and not easily transferable (Gruber and Mandl, 1996). The accumulation of work and life experiences does not automatically lead to better teaching skills, but can only translate into a professional repertoire if reflected upon and purposefully implemented in the classroom (Freidus and Krasnow, 1991; Mayotte, 2003). Established knowledge and routines can even interfere with professional development in the new profession (Bauer et al., 2017).

#### Biographical Agency and Self-Efficacy

It is a basic assumption of life-course research that important personal resources, such as work-related knowledge and control beliefs, are not only shaped by curricular approaches, but also by the individual occupational biography (Moen, 2003; Heinz et al., 2004). Life-course transitions such as career entry and development, including career changes, can be understood as processes of self-socialization, built upon the dynamics of individual agency in varying social contexts over time, when individuals tackle challenges, pursue their aspirations and try to cope with disappointment by selecting pathways and solutions that they perceive as promising. In the context of these processes, experiences related to work options and conditions are translated into biographical action orientations, which are part of a strategic adaptation in order to regain a sense of control, identity and contingence, and which shape further decision-making.

Sense of control is an important personal resource that is associated with a higher chance to move into other occupations or reentering the educational system (Moen, 2003). Starting from this premise, it should be assumed that SCT do not only have a headstart concerning age, but also concerning their biographical agency, and ultimately their control beliefs. Research on SCTs' self-efficacy beliefs corroborates this theoretical assumption. Indeed, SCT have been shown to have self-efficacy beliefs above average compared to teachers in general (Weinmann-Lutz et al., 2006). Self-efficacy beliefs are defined as an individual's conviction about his or her capabilities to accomplish a task when faced with a challenge (Bandura, 1997). Mastery experiences are thought to be one of the most important sources of self-efficacy beliefs. Our own data corroborate that self-efficacy beliefs are higher in SCT than in FCT, but also suggest that they have a higher impact on job stress in SCT (Troesch and Bauer, 2017a), making self-efficacy a crucial control variable in the investigation of professional demands and work outcomes in SCT.

#### Other Characteristics

Other studies suggest that there are even more desirable qualifications and skills associated with career change into teaching, such as pronounced intrinsic motivations to teach (Williams and Forgasz, 2009; Zuzovsky and Donitsa-Schmidt, 2014), high empathy and communication skills (Freidus and Krasnow, 1991), and a great interest in further education and professional development (Weinmann-Lutz et al., 2006; for an overview, see Tigchelaar et al., 2010). On these grounds, it seems safe to assume that SCT bring additional resources into the teaching profession that might facilitate the ability to cope with professional demands, leading to less teacher stress in SCT compared to FCT. Up to now, there is only very limited research concerning this assumption. Keller-Schneider et al. (2016) found no significant differences between SCT and FCT regarding the extent to which they felt challenged by typical professional demands. However, the SCT in Keller-Schneider's study alternatively certified; it seems plausible that potential benefits of prior career experiences had been outweighed by the

shortened teacher training. Our own data revealed no significant differences between SCT and FCT concerning job stress either, but SCT were more satisfied with their jobs than FCT (Troesch and Bauer, 2017a).

Concerning SCTs' long-term retention in the teaching profession, the scarce empirical data is conflicting: While some authors suggest that SCT might be particularly prone to switching jobs or even careers again if working conditions do not meet their standards (Johnson and Birkeland, 2003), other studies show higher intentions to leave the profession in FCT than SCT (Troesch and Bauer, 2017b) or no group differences at all (Boyd et al., 2011a; Kocher et al., 2019). The inconsistent nature of these findings might be due to the heterogeneous samples of SCT involved in the different studies, particularly concerning their prior careers and qualifications, but also their qualifications as teachers. Moreover, there is evidence that the nature of teacher training as well as the previous educational background has an influence on retention (Zuzovsky and Donitsa-Schmidt, 2017). In order to learn more about whether a previous career background per se has any influence on teacher attrition, it would be necessary to compare SCT and FCT with an identical or similar teacher training background.

# Research Questions

The literature reviewed above shows that due to their career pathways, SCT might have knowledge and resources that facilitate the mastery of professional demands, and reduce attrition from teaching (Tigchelaar et al., 2010). However, there is little empirical data concerning SCTs' actual appraisal of professional demands and how they influence their turnover intentions. Drawing on the transactional stress model (Lazarus and Folkman, 1984), professional demands can be conceptualized as challenges, i.e., a type of stressor that can have positive consequences in terms of personal and professional growth, but that can turn into a negative stressor if the challenge is too high and exceeding the individual's resources. As work-related demands and feelings of stress are strong predictors for attrition, it can be assumed that teachers' challenge appraisals of their daily professional tasks are an important influencing factor for their intentions to stay in the profession. However, it is unclear whether SCT and FCT react differently to the specific demands in the teacher profession, and whether different professional demands have a different impact on the intention to leave the profession. In order to address these issues, the present study addresses the following research questions:

	- a. Does the relationship between challenge appraisals and intention to leave vary for different professional demands?
	- b. Do SCT and FCT differ regarding the relationship between challenge appraisals and intention to leave?

# MATERIALS AND METHODS

# Participants

The data are based on the project "Professionals as Teachers," a Swiss research project investigating first and SCT' job wellbeing and retention (Bauer and Hostettler, 2012; Troesch and Bauer, 2017a). All teachers belonging to the 2004–2007 graduation cohorts of the Berne University of Teacher Education (Switzerland) were contacted by questionnaire (912 persons) 7–10 years after graduation, of which 400 took part in the study. The aim of the present study was to examine challenge appraisals in the teaching profession. This appraisal requires current experience in teaching. Therefore, the present analyses do not include persons who were not working as teachers at the time of the survey. This led to a sub-sample of 297 persons who were still working in the teaching profession at that time (average age = 34.21; SD = 6.54; 229 women). Participants who had completed a prior vocational training before entering teacher training, and had thus pursued a prior career, were allocated to the SCT group; this definition applied to 104 persons. Participants who had completed teacher training as a first vocational training were allocated to the FCT group; this definition applied to 193 participants. Although most SCT had entered teacher training via a specific admission for SCT, including a preparatory course and entrance exam, all participants had completed regular teacher training either for kindergarten, primary or secondary education. In the SCT group, the average amount of work experience in their previous career was 7.14 years (SD = 7.87). Approximately one third of the SCT had worked in administrative professions prior to teacher training; in the other two thirds, most had had prior careers in health and craft occupations. Sample descriptives are shown in **Tables 1**, **2**.

# Instruments

#### Challenge Appraisal of Professional Demands

The appraisal of challenge when confronted with specific professional demands were measured using the short scale of Keller-Schneiders' Professional Requirement Scales (EABest-k; Keller-Schneider, 2010, 2014). For this purpose, the respondents were asked to assess 25 typical teacher professional demands regarding the extent to which they perceived them as challenging ("beanspruchend"; 1 = little to 6 = very much). The full scale is available in **Supplementary Material 1**. To analyze the structure of the typical professional demands, Keller-Schneider (2010) conducted a factor analysis revealing four factors: teaching to meet individual students' needs (hereafter referred to as "teaching"; 8 items, e.g., "implement individualized instruction in the classroom"), adaptive classroom management (hereafter referred to as "classroom management"; 6 items, e.g., "Be aware of and lead classroom dynamics"), cooperation within the school (hereafter referred to as "cooperation"; 4 items, e.g., "develop a successful cooperation with the principle") and professional role and identity (hereafter referred to as "professional role"; 7 items, e.g., "taking care of ongoing development"). The mean value of the corresponding items was used to form the subscales. The internal consistencies of the subscales of the present study can

#### TABLE 1 | Descriptive statistics of study variables.

fpsyg-10-03067 January 27, 2020 Time: 15:22 # 6


SCT = second career teachers (n = 104); FCT = first career teachers (n = 193); M = Mean; Class. Management = classroom management. <sup>1</sup>X <sup>2</sup> – statistics for nominal data (percentage female); Independent Samples Mann–Whitney U-Test/z – statistics and r for ordinal data (intention to leave); <sup>2</sup> Instead of means, mean ranks are reported, median was 2 for the total sample, as well as for SCT and FCT; Analyses with imputed data.

#### TABLE 2 | Correlation of study variables.


Gender: 0 = male, 1 = female; Class. Management = classroom management. Analyses with imputed data. N = 297. \*\*p < 0.01; \*\*\*p < 0.001.

be classified as good with α = 0.84 for teaching, α = 0.82 for classroom management, α = 0.73 for cooperation and α = 0.81 for professional role.

#### Intention to Leave the Profession

The intention to leave the teaching profession was measured with a single question: "Can you imagine leaving teaching in the foreseeable future?" Answering options were 1 = no, 2 = rather no, 3 = rather yes, and 4 = yes. This means that a higher value corresponds to a higher intention to leave.

#### General and Teacher Self-Efficacy

General self-efficacy was assessed with the respective instrument by Schwarzer and Jerusalem (1999), and teacher self-efficacy with the respective instrument by Schwarzer and Schmitz (1999). Both scales contain 10 items which had to be rated on a scale from 1 = absolutely do not agree to 4 = "I absolutely agree" (**Supplementary Material 1**). Examples for general self-efficacy are "If someone opposes me, I can find the means and ways to get what I want" and "I can remain calm when facing difficulties because I can rely on my coping abilities." Examples of teacher self-efficacy expectations are "I know that I can maintain a positive relationship with parents, even when tensions arise" and "I know that I can motivate my students to participate in innovative projects." The internal consistencies of the scales can be described as very good with α = 0.84 for general selfefficacy and α = 0.74 for teacher self-efficacy. As recommended by Schwarzer and Jerusalem (1999) and Schwarzer and Schmitz (1999), all the item scores were summed up to form the total score with a possible range of 10 to 40.

#### Analytic Approach

To test for research question 1, multivariate analyses were calculated with the challenges by professional demands as dependent variables and career path (SCT/FCT), age, gender, number of years of teaching experience, degree of employment

as a teacher in percent and self-efficacy (general and teacher selfefficacy) as independent variables. For research questions 2 and 3a, an ordinal regression was calculated to intention to leave the profession. In Model 1, the following independent variables were introduced: age, gender, number of years of teaching experience, degree of employment as a teacher in percent, career path (SCT/FCT) and self-efficacy (general and teacher self-efficacy). In Model 2, the challenge appraisals of the four professional demands were introduced. For research question 3b, ordinal regression analyses were calculated separately for SCT and FCT. In Model 1, age, gender, number of years of teaching experience, degree of employment as a teacher in percent and self-efficacy (general and teacher self-efficacy) were used as independent variables and intention to leave the profession as dependent variable. In Model 2, challenge appraisals of the four professional demands were added.

#### Missing Values

8.08% of the participants had missing on construct-level. One possibility to deal with missing values is multiple imputation. Multiple imputation is regarded as superior to the traditional way of dealing with missing (listwise deletion; pairwise deletion; Graham, 2009). To conduct multiple imputations, we followed the recommendations by Graham et al. (2007) and Newman (2014). Forty datasets were estimated.

# RESULTS

#### Research Question 1: Group Differences Regarding Challenge Appraisals

The descriptive analyses show that SCT reported the highest challenges in relation to teaching tasks in the narrower sense, which were appraised as rather challenging. Cooperation with other professionals was appraised as the least challenging, with classroom management and professional role ranging in between (**Table 1**). As the repeated measures ANOVA<sup>1</sup> showed, SCT rated teaching as significantly more challenging than cooperation with other professionals (Mean Difference = 1.28, SE = 0.10, p < 0.001) and professional role (Mean Difference = 0.36, SE = 0.08, p < 0.001), but not more challenging than classroom management (Mean Difference = 0.20, SE = 0.08, p = 0.10). However, there was no significant difference in challenge appraisals between classroom management and professional role in SCT (Mean Difference = 0.16, SE = 0.07, p = 0.19). Cooperation with other professionals was rated as less challenging than all other job demands (all > 0.001). A similar picture emerged for FCT: teaching was rated as most challenging and significantly more challenging than all other job demands (all > 0.001). Also, cooperation with other professionals was rated as less challenging compared to all other job demands in FCT (all > 0.001). However, no significant difference was found between challenge appraisals in classroom management and professional role (Mean Difference = 0.06, SE = 0.06, p = 1.00). When analyzing the total sample (SCT and FCT), the job demand teaching was rated as most challenging and significantly more challenging than all other job demands (all > 0.001). Moreover, there was no significant difference in challenge appraisals between classroom management and professional role in the total sample (Mean Difference = 0.09, SE = 0.05, p = 0.24). Cooperation with other professionals was rated as less challenging than all other job demands (all > 0.001).

As t-tests revealed, significant group differences between SCT and FCT were only found concerning teaching tasks, where FCT felt significantly more challenged than SCT (**Table 1**). No significant differences between SCT and FCT were found concerning classroom management, cooperation and professional role.

Multivariate analyses revealed that taking important variables into account, the association between career path (SCT/FCT) and challenge appraisals was not significant (F(4,286) = 0.23, p = 0.92, η 2 partial = 0.00). When the challenge appraisals were treated separately for different professional demands, career path had no effect for any of them (all < 0.05). However, general selfefficacy (F(4,286) = 3.09, p < 0.05, η 2 partial = 0.04) and teacher self-efficacy (F(4,284) = 3.37, p < 0.05, η 2 partial = 0.05) were negatively related to challenge appraisals; gender was positively related to challenge appraisals (F(4,286) = 3.33, p < 0.05, η 2 partial = 0.04). As illustrated in **Table 2**, gender was particularly important for challenge appraisal in teaching, with women feeling more challenged by teaching than men (F(1,289) = 11.24, p < 0.01, η 2 partial = 0.04). Age, years of experience and worktime percentage as a teacher, on the other hand, had no significant effect on challenge appraisals (all > 0.05).

#### Research Question 2: Group Differences Regarding Intentions to Leave Teaching

An Independent Samples Mann–Whitney U-Test revealed a significant group difference: FCT showed higher intentions to leave teaching compared to SCT (see **Table 1**). The group effect is small, with both groups expressing a relatively high mean intention to stay in the profession.

Taking important control variables into account, the difference between SCT and FCT in intention to leave the profession disappeared (**Table 3**). In the first ordinal regression analysis (Model 1), gender as well as general and teacher self-efficacy proved to be significant predictors of the intention to leave the teaching profession. Men were more likely to consider leaving the profession than women. Teachers with lower general and teacher self-efficacy reported higher intentions to leave. Age, number of years of teaching experience and degree of employment as a teacher were not significant for the intention to leave the profession. Model 1 explained 13% of the variance.

# Research Question 3: Relationship Between Challenge Appraisals and Intention to Leave

#### Difference Between Professional Demands

As the correlational analyses reveal (**Table 2**), challenge appraisals of the four different professional demands were not significantly related to the intention to leave the profession. Moreover, the

<sup>1</sup>Post hoc analyses; Bonferroni adjusted for multiple comparisons.

TABLE 3 | Ordinal regression analysis predicting intention to leave the teaching profession, total sample (N = 297).


OR = Odds Ratio; Gender: 0 = male, 1 = female; Career path: 0 = first career teachers, 1 = second career teachers; Class. Management = classroom management. Nagelkerke's R<sup>2</sup> = 0.13 for Model 1; Nagelkerke's R<sup>2</sup> = 0.15 for Model 2. Analyses with imputed data.

correlational coefficients were in a similar range (−0.03 until −0.06) indicating that there were no differences in the relations between feeling challenged by specific professional demands and the intention to leave the profession.

This result was confirmed by the ordinal regression analysis: As Model 2 shows (**Table 3**), none of the challenge appraisals of professional demands reached statistical significance taking important controls into account. Again, career path, age, number of years of teaching experience and degree of employment as a teacher were no significant predictors. Moreover, general selfefficacy was not significantly related to intention to leave the profession in Model 2. However, gender and teacher self-efficacy emerged as relevant predictors. In Model 2, 15% of the variance in intention to leave was explained.

#### Group Differences

In a next step, SCT and FCT were analyzed separately (**Tables 4**, **5**). The challenge appraisals for the four professional demands had no significant effect on intention to leave, neither for SCT nor for FCT. In SCT, but not FCT, gender and general selfefficacy played an important role for the intention to leave the profession. In contrast, teacher self-efficacy played a significant role for the intention to leave only in FCT. For both groups, age, number of years of teaching experience and degree of employment as a teacher were not significant for intention to leave the profession, It is also noteworthy that the included variables explained considerably more variance in the SCT than the FCT group (Model 2: SCT: 29%; FCT: 10%).

# DISCUSSION

Due to the increasing flexibilization of educational opportunities and occupational biographies, as well as recurring teacher shortages, SCT are a growing teacher group in many countries. Policy makers who create programs to recruit and train SCT rely heavily on the assumption that having had a previous career is associated with transferable skills and knowledge, resulting in alternative and often shortened training programs for SCT; but in fact, empirical data corroborating this hypothesis are scarce (e.g., Marinell and Johnson, 2014). Findings from mainly qualitative, small-scale studies suggest that SCT are indeed able to draw on their prior work experience and knowledge in certain ways (for an overview, see Tigchelaar et al., 2010). However, there is hardly any data addressing the question whether these experiences can be understood as an expanded set of resources that helps to cope with specific professional demands in the teaching profession. On this background, the present study aimed to investigate to what extent second and FCT feel challenged by typical teacher professional demands and in what way these challenge appraisals influence the intention to stay in the teaching profession.

# Do SCT Feel Less Challenged by Teacher-Specific Professional Demands?

The answer to this question is complex: SCT do feel less challenged, but only to a very limited degree, and it seems that this difference is mainly attributable to SCTs' personal resources as well as to the gender distribution in this subgroup of teachers.

In line with earlier findings (Keller-Schneider et al., 2018), professional demands directly related to student learning and assessment are rated as most challenging compared to the job demands classroom management, professional role and cooperation with other professionals. This finding might be attributable to the fact that teaching and student assessment are often regarded as a teacher's core business, and demand a great amount of a teacher's working hours. FCT felt more challenged by these tasks than SCT, but further analyses showed that this group difference vanishes when age, gender, years of experience and work-time percentage as a teacher as well as general and teacher self-efficacy are being controlled for. Of these control variables, only gender and general as well

TABLE 4 | Ordinal regression analysis predicting intention to leave the teaching profession, second career teachers (n = 104).


OR = Odds Ratio; Gender: 0 = male, 1 = female; Class. Management = classroom management. Nagelkerke's R<sup>2</sup> = 0.27 for Model 1; Nagelkerke's R<sup>2</sup> = 0.29 for Model 2. Analyses with imputed data.

TABLE 5 | Ordinal regression analysis predicting intention to leave the teaching profession, first career teachers (n = 193).


OR = Odds Ratio; Gender: 0 = male, 1 = female; Class. Management = classroom management. Nagelkerke's R<sup>2</sup> = 0.09 for Model 1; Nagelkerke's R<sup>2</sup> = 0.10 for Model 2. Analyses with imputed data.

as teacher self-efficacy were significantly related to challenge appraisals of professional demands. This leads us to two possible interpretations: As the percentage of male teachers is higher in SCT, and males tend to report lower levels of challenge in teaching, the group difference regarding challenge appraisals concerning teaching tasks might also be attributable to gender. On the other hand, SCTs' lower feelings of challenge might be a consequence of their higher general self-efficacy beliefs. This interpretation would be in line with the findings from life-course research, suggesting that life-course transitions such as career changes facilitate the development of a strong sense of control (Heinz et al., 2004).

The finding that establishing a professional role as well as cooperation with other professionals were perceived as equally challenging by both groups, is to some extent surprising. While occupation-specific skills and knowledge from prior fields of work (e.g., in health or administration jobs) may not be transferable to the teaching profession, cross-domain skills should be. Work experiences usually include forms of collaboration, as well as having to establish a flexible professional role toward different interaction partners within the respective work context, i.e., clients, supervisors, colleagues, etc. Moreover, SCT have been shown to integrate more easily into the school organization (Tigchelaar et al., 2008). Therefore, it would have been plausible that establishing a professional role and cooperation with other professionals are appraised as less challenging by SCT. On the other hand, interview data from the same study (Bauer et al., 2017) showed that SCT often struggled with disappointments as they tended to have high expectations regarding team work in their new profession, but also high expectations regarding their own skills and abilities to cope with professional demands. Teaching differs from many

occupations – at least in Switzerland – in that teamwork is only an emerging concept, and that there tend to be less guidelines than in other occupational fields in terms of regular feedbacks or performance standards; substantial differences that might foster disappointments and disorientation in SCT. As SCT are a growing teacher subgroup, it might be beneficial for future research to look more closely into the conditions under which SCTs' prior work experiences can influence the resources that help to establish a professional role as a teacher, as well as cooperation with other professionals in the school environment.

# Do SCT and FCT Differ Regarding Their Intentions to Leave the Profession?

Similar to the results concerning challenge appraisals, our findings show that FCT had higher intentions to leave the teaching profession than SCT, and that this group difference could be attributed mainly to gender and personal resources, as the group difference between FCT and SCT vanished after taking these control variables into account.

Self-efficacy and gender emerged as relevant predictors for the intention to leave the profession. However, the importance of these predictors differed for SCT and FCT: Among SCT, but not FCT, men were more open toward leaving the profession than women. At the same time, teacher self-efficacy was negatively related to the intention to leave in FCT, but not SCT. The importance of self-efficacy beliefs in the workplace for teacher retention has been shown in many studies (for an overview, see Chambers Mack et al., 2019). Our findings corroborate these results, with the addition that for SCT, general self-efficacy is more important regarding their intentions to stay teachers, while for FCT, teacher-related self-efficacy seems to be more relevant. One possible explanation for this finding is that SCTs' cumulative work experiences are reflected in their general selfefficacy rather than their teacher-related self-efficacy, and that their higher general self-efficacy beliefs compensate for possible weaknesses they feel regarding their efficacy as teachers. However, it might also be attributable to a selective drop-out: teachers with limited resources are likely to have left teaching long ago, resulting in them having been excluded from these analyses that included active teachers only. Thus, it is possible that the rate of teachers who have already left teaching is different for SCT or FCT. In order to investigate the connections between career path, personal resources, feelings of challenge and the intention to leave the profession more thoroughly, a longitudinal study would be needed.

On first glance, it might be surprising that gender has an impact on the intention to leave only in SCT, as earlier studies have shown a higher interest in professional development for male teachers in general (Borman and Dowling, 2008). But while one explanation lies in the higher proportion of males in SCT, it is also highly plausible that this effect has to do with different life phases: With a mean age of 40 for SCT and 32 for FCT, FCT matched the life phase between 30 and 35 when female Swiss teachers show a distinct rise in attrition from teaching due to family reasons (SKBF, 2014), leveling the difference between males and females in this subgroup.

The relationship between work experience prior to teaching and intention to leave the profession has rarely been studied, and the findings are inconsistent. On the one hand, a higher age when entering teaching is associated with a higher probability of remaining in the profession (Borman and Dowling, 2008). Because SCT are usually older than FCT, this might indicate that SCT are more prone to remain in the teaching profession. In addition, career changers have particularly strong motivations to teach (Zuzovsky and Donitsa-Schmidt, 2014), which in turn is an important predictor of career retention (Watt et al., 2014). On the other hand, SCT have higher general self-efficacy (Troesch and Bauer, 2017a), which is associated with a higher probability for further career change (Lent et al., 1994). Moreover, SCT have their first career as a fallback career in case teaching should not be as rewarding as anticipated. The present findings corroborate that SCT might be more intent to stay longer in the teaching profession – at least if they have a regular teacher's diploma as was the case in the present sample. It has to be considered that the present sample of SCT has a regular, full teacher's degree as opposed to an alternative certificate. Attrition intentions are linked to a subjective cost-benefit-analysis, and the time and financial investments associated with the completion of often shortened alternative teacher training programs for SCT are comparably low (Weinmann-Lutz, 2007). Indeed, teachers with alternative certificates are more likely to leave the profession again (Chambers Mack et al., 2019). However, the effect of alternative certifications on teacher attrition intention might also depend on the specific curriculum: According to Kocher et al. (2019), retention rates might be associated with the extent to which training on the job is part of the alternative certification program. The degree to which teacher are socialized into their profession seems to play a role for retention and attrition.

# What Is the Role of Teachers' Challenge Appraisals for Their Intention to Leave the Teaching Profession?

The extent to which the teachers felt challenged by typical professional demands of the teaching profession played a subordinate role for their intention to leave the profession. None of the professional demands was a significant predictor in this regard. This finding is somewhat surprising, as professional challenges – while being associated with beneficial outcomes – are still a type of stressor according to the transactional stress model (Lazarus and Folkman, 1984), and can turn into negative stressors if the challenge is too high. Yet, our findings suggest that professional challenges, even if high, are not associated with the intention to leave the profession. There are several interpretations of this phenomenon:

The extent to which an individual feels challenged by a job demand is an indicator for the extent to which relevant resources have to be considered and made available or, if already at disposal, activated. This process can put high pressure on an individual that might be perceived as momentarily stressful. However, these moments of stress are not the same as the long-term, health-impairing process of work stress that leads to exhaustion and attrition due to burnout, as described by many

studies, mainly within the job demands-resources framework (Demerouti et al., 2001). As long as teachers are appreciated by their students and feel that they can accomplish something through their work, i.e., as long as they have a sense of selfefficacy, stress and dissatisfaction remain low (Dicke et al., 2018). Teacher self-efficacy was a relevant predictor for intention to leave the profession in the present analyses, corroborating that self-efficacy beliefs buffer the job demands-strain relationship. From this perspective, our results yield an explanation for the often-cited finding that teachers feel both highly strained and highly satisfied at the same time: The professional demands related to teaching are perceived as moderately to highly challenging – the latter in the case of tasks concerning student learning and assessment –, but not to an extent where the career choice is questioned.

Another possible explanation for the weak association between challenge appraisals and intention to leave the profession might be that the perception of professional demands is contextdependent. Workplace factors such as leadership support, organizational characteristics or school climate have been documented as key factors in teacher attrition throughout the literature (Borman and Dowling, 2008; Kukla-Acevedo, 2009; Van Droogenbroeck and Spruyt, 2016). High levels of challenge might lead to a change of workplace – and not the teaching profession – before excessive demands manifest themselves in forms of chronic stress and exhaustion. This interpretation is supported by the study by Boyd et al. (2011b) which showed that professional demands are related to job turnover but not to attrition from the profession.

Only about 15% of the variance in intention to leave the profession could be explained by the personal and individual characteristics included in the model (including age, gender, and career path), as well as the perception of professional demands, indicating that essential predictors were not considered in this study. Contextual factors such as student composition, school location or general workplace atmosphere may explain further variance (see Borman and Dowling, 2008 for a review). In a recent study based on the job demands resource model (Demerouti et al., 2001), Skaalvik and Skaalvik were able to show that perceived professional demands are only indirectly related to motivation to quit, via low well-being and engagement (Skaalvik and Skaalvik, 2018). In addition to the professional demands, teacher well-being and engagement would probably also explain further variance in the intention to leave the profession.

Another factor that has to be considered is that stress and dissatisfaction are not the only reasons for teacher attrition. While so-called push factors – reasons that push the individual away from his or her field of work – are highly relevant for leaving the teacher profession (Ingersoll and Smith, 2003; Herzog et al., 2007; Borman and Dowling, 2008; Struyven and Vanthournout, 2014), pull-factors – reasons that attract the individual to another field – are important as well. Feelings of challenge when confronted with professional demands may be a push factor if the necessary resources cannot be activated to master the challenges. A teacher's degree yields good opportunities for further professional development, e.g., in educational science or special needs education. In order to explain more variance in the intention to leave the profession, it might be beneficial to look into the question whether teachers acknowledge these professional opportunities and how they are related to teacher attrition.

# Limitations and Future Research

The present study addresses several research gaps, but has limitations that can be seen as guidelines for future research. Firstly, the present study is based on cross-sectional data that do not allow any conclusions about the direction of effects, or about the processes involved. Longitudinal studies could provide insight in the processes that lead to the intention to leave a certain workplace, to leave teaching altogether, or to stay in the teaching profession despite the presence of challenges. Including potential mediating variables such as job well-being or job engagement could help to further clarify the relationship between professional demands, stress and attrition. Secondly, to assess the intention to leave the profession, only one item was used in this study. In order to assess attrition intention more comprehensively, a multiitem scale would be favorable. Thirdly, as the variables included in the model only explained a limited amount of the variance in intention to leave the profession (15% for the total sample and only 10% for FCT), it would be interesting to investigate what other predictors are related to intention to leave the profession. We assume that workplace characteristics and pull-factors might explain further variance (Borman and Dowling, 2008; Ingersoll et al., 2014). Fourthly, in this study we have examined which professional demands are challenging for the individual teachers, but not to which degree they perceive them as relevant and whether they think they can cope. However, there is empirical evidence that job demands are only dealt with thoroughly if they are considered important and manageable, but risk to be ignored or avoided otherwise (Keller-Schneider, 2010, 2014). For future research it would be important to consider the extent to which teachers feel challenged by specific job demands in the context of their subjective relevance and manageability.

# CONCLUSION

In sum, the study shows that SCT are more intent to stay in the teaching profession than FCT, but experience as many challenges regarding most professional demands. One exception are professional demands that are related to student learning and assessment, where SCT feel less challenged. These group differences seem to be attributable mainly to the higher proportion of male teacher among SCT, as well as to their higher general self-efficacy beliefs. Our findings indicate that persons who chose teaching as a second or third career – at least if they are well qualified as teachers – feel as prepared to cope with professional demands as traditional teachers, although not distinctly more so. However, although the previous literature suggests that SCT have additional resources compared to FCT (e.g., Tigchelaar et al., 2010; Bauer et al., 2017), SCTs' cumulative experiences and knowledge do not diminish the challenges that they encounter when entering the teaching profession. This does

not mean that SCTs' previous professional skills are lost when they start a new career as teachers; but it indicates that the skills and knowledge gained in the past are only marginally reflected in the appraisal of professional demands as teachers. The findings also imply that SCT need teacher induction as much as any other teacher, as the professional challenges are high, especially regarding student learning and assessment.

Teachers' career backgrounds and challenge appraisals of professional demands can only explain a small amount of variance in intentions to leave the profession again. This corroborates the previous empirical findings that challenge stressors do not have the negative consequences that are usually associated with stress, including attrition.

#### DATA AVAILABILITY STATEMENT

The datasets generated for this study are available on request to the corresponding author.

# ETHICS STATEMENT

Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. The patients/participants provided their written informed consent to participate in this study.

# REFERENCES

Bakker, A. B., and Demerouti, E. (2017). Job demands-resources theory: Taking stock and looking forward. J. Occup. Health Psychol. 22, 273–285. doi: 10.1037/ ocp0000056

Bandura, A. (1997). Self-Efficacy. The Exercise of Control. New York: Freeman.


# AUTHOR CONTRIBUTIONS

The manuscript was written in close collaboration between the two authors. LT had the idea for the manuscript and conducted all analyses. CB wrote most of the Introduction and the Discussion section. Both authors have contributed to the manuscript to a similar amount.

# FUNDING

The project "Professionals as Teachers" (Bauer and Hostettler, 2012; project nr. 12 s 00501) as well as the present publication in Frontiers in Psychology have been supported by the Berne University of Teacher Education, Switzerland.

#### ACKNOWLEDGMENTS

The authors would like to thank Dr. Loredana Torchetti for her statistical advice.

# SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyg. 2019.03067/full#supplementary-material



Lazarus, R. S., and Folkman, S. (1984). Stress, Appraisal, and Coping. New York, NY: Springer.


inconsistent relationships among stressors and performance. Acad. Manag. J. 48, 764–775. doi: 10.5465/AMJ.2005.18803921



**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2020 Troesch and Bauer. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

# Teaching Careers: Exploring Links Between Well-Being, Burnout, Self-Efficacy and Praxis Shock

Julie Ballantyne\* and James Retell

The University of Queensland, Brisbane, QLD, Australia

Burnout and attrition are issues facing many professions. In a bid to better understand this phenomenon and ways to address it, this paper explores experiences of praxis shock, well-being, burnout, and self-efficacy during teachers' careers. Regression and mediation analyses of 836 responses to a questionnaire reveal that praxis shock may occur at multiple points in a music teachers' career. Findings reveal that praxis shock predicts patterns of reported burnout, well-being and self-efficacy. This impacts on the development of productive professional identities, career satisfaction and success. Evidence is presented regarding praxis shock and its impact across a teaching career.

Keywords: teacher education, music teacher attrition, professional identities, pre-service and in-service education and training, occupational stress

#### Edited by:

Caterina Fiorilli, LUMSA University, Italy

#### Reviewed by:

Ilaria Buonomo, LUMSA University, Italy Maria Popa-Roch, Université de Strasbourg, France

> \*Correspondence: Julie Ballantyne j.ballantyne@uq.edu.au

#### Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology

Received: 16 May 2019 Accepted: 20 September 2019 Published: 18 February 2020

#### Citation:

Ballantyne J and Retell J (2020) Teaching Careers: Exploring Links Between Well-Being, Burnout, Self-Efficacy and Praxis Shock. Front. Psychol. 10:2255. doi: 10.3389/fpsyg.2019.02255

# INTRODUCTION

Professional identities are the ways teachers make sense of themselves within their professional lives. They encompass and are interrelated with personal identities, social and cultural identities and norms, and professional roles and contexts. Research consistently notes the centrality of teachers' professional identities to all aspects of their job. Professional identities particularly impact on teachers' abilities to navigate the complexities of 'being a teacher' (Beijaard et al., 2004; Sachs, 2005). This is because the notion of 'who I am as a teacher' influences, and is influenced by, self-perceptions of agency, self-efficacy, effectiveness, and job satisfaction (Day and Kington, 2008; Skaalvik and Skaalvik, 2014). Although many inroads have been made into researching professional identities and the experiences of beginning teachers (see for example Flores and Day, 2006; Beauchamp and Thomas, 2009; Hong, 2010; Pillen et al., 2013), there is still a need to better understand how teachers traverse through other career periods. This article explores how the components of perceived professional identity (including praxis shock, burnout, well-being, and self-efficacy) interact across the teaching career.

Research on the career and identities intersect where both fields acknowledge that broader issues in a person's job and life are relevant when exploring changes over time (Day and Gu, 2007). As people progress through their chosen profession, their experiences and the ways that they respond to all aspects of their work, change and develop. Research suggests that career development is seen as a "whole-life" endeavor, necessarily encompassing the myriad facets of a person's life that may impact on their career (Litano and Major, 2016). Individuals who are in the profession of teaching are therefore expected to experience changes that will influence their career, over time.

Gu and Day (2007) emphasize that teachers' abilities to "cope" in their professional lives are impacted by their identities and professional life phases, which are in turn mediated by issues in their personal and situated lives, their professional values, beliefs, and external policy agendas. The

"ability to cope", or remain resilient in the face of considerable job uncertainty, job stress, policy changes and career challenges, appears to be most challenging to teachers at the beginning of their careers, when they tend to burn out and leave the profession (Schaefer, 2013). Teachers' responses to career changes and stress vary, but it does seem that teachers are leaving the profession because they are "burning out" (Parliament of Australia Media Release, 2019). As indicated by its name, "burning out" is often the end point of a negative response to work stress and emotional exhaustion (Skaalvik and Skaalvik, 2007). Burnout (and attrition) is affecting the profession (perhaps across the career span), and this is of grave concern to policy-makers, teacher education providers, employers and government bodies alike.

Research on attrition and burnout in teaching necessarily reflects the highly complex and multi-faceted experience of teachers in the workforce (Lindqvist et al., 2014). Burnout has been associated with mental and physical health (Hultell et al., 2013), as well as work environment and motivational factors (Fernet et al., 2012). It can also be attributed to 'clients' or students (Kristensen et al., 2005). Thus, when investigating teacher professional identities, measuring burnout provides a picture of the degree to which teachers feel fatigued as a direct result of their work, and therefore the likelihood of their remaining a teacher. Interestingly, recent work by Hultell et al. (2013), suggests that burnout might not exclusively be the domain of the early career teacher, finding that during their first 3 years of employment, teachers were, on average, relatively healthy with moderately low, albeit increasing, levels of burnout (p. 84).

The day-to-day life of a school teacher often represents a fundamental divergence from the expectations and ideals that are held prior to entering the profession. The feeling of surprise, shock or disequilibrium resulting from this experience has been termed praxis shock and is most common in the early years of teaching (Kelchtermans and Ballet, 2002; Stokking et al., 2003; Ballantyne, 2007). The mismatch experienced when one's expectations of professional life do not align with the realities experienced working within the profession, is likely to be accompanied by a sense of disillusionment and disappointment. Managing praxis shock compounds the burden faced by early career professionals and those who experience praxis shock are likely to be less effective in their work, and may be unable to provide their schools and communities with the best possible provision of education. The experience of praxis shock can result in burnout (Ballantyne, 2007), and is also associated with attrition and reduced job satisfaction (Kelchtermans and Ballet, 2002; Ballantyne, 2005, 2007; Hong, 2010; Skaalvik and Skaalvik, 2010). Burnout can also be reduced or mitigated against "when teachers have a positive perception of their self-efficacy" (Betoret, 2006, p. 534). Arguably, the ways that teachers perceive praxis shock actually may be the key to successful identity development, and longevity in the career (Ballantyne and Zhukov, 2017).

Burnout, praxis shock and attrition are clearly interrelated areas of study examined by researchers in the fields of teacher education and teacher education (Kelchtermans and Ballet, 2002; Ballantyne, 2005, 2007; Hong, 2010; Skaalvik and Skaalvik, 2010). Burnout, praxis shock and attrition are also related to issues of job satisfaction and self-efficacy (Skaalvik and Skaalvik, 2010). Perceived self-efficacy is aligned with success in the profession, with Bandura (2006) maintaining that "the stronger the perceived self-efficacy, the higher the performance attainments" (p. 175). For the purposes of this study, the definition of self-efficacy was borrowed from Skaalvik and Skaalvik (2014), in that it refers to teachers' beliefs about their own abilities to plan, organize, and carry out activities required to attain education goals" (p. 69). Skaalvik and Skaalvik found that better self-efficacy increased job satisfaction, and conversely, poor self-efficacy increased the likelihood of burnout in the profession. Previous research in this area links self-efficacy and psychological well-being, personal accomplishment, job satisfaction, and commitment (Zee and Koomen, 2016). Self-efficacy is strongly associated with images of self, and Beijaard et al. argue that such images "strongly determine the way teachers teach, the way they develop as teachers, and their attitudes toward educational changes" (2004, p. 108). It is important, therefore, when examining teacher professional identities over the career, to also investigate teacher self-efficacy.

Most studies in this area focus on the experiences of beginning teachers (Kelchtermans and Ballet, 2002, Kelchtermans and Ballet, 2002; Ballantyne, 2007; Hong, 2010; Shaw, 2016), and very little research has investigated how teacher efficacy, burnout and teacher identity intersect in the lives of teachers across their careers. In addition, no work has approached the experiences of teachers using praxis shock as a way to investigate the interrelationships that might exist between self-efficacy, wellbeing and burnout.

This study examined the perceptions of music teachers specifically, as it has been consistently argued that the experiences of early career music teachers are characterized by practical challenges associated with teaching the discipline area (Krueger, 2001; Scheib, 2006; Ballantyne, 2007; Conway, 2012; Legette, 2013). This is implicated in the findings by Buonomo et al. (2017), who argue that the contextual stressors provided by teaching roles which require extracurricular responsibilities and workload (like music teachers) are at greater risk of burnout. Much has also been made of the particular nature of music teacher professional identity, whereby performing music (or seeing oneself as a musician), is central to the conception of success as a teacher (Mark, 1998; Ballantyne, 2005, 2007; Ballantyne and Grootenboer, 2012; Ballantyne et al., 2012; Ballantyne and Zhukov, 2017). This work thus extends previous work into the development of productive music teacher identities in pre-service in-service music teachers (Ballantyne, 2007; Ballantyne et al., 2009, 2012; Ballantyne and Zhukov, 2017).

This article explores (1). how teachers' self-reported levels of burnout and self-efficacy, well-being and praxis shock vary as a function of time in the profession, and (2). the effect of praxis shock and the extent to which it can influence the reported experiences of burnout, well-being and self-efficacy in practicing music teachers. Information was gathered around the nature of music teachers' experiences across the career span, so as to explore variation in these experiences at different career stages. Specifically, we predicted that reports of praxis shock and burnout would be more common in earlier career

teachers and decrease across the career span, while wellbeing and self-efficacy would have the opposite relationship. Investigating the relationships between these factors create improved understandings around the experiences of music teachers throughout their career. The findings of this paper thus may be used to inform support mechanisms for teachers who may be experiencing difficulties in their jobs, and are at risk of 'burning out (Day and Gu, 2007).

### MATERIALS AND METHODS

A questionnaire was designed to explore the experiences of music teachers as they progressed through their career.

# Development of the Questionnaire and Conceptual Frame

The questionnaire comprised scales investigating music teacher self-efficacy and socialization into the profession, specifically looking at praxis shock, well-being, burnout and self-efficacy and how they change over time.

#### Praxis Shock

Measuring the experience of praxis shock provides a primary contribution to previous research in this field. The subscale, developed for the purpose of this study, was based on a review of prior theoretical work (Mark, 1998) and qualitative work (Kelchtermans and Ballet, 2002; Ballantyne, 2007) where praxis shock was associated with early-career teachers' surprise or shock in relation to aspects of socialisation, workload, isolation, induction and relationships with students and staff. We considered praxis shock to be the feeling of surprise, shock or disequilibrium that results from a discrepancy between an expectation of a workplace environment and what it should be like, and the reality which may be somewhat different. The questions relating specifically to praxis shock measured 1. the degree of shock or surprise associated with these different components of teaching life, and 2. the degree to which there was dissonance between expectations of teaching life and reality experienced (see **Appendix A**).

#### Burnout

Burnout was examined using items from the Copenhagen Burnout Inventory (CBI) (Kristensen et al., 2005) which defines this construct as "a state of physical, emotional, and mental exhaustion that results from long-term involvement in work situations that are emotionally demanding" (2005, p. 196). The current questionnaire incorporated the CBI items measuring work-related and client-related burnout. CBI items relating to a third domain — personal burnout — were omitted due to redundancy with other questionnaire items. It is notable that the CBI was originally designed to be applicable across a range of work contexts: the "work" in this instance is teaching, and the "clients" are the students. Work-related burnout is defined as the "degree of physical and psychological fatigue and exhaustion that is perceived by the person as related to his/her work", while client (student)-related burnout is defined as the "degree of physical and psychological fatigue and exhaustion that is perceived by the person as related to his/her work with clients" (Kristensen et al., 2005, p. 197). Importantly, the items are intended to measure a direct attribution of work or clients to one's state of fatigue, and not another external factor.

Items measuring work-related burnout were divided into two parts. The first three questions related to intensity of burnout (specifically in relation to teaching) and were measured on a 5 point Likert scale. The final four questions probed the frequency of work burnout on a 5-point Likert scale (see **Appendix A**). The student-related burnout included items explored the extent to which students contributed to teachers' burnout, and the frequency of the student-related burnout. These were also ranked on a 5-point scale. The full complement of questions referred to in this paper (including division into student-related burnout and work-related burnout) are found in **Appendix A**.

#### Well-Being

The questionnaire utilized the World Health Organisation-Five Well-Being Index (1998) version). All items included in this five-item scale are positively worded and designed to measure psychological well-being through such dimensions as mood, vitality and general interest in life<sup>1</sup> (see **Appendix A**). It is scored on a 6-point scale, which was reversed for this questionnaire to reflect the same order as the scales in the other questions. The six points range from 0 – At no time, to 5 – All of the time.

#### Self-Efficacy

Self-efficacy was investigated as it related to teaching activities. Teaching self-efficacy is also related to the individual teacher's estimation of their resilience to changes in circumstances and context. This component of the questionnaire was drawn from the Music Careers Questionnaire II (Hargreaves et al., 2007), although a 6-point Likert scale (strongly disagree to strongly agree) has been utilized. The wording of these items was specifically phrased "in terms of can do rather than will do", to tap "perceived capability" in accordance with Bandura's guidelines (2006, p. 308). The 17 statements covered elements of teaching practice such as lesson planning (e.g., "When I plan lessons, I am certain I can make them work"), perseverance (e.g., "If a lesson goes poorly the first time, I try again until it works better"), avoidance (e.g., "I avoid facing difficult situations in my teaching"), determination (e.g., "When I decide to do something, I go right to work on it") and confidence (e.g., "I feel insecure about my teaching; I am a self-reliant teacher").

Tschannen-Moran and Woolfolk-Hoy (2001) argue that teachers with greater self-efficacy are more able to cope with change. Thus, self-efficacy was also investigated in terms of respondent's self-reported ability to "cope with change" and respondents rated confidence in their capacity to demonstrate a range of skills and behaviors relating to coping with change in schooling environments. This component of the questionnaire borrowed from a bank of questions developed by Skaalvik and Skaalvik (2010) in their investigation of teacher self-efficacy and teacher burnout, and is designed to investigate how "certain"

<sup>1</sup>http://www.who-5.org

TABLE 1 | Reliability of measures.


participants were about various aspects of teaching. A 7-point Likert scale was used with modification (Skaalvik and Skaalvik did not specify descriptors for 2, 4 and 6 in the Likert scale, so for the purpose of this study, the complete range of descriptors was assumed): 1– Not at all certain, 2 – Very uncertain, 3 – Quite uncertain, 4 – Neither certain or uncertain, 5 – Quite certain, 6 – Very certain, 7 – Absolutely certain. The categories have been used as they were in the original, and are from the cluster of questions related to coping with change. They are: "successfully use any instructional method that the school decides to use"; "manage instruction regardless of how it is organized"; "manage instruction even if the curriculum is changed", and; "teach well even if you are told to use instructional methods that would not be your choice". Again, the "can" and "I" wording utilized is reflective of the guides suggested by Bandura (2006). The full complement of questions referred to in this paper (including division into self-efficacy and self-efficacy: cope with change) are found in **Appendix A**.

#### Reliability

The reliability of all measures was assessed using Cronbach's alpha. All measures had a reliability greater than 0.8 (see **Table 1**).

## Participants and Procedure

In order to address the research questions, an online questionnaire was distributed to music teachers via an online community of practice, advertisements in professional list-serves, social media, and professional associations. This questionnaire was sent out over 3 years, in 2012, 2013, and 2014.

A total of 1021 respondents attempted the questionnaire with 836 people completing at least 50% of the questionnaire. Participants who completed less than 50% of the questionnaire were removed from all analyses. Because of the ethics requirements associated with voluntary participation in online questionnaires, if someone chose not to answer a question, they were excluded from the analyses that utilize that item. This explains the varied number of responses in each item of the analyses presented.

As **Table 2** shows, 81% of respondents were female, with 19% male. The vast majority of respondents (85%) came from Australia, with additional respondents from the United States (9%) and other countries (6.2%). Of the Australian respondents, the majority came from Queensland (45%), followed by New South Wales (20%), Western Australia (16%), and Victoria (11%). The sample had a large number of younger (under 25) and older (51–60) respondents (25% in each instance), with the remaining 50% of respondents aged 26–50. 63% of respondents taught at one school, and 20% taught at two schools, with the majority teaching at State Schools (55%), and 29% teaching at Christian or Catholic Schools. 31% of respondents had completed some sort of dual degree in Music or Arts and Education. 20% had completed straight degrees in Education. 26.2% had completed a Bachelor of Music, and 22.9 had completed a Graduate Diploma of Education. 36.8% had completed another (unnamed) degree.

# Missing Data

The data is not as precise for the latter period of the career. In the initial iteration of the questionnaire (2012), respondents who selected "more than 20 years teaching" were given a code of "21" for "years of teaching". In subsequent iterations (2013 and 2014), additional questions associated with how long they had been teaching, in which year they began teaching and the year of their graduation from university. For analysis purposes, cases of yearsteaching greater than 21 were reclassified as 21 years-teaching. That is, 21 years was the upper limit on the years teaching scale.

# RESULTS AND DISCUSSION

# How Teachers' Self-Reported Levels of Burnout and Self-Efficacy, Well-Being and Praxis Shock Vary Over Time in the Profession

Regression analyses were used to evaluate the degree to which the constructs of interest (e.g., praxis shock, self-efficacy, wellbeing, and burnout) varied as a function of time spent teaching in years. Participants' means score on each measure were used in the regression analyses. All analyses were run using Stata (15) software.

Linear regression analyses revealed significant (p < 0.05) linear relationships between years teaching and all the outcome variables except for self-efficacy: cope with change. Consistent with our predictions, self-efficacy and well-being were positively related to years teaching, while work-related burnout, studentrelated burnout, and praxis shock were negatively related to years teaching (see **Table 3**). **Figure 1** provides graphical representations of the linear models described here. Regarding the assumptions of the linear models, in all cases, the error variance was constant as a function of years teaching. That is, the assumption of homoscedasticity was met. Normality was assessed using quantile-quantile plots (see **Figure 2**). Both selfefficacy: cope with change and student-related burnout show some divergence from normality, though only in the tails. Neither a natural log or square root transformation could correct this. Otherwise, the model residuals appear normally distributed.

Early career respondents in our sample tended to report increased levels of praxis shock which declined steadily with time spent within the profession. However, it is important to note that this linear decline was small, and it is clear that praxis shock remained an issue beyond the early years in the profession. Indeed, there are clear cases of respondents who have taught for 21 years or more experiencing high levels of praxis shock. This is

Linking Well-Being, Burnout, Self-Efficacy and Praxis Shock

fpsyg-10-02255 February 17, 2020 Time: 16:58 # 5


 For Australian Respondents,

State n = 477

At how many

schools do you Course completed Type of teacher

n = 1328∗

n = 1063∗

teach? n = 456

Country n = 666

TABLE 2|Demographic information of respondents.

> Age<sup>2</sup> n = 665

n = 632

 Gender n = 675 School type

Years teaching<sup>1</sup>

n = 702 ∗Values exceed the total sample size as respondents could provide multiple responses to many of these questions. <sup>1</sup>It should be noted that responses were coded at '21' years of teaching when respondents selected 'more than 20 years teaching' in the first year of the questionnaire (2012). Although the questionnaire was initially designed for early-career teachers, when it became apparent that this questionnaire was applicable and interesting to more experienced teachers, subsequent iterations asked respondents a series of questions related to years in the profession, to more accurately assess their number of years teaching. Respondents were asked how long they had been teaching, in which year they began teaching, and the year of their graduation from university. In some cases, respondent's age did not match with the number of years that they said they had been teaching, and the years since they graduated university. This is because of the prevalence of private tuition as a source of income during university for music students. However, for the purposes of this study, we considered the number of years teaching to be the number of years teaching since graduation, and therefore adjusted the number of years teaching variable to reflect this. <sup>2</sup>When the data from 2012 was being cleaned, there was the need to assign an actual value for those who were outside of the range option provided for 'age.' It was deemed important to include those people in the sample (those who had provided an age of 'less than 20 years,' and those who had provided an age of 'more than 50 years'), and they were given a nominal age of '19' and '51' respectively. Those respondents who chose not to reveal their age were removed from the 'age' sample. This is standard practice in the cleaning of data. This ensured that the merging of the data from 2012, 2013 and 2014 was consistent.


an interesting phenomenon that was not anticipated in this study and is not previously documented. Persistence of praxis shock may be evident for a number of reasons, but is likely to reflect the fact that praxis shock can emerge when there is a discrepancy between the expectation of teaching life, and the realities faced by teachers. That this was a reality for experienced teachers as well as early career teachers, was unexpected. Praxis shock amongst experienced teachers may reflect the distance between the changing philosophies and expectations held by statutory bodies and management, and which teachers are expected to embody. Or perhaps music teachers commonly find themselves in new (more senior) roles and with new duties, irrespective of teaching experience. Any dissonance between their perceived skill set and the skills required may prompt praxis shock and burnout. Indeed, the praxis shock observed in this data set may be related to recent changes in terms of requirements, and the changes in relation to "what teaching is all about" in many schools. The current data do not shed light on this, but certainly provides the impetus for further investigation and at best may indicate an unexpected need for support for this demographic of the profession.

In summary, the series of linear regressions reported above show that teachers' self-reported levels of burnout, self-efficacy, well-being, and praxis shock vary over their time in the teaching profession. More specifically, we found that self-efficacy and well-being showed a small steady increase with the number of years teaching. While praxis shock, work, and student-related burnout behaved in the opposite manner, where both decreased slightly with number of years teaching.

# How the Constructs of Well-Being, Burnout and Self-Efficacy Relate to Teachers' Experiences of Praxis Shock

The similarities observed between the linear models above suggest that there may be some degree of interrelationship between praxis shock and the other psychological constructs. The working hypothesis was that experience with praxis shock may account for differences in well-being, student- and workburnout and self-efficacy across the teaching career. While it has been argued that the experience of praxis shock can result in burnout (Ballantyne, 2007; Ballantyne and Zhukov, 2017), and is also associated with attrition and reduced job satisfaction (Kelchtermans and Ballet, 2002; Ballantyne, 2005, 2007; Hong, 2010; Skaalvik and Skaalvik, 2010), existing literature is restricted to qualitative documentation of the existence of praxis shock in teachers (Stokking et al., 2003; Smagorinsky et al., 2004; Ballantyne, 2007), indicating that it might be linked to well-being, student- and work-related burnout, and selfefficacy. As such, we ran a series of mediation analyses to test whether praxis shock might mediate the relationship between years of teaching and self-efficacy, well-being, and student- and work-related burnout.

#### Mediation Analyses

Mediation analyses were run to test whether the relationship between years in the profession and all other factors was mediated by praxis shock. In all cases, the relationship was significant (see **Table 4**), with praxis shock partially mediating the direction of the outcomes over time. The one exception was student-related burnout where we observed a full mediation of the relationship.

The mediation analyses provide moderate evidence that praxis shock is a contributing factor to the relationship between years teaching and our outcomes variables self-efficacy, wellbeing and work- and student-related burnout. That is, as the degree of praxis shock reduces as teachers stay in the profession, this reduction partially mediates the relationship between increased years of teaching experience, and the outcomes of burnout, self-efficacy, and well-being. Higher levels of praxis shock contribute to increased levels of burnout and decreased levels of well-being and self-efficacy and vice versa. In characterizing the degree of evidence as moderate, we are acknowledging that the mediation analyses are close to the significance threshold. Although scientific conclusions should not be based on whether p-values passes a specific threshold, future studies may seek to replicate these effects to confirm their reliability.

#### Exploratory Analyses

Finally, we ran one last set of exploratory analyses. It's clear from **Figure 1** that a large amount of variance in the data is unaccounted for by the simple linear models. That is, the relationship between praxis shock and years teaching appears more nuanced than a simple linear one. Certainly, from a theoretical perspective, we might expect the relationships to vary between early- versus late-career teachers. Most of the previous literature has focussed on the existence of praxis shock in the early years of the profession, and in particular the experiences of early career music teachers (Ballantyne, 2007; Ballantyne and Zhukov, 2017). This is logically because it focusses on the discrepancies between university preparation and expectations prior to graduation, and the realities of teaching. To explore possible subtle differences in the relationships explored

TABLE 4 | Statistical output for the mediation tests investigating praxis shock as a mediator of the relationship between years teaching and significant outcomes.


above between the early and late career teacher cohorts, we split the data according to early career (0–5 years) and more advanced career teachers (>5 years). A series of Pearson's correlations between years teaching and our psychological constructs of interest were run for earlier (≤5) and later career teachers (>5). We stress that this set of correlation analyses are very much exploratory, run in response to the complex pattern of data observed above. The results are reported in **Table 5**.

The key result from **Table 5** is that, in some cases, the relationship between years teaching and the respective psychological construct appears to differ as a function of early versus later career. Specifically, for well-being, work burnout, and praxis shock, the direction of the relationship for early and later career teachers is reversed. Well-being looks to decrease very slightly in the first 5 years of teaching and increase thereafter. The opposite is true for praxis shock and work burnout. No effect of self-efficacy or self-efficacy: cope with change was observed here. It appears that early-career teachers are most vulnerable to experiencing praxis shock and burnout. However, increased time (beyond ∼5 years) in the profession may bring only a small reprieve from these experiences.

TABLE 5 | Bivariate relationships between number of years teaching and all outcomes for early-career, mid-career, and late-career periods in the profession.


# CONCLUSION

This paper explored the complexities that lie within the relationships between self-reported levels of self-efficacy, praxis shock, burnout, and well-being for music teachers, and time in a career. This study provides evidence that levels of burnout, well-being, self-efficacy, and most importantly praxis shock, are influenced by the number of years that teachers have been teaching. Additionally, levels of praxis shock predict levels of burnout, well-being and self-efficacy.

Our findings demonstrate that there are distinct challenges in the early career and late career periods of a music teacher's working life. Day and Gu (2007) found that teachers who were 24–30 years post-graduation found themselves in a phase where they have to adjust to changes associated with "deteriorating pupil behaviors, adverse personal events, resentment at 'being forced to

jump through hoops by a constant stream of new initiatives, and career stagnation". Teachers in this life phase are at greater risk of being less effective, largely due to the likelihood of experiencing "extreme professional life phase scenarios (p. 437). It is highly likely that the experiences both early in the career and later in the career are associated with attrition from the profession. At the very least, providers of pre-service teacher education and inservice teacher education would be well advised to consider the impact of praxis shock across the career in terms of burnout and well-being.

As praxis shock is largely about the dissonance between expectations and the realities of working life, it is likely that awareness of the trends identified in this study will enhance awareness of the support necessary for teachers across the career - where praxis shock appears to be evident even later into the career. Targeted career-stage support needs to be developed for teachers, to assist them in navigating difficult periods successfully. Perhaps the development of a "proactive praxis" strategy in in-service teacher support will enable the pre-empting of the likelihood of praxis shock. For example, Gu and Day's (2007) work examining teacher identities and life stages suggest that in-service support focussing on ways to address external policy initiatives, helped more experienced teachers maintain enthusiasm and commitment to teaching during periods of praxis shock. Another suggestion is the use of strategies from positive psychology - qualitative findings with early-career teachers suggest that resilience developed through a positive psychology approach is indicated as protective against the negative impacts of praxis shock (Ballantyne and Zhukov, 2017). It is likely that support designed for experienced teachers might initially address similar issues to those addressed early in the career, but further research is warranted to investigate the nature of later-career praxis shock, in order for a nuanced support approach to be developed.

Support strategies for pre-service teachers are equally important. Demystifying the teaching experience and actively encouraging pre-service teachers to engage critically with their teaching experiences is likely to be key to negating praxis shock. Support strategies might involve addressing well-being and selfefficacy explicitly during university education, so that prior to embarking on their career, pre-service teachers might be better equipped to navigate successfully through the difficult first five years, minimizing their likelihood of experiencing burnout.

As identified by Hong (2010), emotional burnout is critical in influencing success in the profession. The current study goes further by showing that reported praxis shock impacts on burnout. It should be noted that the findings by Fiorilli et al. (2017) indicate that teachers who may be undergoing the experience of praxis shock may interpret students' actions in a more negative way - this may be associated with student-related burnout.

The links found between praxis shock and self-efficacy echo the arguments put forward by Gibbs (2001), who claimed that self-efficacy is the key to managing anxiety and stress associated with praxis shock. Gibbs emphasized that strategies specifically aimed at supporting teachers should result in enhanced self-efficacy.

This study also raises questions about other similar professions where burnout has been identified at the beginning of the career (for example, in health professions, psychology, law enforcement, legal, and even academia). Indeed, as this paper looked exclusively at teachers identifying as music specialists, it would be interesting to see to what extent this trend is evident in teachers with other specializations, and even generalist primary teachers. The praxis shock questionnaire items developed for the current study provide a useful tool for the investigation of other professions.

Our data do not reveal whether some of the issues have to do with time in the teaching role or perhaps changes to the nature of teaching over time. To address this question, a longitudinal study is required. The changing nature of schools, teaching methods, accountability demands, and technology infiltrating the classroom may damage self-efficacy for teachers who have been in the industry for many years (>5), and this may be the reason for persistent praxis shock in the later parts of the career. In addition, the varied contexts from which the respondents came (including across different countries), means that contextualization in terms of common experiences is not possible. Indeed, even if the data were limited to one country, common experiences are unlikely to be evident; even within Australia, the experiences of music teachers vary from school to school. Future analyses of the longitudinal data (the 2nd and 3rd phases of questionnaire analysis with the same respondents) are likely to shed more light on how the experience of praxis shock and burnout, as well as teachers' well-being and self-efficacy change over time.

It should also be noted that this data was collected from teachers who were currently teaching, and it therefore, does not include those teachers who had left the profession. Hong (2010) found that those teachers who did burn out and consequently left teaching were more likely to admit to being extremely stressed about the work itself (work burnout in this study). Also evident in Hong's research is the impact of student-related stress (p. 1537). Although teachers who remained in the profession reported less emotional burnout, it did not seem to affect their professional identities and lives (Hong, 2010).

The apparent occurrence of praxis shock throughout the career is likely to have an impact on the development of sustainable and productive professional identities. In the case of music teachers, it has been argued that how music teachers think about themselves as musicians and as educators, and how proficient they are in these areas (their self-efficacy), is demonstrably associated with their identities (Hargreaves et al., 2007; Ballantyne and Grootenboer, 2012; Ballantyne et al., 2012; Ballantyne and Zhukov, 2017). Professional identity development has consistently been related to self-efficacy (Hargreaves et al., 2007) and teacher identity arguably has a prominent role to play in teacher praxis (Beijaard et al., 2004; Sachs, 2005; Diez and Raths, 2007). More needs to be known about how professional contexts, praxis shock and identity interrelate, and how these factors over the career play a role in the development of confidence and effectiveness as a professional. Investigating the nuances in these interrelationships is an area for future research and one that will hopefully emerge from the analysis of the fuller data from this project.

# DATA AVAILABILITY STATEMENT

The datasets generated for this study will not be made publicly available as it was a condition of ethical approval that only the authors are able to access the data.

# ETHICS STATEMENT

fpsyg-10-02255 February 17, 2020 Time: 16:58 # 10

The studies involving human participants were reviewed and approved by the University of Queensland Ethics Committee. Participants read a full information statement about the study, and gave their consent to participate prior to beginning the questionnaire. The participants provided their written informed consent to participate in this study.

# REFERENCES


### AUTHOR CONTRIBUTIONS

JB led all aspects of this project and writing of the manuscript. JR assisted with the statistical analysis.

# ACKNOWLEDGMENTS

The authors gratefully acknowledge the support for this project provided by the University of Queensland Foundation Research Excellence Awards. Many thanks also go to the research assistants who worked on the project – Michael Ireland, Tammie Olm-Madden, Claire Petherick, Nigel Bond, Rachael Dwyer, and Katie Zhukov. For this paper, the support provided by Michael Ireland for the initial work establishing the scales for the questionnaire was particularly appreciated.


House/Employment\_Education\_and\_Training/TeachingProfession/Media\_ Releases (accessed November 11, 2019).


**Conflict of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2020 Ballantyne and Retell. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

#### Ballantyne and Retell Linking Well-Being, Burnout, Self-Efficacy and Praxis Shock

#### APPENDIX A | Questionnaire items.

fpsyg-10-02255 February 17, 2020 Time: 16:58 # 12


(Continued)

#### Ballantyne and Retell Linking Well-Being, Burnout, Self-Efficacy and Praxis Shock

#### APPENDIX A | Continued

fpsyg-10-02255 February 17, 2020 Time: 16:58 # 13

