# NEW PERSPECTIVES ON DOMESTIC VIOLENCE: FROM RESEARCH TO INTERVENTION

EDITED BY : Luca Rollè, Shulamit Ramon and Piera Brustia PUBLISHED IN : Frontiers in Psychology and Frontiers in Sociology

#### Frontiers Copyright Statement

© Copyright 2007-2019 Frontiers Media SA. All rights reserved. All content included on this site, such as text, graphics, logos, button icons, images, video/audio clips, downloads, data compilations and software, is the property of or is licensed to Frontiers Media SA ("Frontiers") or its licensees and/or subcontractors. The copyright in the text of individual articles is the property of their respective authors, subject to a license granted to Frontiers.

The compilation of articles constituting this e-book, wherever published, as well as the compilation of all other content on this site, is the exclusive property of Frontiers. For the conditions for downloading and copying of e-books from Frontiers' website, please see the Terms for Website Use. If purchasing Frontiers e-books from other websites or sources, the conditions of the website concerned apply.

Images and graphics not forming part of user-contributed materials may not be downloaded or copied without permission.

Individual articles may be downloaded and reproduced in accordance with the principles of the CC-BY licence subject to any copyright or other notices. They may not be re-sold as an e-book.

As author or other contributor you grant a CC-BY licence to others to reproduce your articles, including any graphics and third-party materials supplied by you, in accordance with the Conditions for Website Use and subject to any copyright notices which you include in connection with your articles and materials.

All copyright, and all rights therein, are protected by national and international copyright laws.

The above represents a summary only. For the full conditions see the Conditions for Authors and the Conditions for Website Use. ISSN 1664-8714 ISBN 978-2-88945-870-7 DOI 10.3389/978-2-88945-870-7

#### 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

# NEW PERSPECTIVES ON DOMESTIC VIOLENCE: FROM RESEARCH TO INTERVENTION

Topic Editors: Luca Rollè, University of Turin, Italy Shulamit Ramon, University of Hertfordshire, United Kingdom Piera Brustia, University of Turin, Italy

Citation: Rollè, L., Ramon, S., Brustia, P., eds. (2019). New Perspectives on Domestic Violence: From Research to Intervention. Lausanne: Frontiers Media. doi: 10.3389/978-2-88945-870-7

# Table of Contents

*05 Editorial: New Perspectives on Domestic Violence: From Research to Intervention*

Luca Rollè, Shulamit Ramon and Piera Brustia

#### SECTION 1

#### FACTORS ASSOCIATED WITH DOMESTIC VIOLENCE / INTIMATE PARTNER VIOLENCE / FEMINICIDE


Luca Rollè, Giulia Giardina, Angela M. Caldarera, Eva Gerino and Piera Brustia


# SECTION 2

# INTIMATE PARTNER VIOLENCE AND DOMESTIC VIOLENVE IN SPECIFIC CONTEXT


Eva Gerino, Angela M. Caldarera, Lorenzo Curti, Piera Brustia and Luca Rollè

*119 The Importance of Culture in Addressing Domestic Violence for First Nation's Women*

Donna M. Klingspohn

# SECTION 3

# THE IMPACT OF INTIMATE PARTNER AND DOMESTIC VIOLENCE

*126 Informal Caregiving Relationships in Psychosis: Reviewing the Impact of Patient Violence on Caregivers*

Juliana Onwumere, Zheng Zhou and Elizabeth Kuipers


# SECTION 4

# INTERVENTION AND PREVENTION ON IPV/DV

*160 The Willingness to Intervene in Cases of Intimate Partner Violence Against Women (WI-IPVAW) Scale: Development and Validation of the Long and Short Versions*

Enrique Gracia, Manuel Martín-Fernández, Miriam Marco, Faraj A. Santirso, Viviana Vargas and Marisol Lila

*176 New Directions for Preventing Dating Violence in Adolescence: The Study of Gender Models*

Chiara Santoro, Belén Martínez-Ferrer, Carmen Monreal Gimeno and Gonzalo Musitu

*181 Intimate Partner Violence and Child Custody Evaluation: A Model for Preliminary Clinical Intervention*

Marialuisa Gennari, Giancarlo Tamanza and Sara Molgora

# Editorial: New Perspectives on Domestic Violence: From Research to Intervention

Luca Rollè<sup>1</sup> \*, Shulamit Ramon<sup>2</sup> and Piera Brustia<sup>1</sup>

<sup>1</sup> Department of Psychology, University of Turin, Turin, Italy, <sup>2</sup> School of Health and Social Work, University of Hertfordshire, Hertfordshire, United Kingdom

Keywords: domestic violence, intimate partner abuse, intimate partner violence (IPV), gender violence against women, same sex intimate partner violence, systematic review, perpetrator and victim of violence, perpetrator

**Editorial on the Research Topic**

#### **New Perspectives on Domestic Violence: from Research to Intervention**

In a document dated June 16th 2017, the United States Department of Justice stated that Domestic Violence (DV) has a significant impact not only on those abused, but also on family members, friends, and on the people within the social networks of both the abuser and the victim. In this sense, children who witness DV while growing up can be severely emotionally damaged. The European Commission (DG Justice) remarked in the Daphne III Program that 1 in 4 women in EU member states have been impacted by DV, and that the impact of DV on victims includes many critical consequences: lack of self-esteem, feeling shame and guilt, difficulties in expressing negative feelings, hopelessness and helplessness, which, in turn, lead to difficulties in using good coping strategies, self-management, and mutual support networks. In 2015 the EU Agency for Fundamental Rights affirmed that violence against women can be considered as a violation of human rights and dignity. Violence against women exists in each society and it can be related to any social, economic and cultural status and impact at the economic level. It includes physical, sexual, economic, religious, and psychological abuse.

Although men experience domestic violence by women, the rate of DV among women is much higher than that of men, especially in the category of being killed due to DV.

Recent studies have shown that between 13 and 61% of women (15–49 years old) report to have been physically abused at least once by an intimate partner. Domestic Violence takes place across different age groups, genders, sexual orientations, economic, or cultural statuses. However, DV remains largely under-reported due to fear of reprisal by the perpetrator, hope that DV will stop, shame, loss of social prestige due to negative media coverage, and the sense of being trapped with nowhere to go:

Hence, it is estimated that 90% of cases of DV continue to be identified as a nondenounced violence.

The aim of this Special Issue of Frontiers of Psychology is to gather updated scientific and multidisciplinary contributions about issues linked to domestic violence, including intimate partner violence (IPV). We encouraged contributions from a variety of areas including original qualitative and quantitative articles, reviews, meta-analyses, theories, and clinical case studies on biological, psycho-social and cultural correlates, risk and protective factors, and the associated factors related to the etiology, assessment, and treatment of both victims and perpetrators of DV.

We hope that this Special Issue will stimulate a better informed debate on Domestic Violence, in relation to its psychosocial impact (in and outside home, in school, and workplace), to DV prevention and intervention strategies (within the family and in society at large), in addition to specific types of DV, and to controversial issues in this field as well.

#### Edited and reviewed by:

Kath Woodward, The Open University, United Kingdom

> \*Correspondence: Luca Rollè l.rolle@unito.it

#### Specialty section:

This article was submitted to Gender, Sex and Sexuality Studies, a section of the journal Frontiers in Psychology

> Received: 25 February 2019 Accepted: 07 March 2019 Published: 28 March 2019

#### Citation:

Rollè L, Ramon S and Brustia P (2019) Editorial: New Perspectives on Domestic Violence: From Research to Intervention. Front. Psychol. 10:641. doi: 10.3389/fpsyg.2019.00641

**5**

The Special Issue comprises both theoretical reviews and original research papers. 7 research papers, 6 reviews (policy and practice review, systematic review, review and mini-review) and 1 methodological paper are included.

The first section comprises 2 systematic review and 3 original research papers focused on factors associated with Domestic Violence/Intimate Partner Violence/feminicide. Velotti et al. conducted a systematic review focused on the role of the attachment style on IPV victimization and perpetration. Several studies included failed to identify significant associations. The authors suggest to consider other variables (e.g., socioeconomic condition) that in interaction with attachment styles could explain the differences found between the studies. Considering the clinical contribution that these findings can provide to the treatment of IPV victims and perpetrators, future studies are needed. From a systematic review conducted by Gerino et al. focused on IPV in the "golden age" (old age), economic and educational conditions, younger age (55–69), membership in ethnic minorities, cognitive and physical impairment, substance abuse, cultural and social values, sexism and racism, were found as risk factors; depression emerged as risk factor and consequence of IPV. However, social support was identified as main protective factor. Also help-seeking behaviors and local/national services had a positively impact the phenomenon. Furthermore, the role of the parental communication was highlighted (Rios-González et al.) In that mothers encourage daughters to engage in relationship with ethical men, while removing from their representation attractive features and enhancing the double standard of viewing ethical man as unattractive vs. violent and attractive man. Fathers' communication directed toward young boys supports the dominant traditional masculinity, objectifying girls and emphasizing chauvinist values. These communicative dynamics impact males' behavior and females' choice of the partner while increasing the attraction toward violent men, and thus influencing the risk to be involved in IPV episodes.

Furthermore, factors associated with multiple IPV victimization by different partners were identified. From the study of Herrero et al., experiencing child abuse emerged as a main predictor ("conditional partner selection process"). Similarly, adult victimization perpetrated by other than the intimate partner influences multiple IPV episodes. Moreover, this phenomenon is more frequent among younger women and those with lower income satisfaction. Length of relationship and greater psychological consequences to previous IPV are positively associated with multiple IPV episodes, while previous physical abuse is negatively related with subsequent victimization. The risk of multiple IPV episodes is reduced in countries with greater human development, suggesting the role of structural factors.

Regarding reasons of feminicide, passion motives assume the main role, followed by family problems, antisocial reasons, predatory crimes that comprise sexual component, impulsivity and mental disorders. The risk of overkilling episodes is higher when the perpetrator is known by the victim and when the murder is committed for passion reasons (Zara and Gino).

The second section includes papers focused on IPV/DV in particular contexts (one research paper, two reviews). Within separated couples, where conflicts are common, both men and women experience psychological aggression. However, some particularities emerged: women started to suffer of several kinds of psychological violence that was aimed to control (complicating the separation process), dehumanize and criticize them. Men report only few forms of violence experienced (likely due to the men's social position that narrows their disclosure opportunity), which mainly concern the limitation of the possibility to meet children (Cardinali et al.). Regarding same-sex couples (Rollè et al.), both similarities and differences in comparison with heterosexual couples emerged. IPV among LGB people is comparable or even higher than heterosexual episodes. Unique features present in same-sex IPV concern identification and treatment aspects, mainly due to the absence of solutions useful in addressing obstacles to help-seeking behaviors (related to fear of discrimination within LGB community), and the limitation of treatment programs tailored to the particularities of the LGB experience. Similarly, within First Nation's communities in Canada, IPV is a widespread phenomenon. However, the lack of preventing programs and the presence of intervention solutions that fail to address its cultural origins, limit the reduction of the problem and the recovery of victims. Klingspohn suggests the development of interventions capable to guarantee cultural safety and consequently to reduce discrimination and marginalization that Aboriginal people experience with mainstream health care system and which limit help-seeking behaviors.

The third section comprises two reviews and one research paper concerned with the impact of Intimate Partner and Domestic Violence. The systematic review conducted by Onwumere et al. highlighted the financial and emotional burden that violence perpetrated by psychotic patients entails for their informal carers (mainly close family relatives). Moreover, the authors identified within the studies included positive association between victimization and trauma symptoms, fear, and feeling of powerless and frustration.

Among people who suffered of Domestic Violence with a romantic or non-romantic partner who became their stalker, stalking victimization entails physical and emotive consequences for both male and female victims. Females suffered more than males of depressive and anxiety symptoms (although for both genders symptoms were minimal), while males experienced more anger. Furthermore, both genders adopted at least one "moving away" strategy in coping with stalking episodes, and the increasing of stalking behaviors determined a reduction in coping strategies use. This latter finding is likely to be due to the distress experienced (Acquadro Maran and Varetto).

Children abuse—which occurs often in Domestic Violence results in emotional trauma as well as physical and psychological consequences that can negatively impact the learning opportunities. The school staff's ability to identify abuse signals and to refer to professionals constitute their main role. However, lack of skills and confidence among teachers regarding this function emerged, and further training for the school staff to increase support provided to abused children is needed (Lloyd).

Lastly, the fourth section includes two papers (one review and one methodological paper) that provide information on intervention and prevention programs and one research paper which contributes to the development and validation of the Willingness to Intervene in Cases of Intimate Partner Violence Against Women (WI-IPVAW) Scale. Gracia et al. The instrument demonstrated—both in the long and in its short form—high reliability and construct validity. The development of WI-IPVAW can contribute to the evaluation of the t role that can be played by people who are aware of the violence and understand attitudes toward IPV that can influence perpetrator's behavior and victim disclosure. The origin of violence within intimate relationship during adolescence calls for the development of preventive programs able to limit the phenomenon. The mini-review conducted by Santoro et al. highlighted the necessity to consider the relational structure where women are involved (history of poly-victimization re-victimization), and the domination suffered according to the gender model structured by the patriarchal context. Moreover, considering that violence can occur after separation or divorce, requires in child custody cases the evaluation of parenting and co-parenting relationship. This process can provide an opportunity to assess and treat some kind of violent behavior (Conflict-Instigated Violence, Violent Resistance, Separation-Instigated Violence). According to these consideration, Gennari et al. elaborated a model for clinical intervention (relational-intergenerational model) useful to address these issues during child custody evaluation. The model is composed of three levels aimed at understanding intergenerational exchange and identify factors that contribute to safeguard family relationship. This assessment process allows parents to reflect on information emerged during the evaluation process and activate resources useful to promote a constructive change of conflict dynamics and violent behaviors.

# AUTHOR CONTRIBUTIONS

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

# ACKNOWLEDGMENTS

We would like to thanks all the authors and the reviewers who contributed to the present article collection, for their dedication to our topics and to their readiness to share their knowledge, and thus to increase the research in this field; KathWoodward, Specialty Chief Editor of Gender, Sex, and Sexuality Studies that believed in our project, and to Dr. Tommaso Trombetta for his collaboration during last year.

**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 Rollè, Ramon and Brustia. 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.

# Exploring Relationships: A Systematic Review on Intimate Partner Violence and Attachment

Patrizia Velotti <sup>1</sup> \*, Sara Beomonte Zobel <sup>2</sup> , Guyonne Rogier <sup>2</sup> and Renata Tambelli <sup>2</sup>

<sup>1</sup> Department of Educational Sciences, Università di Genova, Genoa, Italy, <sup>2</sup> Department of Dynamic and Clinical Psychology, Sapienza Università di Roma, Rome, Italy

Background: Intimate Partner Violence (IPV) is an important public health challenge. In recent years, there has been a greater awareness concerning this phenomenon, its causes and consequences. Due to the relational nature of IPV, attachment theory (Bowlby, 1988) appears a useful framework to read the phenomenon and to better understand its components and its dynamics to provide more precise and tailored interventions in the future.

#### Edited by:

Luca Rollè, Università degli Studi di Torino, Italy

#### Reviewed by:

Giuseppe Craparo, Kore University of Enna, Italy Rosaura Gonzalez-Mendez, Universidad de La Laguna, Spain

> \*Correspondence: Patrizia Velotti patrizia.velotti@unige.it

#### Specialty section:

This article was submitted to Gender, Sex and Sexuality Studies, a section of the journal Frontiers in Psychology

> Received: 25 April 2018 Accepted: 18 June 2018 Published: 05 July 2018

#### Citation:

Velotti P, Beomonte Zobel S, Rogier G and Tambelli R (2018) Exploring Relationships: A Systematic Review on Intimate Partner Violence and Attachment. Front. Psychol. 9:1166. doi: 10.3389/fpsyg.2018.01166 Purpose: To summarize our knowledge of the research concerning IPV and attachment with an aim to better design and implement future research.

Methods: Computer database researches were conducted using the following databases: Psychinfo, Psycharticle, Medline, Scopus, Web of Science, and PubMed (all years to the 01/02/2018). Search terms were compiled into two concepts for all database namely Attachment and IPV.

Results: After removing the duplicates, a total of 3,598 records was screened, resulting in the identification of 319 full-text articles to be further scrutinized. Upon closer examination, there was consensus that 113 of those studies met the study inclusion criteria. Data was organized considering specifically studies concerning (1) IPV victimization and attachment, (2) IPV perpetration and attachment (both these sections were articulated in Physical, Psychological, and Sexual IPV), and (3) New research (comprising same-sex couples, IPV and attachment in couple contexts and IPV profiles and attachment among perpetrators).

Conclusion: A number of studies failed to find significant associations between insecure attachment and IPV victimization or perpetration. Additional research is needed to provide a greater understanding of different IPV forms and to aid in the development of prevention and treatment interventions.

Keywords: attachment, intimate partner violence, systematic review, victimization, perpetration, mutual violence, homosexuality

**8**

# INTRODUCTION

#### Rationale

Intimate Partner Violence (IPV) is defined by World Health Organization (WHO) as "any behavior within an intimate relationship that cause physical, physical or sexual harm to those in the relationship" (Heise et al., 1999). The term IPV comprises different forms of violence, that go from manipulation to sexual coercion, that can be divided in three main categories: physical violence, psychological violence and sexual violence.

To understand violence, due to its complexity, the ecological model was applied: IPV seems to be a result of the explosive interaction between individual, relational, community and societal factors (Garcia-Moreno et al., 2005).

Physical and mental health are affected by IPV through both direct pathways, like wounds and injuries, and indirect pathways, like chronic health problems or psychological consequences of trauma and stress (Krug et al., 2002).

Due to the relational nature of IPV, we thought that Attachment Theory can be a useful framework to read the phenomenon and to better understand its components and its dynamics to provide more precise and tailored interventions in the future. At the end of the Eighties, attachment theory has also been used to investigate the quality of adult attachment relationships (Hazan and Shaver, 1987; Mikulincer and Shaver, 2007). Individual differences in adult attachment are assessed via self-report (e.g., Brennan et al., 1998) or interview (e.g., George et al., 1985; Velotti et al., 2011; Castellano et al., 2014). In both these traditions individuals can be classified into categories—secure, insecure-dismissing, insecure-preoccupied, disorganized—corresponding to those obtained among children. Also, research suggests that adult attachment is best described by two dimensions, avoidance, and anxiety (Shaver and Mikulincer, 2002; Mikulincer and Shaver, 2007). Individuals scoring high on the avoidance dimension are characterized by feelings of fear and uneasiness regarding intimacy as well as the difficulty to accept dependency on others within an affective bond (for example discomfort when the partner becomes too intimate or dependent). High scores on the anxiety dimension appear to reflect preoccupation about the reliability of the attachment figure and the availability to face the needs of attachment (for example, one might think the partner may be interested in someone else or that he/she does not desire closeness). The combination of anxiety and avoidance leads to four prototypes (Brennan et al., 1998): the secure (low levels of avoidance and low levels of anxiety), preoccupied (low levels of avoidance, but high levels of anxiety), dismissing (which is the same as the avoidant style mentioned above, with high levels of avoidance and low levels of anxiety), and lastly, the fearful style.

Each one of these lines of research has contributed to enrich the knowledge of the mechanisms, which come into play in the formation, functioning, and evolution of couple relationships (Mikulincer and Shaver, 2007).

Concerning IPV in romantic relationships, violence has been interpreted by researchers as a dysfunctional attempt to maintain proximity to the partner, that assumes the role of an attachment figure, when attachment needs are threatened (Simpson and Rholes, 1994).

According to Shaver and Mikulincer (2011), people with anxious attachment would tend to be ambivalent toward power and domination; on one hand, in fact, they would like to have control of the relationship, but on the other they may fear to obtain it, because this could provoke the resentment of the partner, and therefore constitute a threat to the stability of the relationship. People with an avoidant attachment would instead tend toward autonomy and distance, the critical vision of others, and the perception of others as objects to be used instrumentally for the satisfaction of their needs.

The joint between insecurely attached partners is peculiar: one partner may perceive a threat when the other partner claims for autonomy, as if leaving he won't ever get back again, and he gains reassurance only maintaining proximity and control over him. In reverse, the other partner may perceive partner's need for closeness and intimacy as oppressing and threatening for its autonomy. This conflicting perspectives can easily lead to a misunderstanding that often generates violence, perpetrated by one partner or both (Hazan and Shaver, 1987).

# Objectives

In the years, several reviews of different nature have been conducted to explore the relationship between the two constructs.

McClellan and Killeen (2000) produced the first narrative review exploring the use of aggression by males in couples in the light of attachment theory: the paper is focused on the evidence that adults internal working models have a consistent role in their adult relationship with intimate partners, making a parallel between infant experiences of attachment and the replications of insecure patterns in adulthood.

A review on risk and protective factors for male psychological abuse toward partners has been conducted by Schumacher et al. (2001): according to this review, adult attachment, along with other factors such as communication partners and marital adjustment, is significantly associated with psychological IPV.

A review on literature on female perpetrators of IPV has been written by Carney et al. (2007): the narrative review makes an interesting confrontation on male and female offenders and includes a summary of existing intervention programs for these women.

Finkel and Slotter (2006) have discussed a narrative review, adopting an attachment perspective to reconsider IPV as an impulsive behavior that occurs when an individual feels threatened in the relationship.

Langhinrichsen-Rohling (2010) wrote an interesting paper on controversial discussions regarding gender and IPV in US, addressing topics about subtypes of IPV, differences between male and female perpetrators and gender-related challenges concerning the phenomenon.

Ogilvie et al. (2014) wrote a meta-analysis focused on attachment and violent offending, investigating controversial results about the correlation between attachment and several typologies of criminal offending (i.e., IPV, violent offending, sexual offending, non-violent offending).

The narrative review produced by Park (2016) is a very useful dissertation about implications of attachment theory applications to IPV, focused on theory's strengths and limitation in both understanding and facing the phenomenon.

Tapp and Moore (2016) produced a very useful article on instruments to assess the risk of IPV in late adolescents and young adults. It provides a very exhaustive review on the most used measures, highlighting their characteristics and efficacy, to explore the phenomenon and detect risk potential.

In the end, Karantzas et al. (2016) provided a very complete systematic review concerning the topic of attachment style and less severe forms of sexual coercion, taking in consideration the phenomenon not only in couple setting but also related to acts perpetrated toward other people.

After exploring existing reviews discussing the relationship between IPV and attachment, we found a gap in research: no other review examined studies that explored the relationship between attachment and IPV in all its manifestations nor it adopted a systematic approach nor it considered studies conducted among male and female samples.

Therefore, this systematic review has the objective to collect and draw conclusions from all the studies available that investigated the relationship between attachment and all forms of IPV, considering researches conducted among male and female samples and not only among couples. It is crucial for both clinicians and researchers to have a clearer view of the correlations between the two constructs, with the goal to elaborate specific programs, to prevent and to intervene properly on both perpetrators and victims.

## Research Question

Two main research questions have led to the preparation of this review: Is attachment involved in IPV? How does attachment explain the process that leads to IPV?

# METHOD

A systematic search was conducted according to PRISMA guidelines (Moher et al., 2009). The full process of study identification, inclusion and exclusion is illustrated in **Figure 1**.

# Search Strategy

Computer database researches were conducted using the following databases: Psychinfo, Psycharticle, Medline, Scopus, Web of Science, and PubMed (all years to the 01/02/2018). Search terms were compiled into two concepts for all database namely Attachment and IPV. For Psycinfo, Psycharticle, Medline and Scopus, terms of Appendix A were searched in Title, Abstract and Key-words fields and results were refined including only articles. For Web of Science, terms presented in Appendix B were searched into Topic field. Then, we refined results including articles and excluding the Medline database. Finally, for PubMed, terms presented in Appendix B (including Mesh Terms) were searched into Title and Abstract fields.

# Selection of Studies

We screened every title and abstract to determine the eligibility of the study for inclusion. Criteria for inclusion of studies were the following: (1) To investigate both attachment and IPV constructs; (2) To conduct study on adults or adolescents; (3) To provide original research; (4) To use a qualitative perspective; (5) To use validated instruments for the measurement of both attachment and IPV.

Two reviewers (SBZ, GR) independently conducted the electronic searches using the aforementioned databases. Together, independent review of these electronic databases identified a total of 6,129 articles with the initial search terms, which were then examined by each reviewer for eligibility. After removing the duplicates, a total of 3,598 records was screened, resulting in the identification of 319 full-text articles to be further scrutinized. Upon closer examination, there was consensus that 113 of those studies met the study inclusion criteria.

# Data Extraction and Reporting

A coding protocol was prepared and used to extract relevant information from the selected primary studies. In particular, six classes of information were coded: (1) characteristics of the publication (i.e., year); (2) characteristics of the sample (i.e., total sample size; gender; age was coded as the mean, standard deviation in years, sample composition); (3) information about the methodological characteristics (i.e., the context of the study was coded as the country in which the research was conducted; the type of design was coded as cross-sectional or longitudinal; the instruments used to measure attachment and IPV were reported) (4) Main results (the dimensions of attachment significantly associated with IPV were reported together with the statistical index used in the study).

# RESULTS

#### IPV Victimization and Attachment

We found 47 studies examining attachment among victims of IPV. Despite the fact that the first papers on the topic were written on 1997, 72.92% of the studies have been published in the last 10 years. Importantly, 1.46% of these studies did not distinguish between different forms of IPV. In contrast, 60.42% of the papers focused on physical IPV, 45.83% examined psychological forms of violence and only 1.25% investigated sexual IPV. Results are presented within these four categories in the following sections.

#### Generic IPV Victimization and Attachment

**Table 1** displays the characteristics of the seven studies investigating attachment among victims of IPV without distinguishing between the different forms of violence. Most of researches have been conducted in USA (57.14%) and Canada (42.86%). Research is mainly cross-sectional with only two contributions adopting a longitudinal prospective (Weiss et al., 2011; La Flair et al., 2015).

Researchers generally decided to investigate the topic among samples balanced for gender with only two exceptions (Shechory, 2013; Yarkovsky and Timmons Fritz, 2014). Also, they often used participants with no previous report of IPV such as students

(Wekerle and Wolfe, 1998; Yarkovsky and Timmons Fritz, 2014; McClure and Parmenter, 2017) or participants recruited among minority populations. For example, some studies have been conducted on couples with men doing military service (Frey et al., 2011), on adolescents with intellectual disability (Weiss et al., 2011) and on healthcare female workers (La Flair et al., 2015). Noteworthy, only two studies recruited participants with a previous reported history of IPV (Shechory, 2013; Lewis et al., 2017).

Also, studies showed heterogeneity regarding age of the participants with three studies investigating the topic among adult population (Frey et al., 2011; Shechory, 2013; La Flair et al., 2015), and five studies recruiting young adults or adolescents (Wekerle and Wolfe, 1998; Weiss et al., 2011; Yarkovsky and Timmons Fritz, 2014; Lewis et al., 2017; McClure and Parmenter, 2017).

Studies were highly heterogeneous regarding the instrument used to measure IPV with the most administering the Conflict in Adolescents Dating Relationships (CADRI, Wolfe et al., 2001). In contrast, measurment of attachment appeared more consistent with the Experiences in Close Relationships (ECR; Brennan et al., 1998) and its revised version being the most used.

Results seem to support the hypothesis of a relationship between the attachment dimensions anxiety and avoidance and IPV victimization. Indeed, almost all correlational studies, with two exceptions (Frey et al., 2011; Weiss et al., 2011), found significant and positive correlation between the anxious dimension of attachment and IPV victimization. However, coefficient indicated only weak associations, ranging from 0.13 to 0.30. The fact that Weiss et al. (2011) did not replicate this result may be explained by the specificity of their sample, being constituted by adolescents with intellectual disability. Also, Frey et al. (2011) found a negative and significant correlation (albeit very weak: r = −0.03) between IPV victimization and anxious attachment in female partners of men doing military service. It has to be noted that the sample in this study is particularly small and results are consequently difficult to generalize to the whole population. Supporting the idea that victims of IPV may have high levels of anxious attachment, two studies successfully compared groups of females with a history of IPV with groups of females without previous reported victimization (Shechory, 2013; Lewis et al., 2017). Both found that females belonging to the IPV group scored higher on the anxious dimension of attachment compared to control participants.

Turning to the attachment dimension of avoidance, results are more contrasting with five studies showing associations with IPV victimization (Wekerle and Wolfe, 1998; Frey et al., 2011; Weiss et al., 2011; Shechory, 2013; La Flair et al., 2015) and others failing to replicate such results (Yarkovsky and Timmons Fritz, 2014; Lewis et al., 2017; McClure and Parmenter, 2017). As for the anxiety dimension, correlational coefficients indicate weak associations between avoidance and IPV victimization, ranging from 0.27 to 0.33.

Moreover, some of these studies brought additional contributions for the understanding of the relationship between IPV and attachment. For example, the role of individual differences has been pointed out, underlying that intellectual ability (Weiss et al., 2011) and gender (Wekerle and Wolfe, 1998; Lewis et al., 2017) may play a moderating role in such link. Also, two studies, using structural equation modeling, evidenced that attachment insecurity may mediate the relationship between IPV and psychological symptoms as Post-Traumatic Stress Disorder (PTSD) (Frey et al., 2011) or depression (La Flair et al., 2015). Importantly, a study showed that attachment insecurity predicted no longer IPV victimization after controlling for social desirability (Yarkovsky and Timmons Fritz, 2014).

#### Physical IPV Victimization and Attachment

Studies examining attachment in victims of physical IPV were 30 (all displayed in **Table 2**) with 62% conducted in USA, 24.14% in Europe, 10.34% in Canada and one in Chile. Only five studies adopted a longitudinal design of research with the others being cross-sectional.

Eight studies were conducted on women with a previous reported history of IPV whereas more than a third of the studies used students as participants. Seven studies were conducted on couples whereas 15 groups of researchers focused exclusively on female population. Interestingly, two studies examined the topic among clinical population suffering from PTSD and depression.

The Conflict Tactics Scale (CTS, Straus, 1979) and its revised version (CTS2, Straus et al., 1996) were the most used instrument for the assessment of physical IPV (65.52% of the studies). Instruments measuring attachment were more heterogeneous with the ECR being the most used (44.83% of the studies).

Noteworthy, all studies merging the avoidance and anxiety dimensions into a unique insecure one, evidenced a positive and significant association with physical victimization (Toews et al., 2005; Higginbotham et al., 2007; Karakurt et al., 2013). Interestingly, whereas these first evidences were brought by two studies focusing exclusively on females, Karakurt et al. (2013) successively found that such association was significant only among male participants.

Considering the specific dimensions of attachment, only four studies failed to find some kind of relationship with physical IPV victimization (Bookwala, 2002; Orcutt et al., 2005; Shurman and Rodriguez, 2006; Rapoza and Baker, 2008). In relation to the anxious facet of attachment, results are highly contrasting with 14 studies finding a role played by this dimension in physical IPV victimization and other 10 failing to replicate such result. Correlational studies providing support to the hypothesis of a relationship between physical IPV victimization and attachment reported coefficients ranging from 0.14 to 0.42. Also, studies conducted on clinical participants suffering from PTSD or depression are in line with these results (Owens et al., 2014; Karakoç et al., 2015).

Regarding the avoidant dimension, again, studies are split in two balanced categories of results with 14 failing to find any association with physical IPV victimization and 11 pointing out a significant relationship between the two variables. However, it has to be noted that from the last category, three studies suffer from methodological concerns related to the lack of a validated measure of IPV (Craparo et al., 2014; Hellemans et al., 2015; Karakoç et al., 2015). Anyway, the intensity of the significant associations reported vary from 0.10 to 0.53 with one study not reporting any statistical index (Rogers et al., 2005). Interestingly, the strongest correlation was obtained among a sample of men being in treatment for IPV perpetration (Bélanger et al., 2015).

Some studies did not limit the investigation to the relationship between attachment and IPV victimization but add other insightful considerations. For example, several studies tested this relationship considering the role of early trauma. First, results brought by Sandberg et al. (2016) underlined that anxious attachment significantly predicted physical IPV victimization also after controlling the role of trauma. In contrast, Karakoç et al. (2015) showed that, when controlling for the effect of trauma, insecure attachment no longer predicted IPV victimization among patients suffering from depression. Also, Smith and Stover (2016) found that childhood maltreatment predicted IPV victimization only when participants scored high on the anxious dimension of attachment. However, Gay et al. (2013) showed that insecure attachment did not mediate the relationship between childhood maltreatment and IPV victimization. No other studies



Frontiers in Psychology | www.frontiersin.org


Frontiers in Psychology | www.frontiersin.org


tested such mediation model. Then, insecure attachment has been showed to moderate the link between IPV victimization and PTSD symptoms (Scott and Babcock, 2009) and to mediate the relationship between physical victimization and depressive symptomatology (Smagur et al., 2018). Furthermore, others variables seem to play a role in the relationship between attachment and IPV victimization as conflict resolution strategies (Bonache et al., 2017), anger (Kuijpers et al., 2012) and religiosity (Higginbotham et al., 2007). Finally, gender differences emerged in the study of Hellemans et al. (2015) suggested that physical IPV victimization was related with attachment avoidance among women and with attachment anxiety among men.

#### Psychological IPV Victimization and Attachment

Since 1997, 23 studies investigating attachment among victims of psychological IPV have been published. Characteristics of these studies are illustrated in **Table 3**.

Among them, 54.54% were conducted in USA, 22.72% in Europe and 18.18% in Canada. Despite the fact that the very two first studies were published in 1997, 81.81% of them have been published in the last 10 years. Only 18.18% of the studies were longitudinal in their design with the others being crosssectional. Regarding sample types, most of the researches were conducted on couples (Péloquin et al., 2011; Karakurt et al., 2013; Oka et al., 2014, 2016; Seiffge-Krenke and Burk, 2015; Goncy and van Dulmen, 2016; Tougas et al., 2016; Sommer et al., 2017). Fortunately, only five groups of researchers used student samples (O'Hearn and Davis, 1997; Wigman et al., 2008; Riggs and Kaminski, 2010; Bonache et al., 2016, 2017). Two additional studies were conducted among male-only samples being veterans suffering from PTSD (Owens et al., 2014) or batterers (Bélanger et al., 2015). Unfortunately, only a small proportion of studies recruited women reporting experiences of psychological IPV (Henderson et al., 1997; Shurman and Rodriguez, 2006; Kuijpers et al., 2012; Smagur et al., 2018). Finally, two studies examined the topic among minority populations of women (Weston, 2008; Hellemans et al., 2015). Instruments used to evaluate both IPV and attachment were homogenous with ECR being mostly used to evaluate attachment styles and the subscale of CTS used to measure the intensity of psychological IPV.

Three studies, merging the anxiety and avoidance dimensions in a unique index of insecure attachment, found that psychological IPV victimization was positively correlated with insecure attachment (Toews et al., 2005; Karakurt et al., 2013; Oka et al., 2016) with coefficient ranging from 0.31 to 0.53. Noteworthy, almost half of the studies found that psychological IPV victimization was not associated with anxious or avoidant dimensions of attachment (Henderson et al., 1997; Shurman and Rodriguez, 2006; Wigman et al., 2008; Oka et al., 2014; Bélanger et al., 2015; Tougas et al., 2016).

Regarding the anxious dimension, studies conducted on women with reported history of IPV mainly failed to find an association between anxious attachment and psychological victimization (Henderson et al., 1997; Shurman and Rodriguez, 2006; Kuijpers et al., 2012). In contrast, studies recruiting students or community participants mostly indicated a relationship between anxious attachment and IPV among victims, suggesting a potential role played by sample type (O'Hearn and Davis, 1997; Henderson et al., 2005; Riggs and Kaminski, 2010; Bonache et al., 2016, 2017). Noteworthy, such studies greatly vary in the intensity of reported association with correlational coefficients ranging from 0.15 to 0.58. Finally, whereas some studies conducted on couples reported association between psychological IPV and anxious attachment among victims (Péloquin et al., 2011; Seiffge-Krenke and Burk, 2015; Goncy and van Dulmen, 2016; Sommer et al., 2017), two others studies failed to replicate such results (Oka et al., 2014; Tougas et al., 2016). However, the study of Oka et al. (2014) may be biased by methodological issues as IPV was measured throughout only three items extracted from the Conflict Tactics Scale-Revised.

Then, from 20 studies examining the relationship between avoidant attachment and psychological IPV, only 11 found a significant association between the constructs with correlational coefficients ranging from 0.20 to 0.50. In relation to the role played by gender in such relationship, Péloquin et al. (2011) found that this association was significant only among females whereas two other studies indicated significant correlations in both gender (Seiffge-Krenke and Burk, 2015; Goncy and van Dulmen, 2016; Sommer et al., 2017).

Some studies shed light on additional interesting aspects related to the link between attachment and psychological IPV victimization. First, gender differences emerged in some studies (Péloquin et al., 2011; Hellemans et al., 2015). Also, studies showed that insecure attachment not only predicted psychological IPV victimization beyond the role of depression (Riggs and Kaminski, 2010) but also mediated such relationship (Smagur et al., 2018). Finally, the use of destructive conflict strategies has been showed to explain the pathway by which insecure attachment leads to psychological IPV victimization (Bonache et al., 2016, 2017).

#### Sexual IPV Victimization and Attachment

Only six studies investigated the relationship between sexual IPV and attachment among victims (see **Table 4**). Interestingly, these studies are relatively recent with the majority having been published in the last 3 years. All of them, except one (Bonache et al., 2016), were conducted in USA (Weston, 2008; Karakurt et al., 2013; Ross et al., 2016; Sommer et al., 2017; Smagur et al., 2018). Half of the studies were cross-sectional in their design with the remainders being longitudinal. Studies recruited large samples ranging from 51 to 574 participants by group (Weston, 2008; Bonache et al., 2016). Noteworthy, only one research was conducted on participants with a reported history of IPV (Smagur et al., 2018) with most of others recruiting couples extracted from general population (Karakurt et al., 2013; Sommer et al., 2017) or undergraduate students (Bonache et al., 2016; Ross et al., 2016).

Results regarding the relationship between the anxious dimension of attachment among victims of sexual IPV are contrasting. Some found that sexual IPV victimization correlated positively and significantly with anxiety (Sommer et al., 2017; Smagur et al., 2018) with coefficient reaching 0.53. Also, results of Ross et al. (2016) indicated that individuals with history of sexual



July 2018 | Volume 9 | Article 1166

**19**

Frontiers in Psychology | www.frontiersin.org



IPV scored higher on the anxious dimension of the Experiences in Close Relationships-Short form (ECR-S, Wei et al., 2007) compared to participants without experiences of sexual IPV. In contrast, two studies failed to find significant association between sexual IPV victimization and anxious attachment (Weston, 2008; Karakurt et al., 2013).

In relation the attachment dimension of avoidance, data brought by correlational studies mostly indicated a positive association between anxious attachment and sexual IPV with coefficients ranging from 0.22 to 0.39. Noteworthy, one of them found that the association was significant only among men. However, two studies did not go in the same direction, finding no association between avoidance and sexual IPV victimization (Karakurt et al., 2013) or no differences on avoidance scores between individuals with and without sexual IPV victimization (Ross et al., 2016).

Finally, two studies examined other variables accounting for the relationship between attachment and sexual IPV victimization showing that such link was mediated by the use of destructive conflict resolution strategies (Bonache et al., 2016) and that insecure attachment fully mediated the pathway by which childhood maltreatment leads to sexual IPV victimization (Smagur et al., 2018).

As a whole, research examining the role of attachment in IPV victimization appears widely unbalanced in relation to the type of violence investigated, with most studies measuring physical manifestation and only few including a separate measurement of sexual victimization. Despite the fact that the majority of studies found some kind of association between insecure attachment and IPV victimization, results are highly contrasting regarding the specific dimensions of attachment.

# IPV Perpetration and Attachment

In the present review, we found 72 studies that explored the attachment dimensions among IPV perpetrators. Contrary to the studies on IPV victimization, most of the studies focused on psychological IPV (40.74%), whereas 15.52% of the studies did not make differences between the different forms of IPV, 31.04% of the studies investigated physical IPV and on 6.79% of the studies were focused on sexual IPV. The studies we examined cover a wide range of years, comprised between 1994 and 2017, even though the majority has been published in the last 10 years.

#### Generic IPV Perpetration and Attachment

Over the years, 15 studies decided to investigate the relationship between the perpetration of violence in general, not discriminating between different forms of expression, and attachment. These studies are displayed in **Table 5**.

America and Europe have been the continents in which studies were mainly conducted: most of the studies were run in USA (33.3%), followed by Spain (13.3%), Chile (13.3%), and Belgium (13.3%). There was only one studied conducted in France, one in Canada and one in UK.

Research are mainly cross-sectional in their design with only one study adopting a longitudinal design of research (Ulloa et al., 2014).

Only six studies (Babcock et al., 2000; Carraud et al., 2008; Gay et al., 2013; Genest and Mathieu, 2014; Muñoz, 2015; Pimentel and Santelices, 2017) did not investigate the topic among samples balanced for gender.

Regarding sample types, researchers often used participants with no previous report of IPV such as students (Wigman et al., 2008; Grych and Kinsfogel, 2010; Gay et al., 2013; Tassy and Winstead, 2014; Ulloa et al., 2014; Aizpitarte et al., 2017; Gonzalez-Mendez et al., 2017; McClure and Parmenter, 2017) or minority population such as divorced couples (De Smet et al., 2012, 2013) and jail population (Carraud et al., 2008). In this case, only four studies recruited participants with a previous reported history of IPV (Babcock et al., 2000; Genest and Mathieu, 2014; Muñoz, 2015; Pimentel and Santelices, 2017).

Surprisingly, most studies investigated the topic among young adult population and adolescents and only six studies recruited adult population (Babcock et al., 2000; De Smet et al., 2012, 2013; Genest and Mathieu, 2014; Muñoz, 2015; Pimentel and Santelices, 2017).

Due to the age variability of the samples, there was a relevant heterogeneity concerning the instruments used to measure IPV: CTS, both in its revised and its short version (Control Tactics Scale-Short form; CTS-S), has been the most used in adult samples, whereas CADRI was the most used with adolescents. As for attachment measures, ECR, both in its revised (Experiences in Close Relationships-Revised; ECR-R, Fraley et al., 2000) and its short version, turns out to be the most used tool (62.5%) both for adult and for adolescent population. Also, an American study conducted in 2000 used Adult Attachment Interview (AAI; George et al., 1985) on batterers.

Results of the studies supported the hypothesis of a relationship between attachment dimensions and being a perpetrator of IPV. All the studies that confronted a clinical group of violent men with a control group of non-violent men (Babcock et al., 2000; Carraud et al., 2008; Muñoz, 2015; Pimentel and Santelices, 2017) proved that violent men tend to have insecure attachment (Babcock et al., 2000), showing a higher level of anxiety in close relationships (Muñoz, 2015; Pimentel and Santelices, 2017) compared to non-violent men, even though results do not agree with each other about the prevailing attachment style of clinical groups. There has been found a prevalence of preoccupied and dismissing attachment (Carraud et al., 2008) over other attachment styles.

Most of the studies support the existence of a positive correlation between the attachment dimension of anxiety and IPV perpetration, even though coefficient did not indicate any strong association. They ranged from 0.06 to 0.33.

Both the weak correlation of 0.06 and a study that did not obtain any significant outcome (McClure and Parmenter, 2017) may be explained by the use of the Adult Attachment Scale (AAS, Hazan and Shaver, 1987), which might be not enough sensitive as a tool for this specific target group, as claimed by Tasso et al. (2012).

Concerning the attachment dimension of avoidance, only two studies found a significant correlation between IPV perpetration and avoidance (Weiss et al., 2011; Aizpitarte et al., 2017), where other studies failed to replicate the same result.



Correlational coefficients range from 0.30 to 0.13: as for anxiety dimension scores indicate a weak association between the two constructs.

#### Physical IPV Perpetration and Attachment

Since 1998, the relationship between attachment and physical IPV as perpetrators has been investigated in 32 studies, illustrated in **Table 6**.

Most of the studies have been conducted in America: in USA (62.5%) and Canada (21.87%). Only three studies have been conducted in Europe, two in UK and one in Germany. Other two studies have been conducted respectively in Chile and in Australia.

In this group of studies, the most common design is crosssectional with only one study adopting a longitudinal design, performed in USA (Lawson and Brossart, 2009).

Among all the studies, 23 of them investigated the relationship between the two dimensions among samples balanced for gender. Only four studies had female-only samples and 11 had only male samples.

Contrary to what one might thing, only ten studies (Rankin et al., 2000; Kim and Zane, 2004; Lawson et al., 2006; Goldenson et al., 2007; Mauricio et al., 2007; Lawson, 2008; Lawson and Brossart, 2009; Brown et al., 2010; Lawson and Malnar, 2011; Bélanger et al., 2015) recruited participants that had a previous history of IPV or that are convicted or in therapy because of it. Most of the samples are made up of community population: a relevant number of researches enrolled high school or college students whereas others recruited veterans in treatment for PTSD (Owens et al., 2014), men in treatment for relationship issues (Fournier et al., 2011; Brassard et al., 2014), divorced mothers (Toews et al., 2005), and female prisoners (McKeown, 2014).

All studies, except two (Wekerle and Wolfe, 1998; Burk and Seiffge-Krenke, 2015), enrolled adults or young adults in their samples, so the age of participants is quite homogeneous.

Due to this homogeneity, we can observe quite an accordance in the choice of the physical IPV measure: most of the studies used CTS, both in its revised and its short version, whereas three studies used CADRI to assess adolescents. Concerning attachment assessment, there is much more heterogeneity with most studies making use of ECR, both in its revised and its short version.

The hypothesis of a relationship between physical IPV perpetration and attachment dimensions has been supported by the results of the studies. Compared to non-violent samples, physical IPV perpetrators show higher level of anxiety and avoidance (Goldenson et al., 2007) and a preoccupied attachment style (Henderson et al., 2005). Most of the studies supported a positive correlation between physical IPV perpetration and the attachment dimension of anxiety, even though coefficient ranged from 0.56 to 0.12, so they're not really strong.

Concerning the attachment dimension of avoidance, several studies found a positive correlation with physical IPV perpetration. Correlational coefficients range from 0.30 to 0.12, so, as for attachment anxiety, they indicate a weak correlation between the constructs.

Concerning the dimension of closeness, one study found a negative correlation (r = −0.32) with physical IPV (Lawson, 2008).

Regarding gender differences, they are consistent with the trend, showing a prevalence of anxiety and avoidance in both male and female physical IPV perpetrators (González et al., 2016; Sommer et al., 2017).

Noteworthy, eight studies didn't obtain any significant outcome. However, two of them used the AAS as attachment measure, which is claimed to be not much sensitive for such samples by Tasso et al. (2012).

#### Psychological IPV and Attachment

There are 42 studies that investigated the relationship between attachment and psychological IPV, focusing on IPV perpetration (see **Table 7**).

Most of the studies have been conducted in America (64.28% in USA and 26.19% in Canada), with only two studies conducted in UK, one in Germany and one in Australia.

Studies have predominantly a cross-sectional design; only three studies present a longitudinal design (Lawson and Brossart, 2009; Wright, 2015; Gou and Woodin, 2017).

Surprisingly, only 22 studies investigated the topic among samples balanced for gender; instead, 16 studies had an exclusively male sample and 4 studies had an exclusively female sample.

Concerning sample nature, a relevant number of researches enrolled participants from community samples: most of them were conducted among high school and college students, couples and veterans in treatment for PTSD. Only 11 studies were focused on subjects with a history of IPV (Dutton et al., 1994, 1996; Dutton, 1995; Rankin et al., 2000; Mahalik et al., 2005; Mauricio et al., 2007; Lawson, 2008; Lawson and Brossart, 2009; Brown et al., 2010; Lawson and Malnar, 2011; McKeown, 2014).

Most of the studies were conducted among adult and young adult population, whereas only four studies investigated the topic among adolescents. According to the age of the samples, there was a consistent homogeneity over the instruments used to measure attachment: the most used instrument in adult sample was the ECR, both in its revised and its short version, although several studies used other measures.

Instead, concerning instruments to measure IPV, there was a remarkable variability presumably imputable to construct complexity, as psychological violence comprises very different forms of violence from verbal abuse to cyber aggression. It follows that the most used tool has been CTS, both in its revised and its short version, because of its ability to detect different types of psychological violence, less or more severe. Another common measure is the Psychological Maltreatment of Women Inventory (PMWI, Tolman, 1999), made up of several abuse typologies subscales. Other studies measured specific forms of violence with relevant instruments: as Controlling Behavior Index (BCI, Dobash et al., 1996), Dominance Scale (DS, Hamby, 1996), Intimate Justice Scale (IJS, Jory, 2004) and Mate Retention Inventory (MRI, Buss et al., 2008). In a longitudinal study conducted among high school students



Frontiers in Psychology | www.frontiersin.org


Frontiers in Psychology | www.frontiersin.org

July 2018 | Volume 9 | Article 1166


in Adolescent Dating Relationships

 Inventory; LEQ, Love Experiences

 Questionnaire;

 M, Male.


(Continued)

**30**


Frontiers in Psychology | www.frontiersin.org



Frontiers in Psychology | www.frontiersin.org


Scale-Revised; DS, Dominance Scale; NVAWS, National Violence Against Women Survey modified to assess perpetration; ABI, Abusive behavior inventory; SSDS, Spouse-specific dependency scale; PMI, gender neutral Psychological Maltreatment of Women Inventory; IJS, Intimate Justice Scale; MIMARA, Multi-Item Measure of Adult Romantic Attachment; MCTS, Modified Conflict Tactics Scale; ECR-S, Experiences in Close Relationships Scale-Shortform; CADRI, Conflict in Adolescence Rating Scale; LEQ, Love Experiences Questionnaire; MRI-S: Mate Retention Inventory-Short Form.

of

 of

(Wright, 2015) there have been developed two specific selfreport questionnaires to measure both attachment and cyber aggression.

Even though studies present a relevant heterogeneity, mainly for kind of violence and for sample composition, they supported the hypothesis of a relationship between insecure attachment and psychological IPV perpetration.

Four studies that confronted clinical groups of IPV perpetrators with control groups of non-perpetrators showed that attachment anxiety or fearful attachment make the difference between the two groups (Dutton et al., 1994, 1996; Dutton, 1995) and that IPV has a negative correlation (r = −0.30) with the attachment dimension of closeness (Lawson, 2008).

Concerning gender difference, there are contrasting results: even though most studies state that both male and female perpetrators tend to present attachment anxiety, some studies have found a prevalence of avoidant attachment in male samples compared to female (Lafontaine and Lussier, 2005; Gormley and Lopez, 2010; Frey et al., 2011; McKeown, 2014; Sommer et al., 2017).

The majority of studies support a positive correlation between psychological violence perpetration and the attachment dimension of anxiety, albeit the association indicated by coefficients ranged from 0.66 to 0.16. Regarding the association between avoidance and psychological IPV, several studies support the positive correlation between the two constructs, but the coefficient tend to be weak also in this case, ranging from 0.65 to 0.15.

#### Sexual IPV and Attachment

As shown in **Table 8**, only seven studies investigated the relationship between sexual IPV and attachment from the perspective of the perpetrator.

The majority of these studies have been conducted in USA (71.42%), one study has been run in Australia and another one in China. No studies have yet been conducted in Europe. All the studies present a cross-sectional design.

Although sexual partner violence is usually considered a male preserve, most researches explored the construct among samples balanced for gender and only three studies present a male-only sample (Kalichman et al., 1994; Rankin et al., 2000; Smallbone and Dadds, 2001).

Regarding sample composition, most of the studies have been conducted among community population samples, like college students (Kalichman et al., 1994; Smallbone and Dadds, 2001; Ménard et al., 2010; Karakurt et al., 2013; He and Tsang, 2014) and community couples (Sommer et al., 2017). Surprisingly, only one study has been focused on male subjects convicted for IPV (Rankin et al., 2000).

All the studies were conducted among young adult and adult population, so all the samples are relatively homogeneous for age; due to the peculiarity of the construct, it has not been investigated among adolescents.

Even though the populations of different samples are not homogeneous, there is a consistent heterogeneity in the choice of instruments, both for IPV and for attachment measures. Concerning attachment measures, ECR, both in its standard and its short version and Attachment Style Questionnaire (ASQ, Feeney and Noller, 2001) are the most common instruments. There is much more variability regarding the measures that assess sexual IPV: only three studies made use of CTS, in its original and in its revised version, which is the most widespread instrument to assess different forms of IPV. Some studies employed very specific measures that consider sexual experiences more broadly, among which we find sexual coercion and sexual abuse perpetration.

All studies, except two that did not have any significant outcome, support the hypothesis of a positive correlation between attachment dimensions and sexual IPV perpetration. An equal number of studies support the correlation between sexual IPV and the attachment dimension of anxiety and the correlation between sexual IPV and the attachment dimension of avoidance. Concerning the dimension of anxiety, several studies supported the correlation with sexual IPV (Smallbone and Dadds, 2001; Ménard et al., 2010; He and Tsang, 2014; Sommer et al., 2017). The coefficients presented ranged from 0.32 to 0.16, that indicate a weak correlation. The positive correlation between the attachment dimension of avoidance and sexual IPV is supported by several studies, but the coefficient is not strong, as for anxiety, ranging from 0.29 to 0.22.

Regarding gender difference, one study (Sommer et al., 2017) confronted male and female groups and it resulted that both female and male sample presented a significant rate of attachment anxiety but only male results displayed a correlation between sexual IPV and avoidance.

#### New Lines of Research

In the studies we examined, there were several studies that studied the construct of IPV and its relationship with attachment in peculiar groups or from an unconventional perspective, which could represent interesting future trends in research.

Some studies investigated IPV in homosexual individuals and couples, to see if they are consistent with results obtained in heterosexual couples. Other studies were focused on couples and mutual violence, which are a privileged point of view to investigate how the phenomenon of IPV arise in dyads, considering couple attachment. In the end, there is a specific current of studies that explored the chance to categorize IPV perpetrators in specific groups, to develop specific approaches to treatment that take into account the peculiar characteristics of the specific groups.

#### IPV in Same-Sex Couples

As **Table 9** illustrates, five studies (Landolt and Dutton, 1997; Stanley et al., 2006; Bartholomew et al., 2008; Craft et al., 2008; Gabbay and Lafontaine, 2017), in a time lapse that runs from 1997 to 2017, investigated the relationships between different forms of IPV among same-sex couples and homosexual individuals. These studies have been conducted only in Canada (60%) and USA (40%), maybe due to the major cultural acceptance of homosexuality in these Countries. All the studies, except one, investigated the correlations between the two constructs in community samples, mainly


recruited by advertising on newspaper and on the internet. All the studies have been conceived with a longitudinal design.

Concerning the instruments, the majority of studies made use of CTS, both in its original and in its revised version; one study used a specific interview about history of IPV experiences and another one used both CTS and Psychological Maltreatment Inventory (PMI) to assess both physical and psychological violence. Regarding attachment, the most used instrument is the Relationship Scale Questionnaire (RSQ, Bartholomew and Horowitz, 1991).

Concerning the composition of samples, only two studies assessed samples with a gender balanced composition; two studies were conducted among male populations; one study, instead, enrolled male couples.

All the studies confirm the positive correlation among homosexual population between different typologies of IPV and attachment, both for perpetrators and victims, in accordance with the findings obtained in researches that investigated the subject in heterosexual population.

Among all the studies, only two investigated the constructs from both the perspectives; the others were focused on perpetrators.

According to Bartholomew et al. (2008), IPV victims in same sex relationships show a negative correlation between avoidance and both psychological (−0.17) and physical (−0.16) IPV, whereas there was a positive correlation between anxiety and physical IPV (r = 0.17). For what concerns IPV perpetrators in same sex relationships, the study claimed a positive correlation between anxiety and both psychological (r = 0.18) and physical (r = 0.19) IPV and a negative correlation between avoidance and physical IPV (r = −0.22).

Concerning studies that assessed the construct only among perpetrators, there is a general accordance about the positive correlation between insecure attachment and generic IPV perpetration (Craft et al., 2008). Concerning psychological IPV (Landolt and Dutton, 1997; Bartholomew et al., 2008), perpetration turned out to be positively correlated to the attachment dimension of anxiety (r = 0.18) and to fearful (r = 0.40) and preoccupied (r = 0.26) and negatively correlated to secure attachment styles (r = −0.37). For physical IPV (Landolt and Dutton, 1997; Bartholomew et al., 2008), correlations are different: perpetration of this form of violence is positively correlated with fearful attachment style (r = 0.34) and with the attachment dimensions of anxiety (r = 0.19), but it is negatively correlated with avoidance (r = −0.22). Only one study investigated the relationship between sexual IPV perpetration and attachment and both the attachment dimensions of anxiety and avoidance are positively correlated with the construct.

The study that assessed the relationship between IPV and attachment using History of Attachments Interview (Stanley et al., 2006) deserve a particular mention: due to the peculiar structure of the interview, that explores experiences concerning significant attachment episodes along lifespan, several significant themes recurred in people narratives. Unmet or threatened emotion needs, such as need for closeness or desire for commitment and monogamy or loss of relationships, recurred as consistent themes in participant's stories. So the authors agreed that they can be interpreted as attachment wounds linked to dimensions of anxiety and avoidance.

#### IPV and Attachment in Couple Contexts

To date, 15 studies investigated the relationship between IPV and attachment in couples, taking in consideration both partners attachment (see **Table 10**).

Most of the studies were run in North America, in USA (66.6%) and Canada (33.3%), and only one study was conducted in Europe (Germany). All the studies, except two (Bookwala and Zdaniuk, 1998; Miga et al., 2010), were conducted among couple samples and only one study investigated the construct among a sample not balanced for gender (Miga et al., 2010). Concerning the design of the studies, all adopted a cross-sectional design except one (Miga et al., 2010).

There was an interesting variability concerning sample composition: some studies investigated the construct among community samples (Bond and Bond, 2004; Doumas et al., 2008; Rapoza and Baker, 2008; Péloquin et al., 2011; Wilson et al., 2013), some among college students (Bookwala and Zdaniuk, 1998; Bookwala, 2002; Rogers et al., 2005; Goncy and van Dulmen, 2016), three studies among adolescents (Miga et al., 2010; Seiffge-Krenke and Burk, 2015; Lewis et al., 2017) and one study among veterans in treatment for PTSD and their spouses (Frey et al., 2011).

Surprisingly, only two studies regarded couples presenting previous or current IPV (Allison et al., 2008; Bélanger et al., 2015).

Due to the heterogeneity of the samples, there was a remarkable variability of instruments used. Concerning attachment, the most common instrument was ECR, also in its revised and in its short version, but other instruments have been used. As for attachment, a wide range of instruments has been used also to assess IPV: CTS, also in its revised version, has been used in the majority of researches.

Even though the studies adopted different instruments to measure the construct and different methods to test the hypothesis, all of them agree on the positive correlation between IPV and insecure attachment of both partners. According to studies that investigated the relationship between partners attachment and IPV, not making distinctions between different forms of violence, mutual violence seems to occur when there is at least on partner with preoccupied attachment (Bookwala and Zdaniuk, 1998; Bookwala, 2002), in the dismissing-anxious couple pattern (Bond and Bond, 2004), when both partners present anxious attachment (Allison et al., 2008), when they show high rates on both attachment dimensions of anxiety and avoidance (Rogers et al., 2005).

The dismissing-anxious pattern seems to be the most common configuration in couples with one-sided IPV (Bookwala, 2002; Frey et al., 2011; Wilson et al., 2013). This pattern of interaction has been referred to as a pursue-withdraw cycle (Johnson, 2004). But several studies claim that one sided IPV couples have at least one partner showing high scores on the attachment dimension of anxiety (Lewis et al., 2017).

TABLE 9 | Studies investigating the relationship between generic IPV and attachment among homosexuals.


NA, Not available; PMI, Psychological Maltreatment Inventory; CTS, Conflict Tatics Scale; RSQ, Relationship Style Questionnaire; IPV, Intimate partner Violence; HAI, History of Attachments Interview; CTS2, Conflict Tactics Scale-Revised; ECR, Experiences in close relationship.

Concerning physical IPV and partner's attachment, the dismissing-anxious couple pattern seems to foster physical violence in the couple (Rogers et al., 2005; Doumas et al., 2008), even in adolescents (Miga et al., 2010). One study, conducted by Bélanger et al. (2015), reported a positive correlation between physical IPV victimization and avoidant attachment in males.

Other studies claimed that partner presenting high rates of anxiety tend to foster physical IPV perpetration by the other partner (Doumas et al., 2008; Rapoza and Baker, 2008; Seiffge-Krenke and Burk, 2015) or even by themselves (Péloquin et al., 2011).

Concerning the relationship between psychological IPV and attachment in couples, some studies claim that high levels of attachment anxiety in one of the members of the couple foster IPV perpetration by the other (Péloquin et al., 2011; Goncy and van Dulmen, 2016), while other studies assert that high levels of both anxiety and avoidance by one of the partners are positively related to psychological IPV (Seiffge-Krenke and Burk, 2015).

Two interesting studies that tried to explain the complex mechanism of mutual IPV in couples have been conducted in Netherland (Kuijpers et al., 2012) and in USA (Smith and Stover, 2016). These longitudinal studies accurately investigated the phenomenon of IPV revictimization and use of violence by IPV victims among samples of IPV victims although they got contrasting findings. In accordance with Kuijpers et al. (2012) avoidant attachment is a significant predictor for revictimization of both psychological and physical IPV; conversely, Smith and Stover (2016) found a positive correlation with anxious attachment in victims and IPV revictimization and use of violence.

#### IPV Profiles and Attachment Among Perpetrators

Between 1998 and 2014, eight studies investigated the relationship between IPV perpetration and attachment, identifying different typologies of perpetrators. Characteristics of these studies are showed in **Table 11**. The studies were conducted in USA (75%) and in Netherlands (25%) and they all adopted a cross-sectional study design.

All the studies enrolled male participants. Seven of them were conducted among clinical populations of people in treatment for IPV, only two studies investigated the construct among a community sample, recruited by advertising (Holtzworth-Munroe et al., 2000; Waltz et al., 2000).

Concerning instruments, there was a surprising homogeneity about the instruments used to assess violence: all the researches make use of CTS, both in its original and in its revised version. To evaluate the construct of attachment, several instruments have been chosen: the most common instruments is ECR, both in its original and in its revised version (ECR-R).

Concerning classifications of perpetrators made by the studies, there are several interesting differences. Three studies distinguished perpetrators according to attachment style (Mauricio and Gormley, 2001; Buck et al., 2012, 2014); four other studies according to violence level or typology (Holtzworth-Munroe et al., 2000; Waltz et al., 2000; Chiffriller and Hennessy, 2009; Mauricio and Lopez, 2009); one study identified specific categories of perpetrators, based on several characteristics (Tweed and Button, 1998).

Considering classifications based on attachment styles, Mauricio and Gormley (2001) spotted two categories among violent men in treatment for IPV: Insecurely attached and TABLE 10 | Studies investigating the relationship between IPV and attachment in couple context.


(Continued)

#### TABLE 10 | Continued


IP, Intimate Partner Violence; NA, Not available; CTS, Control Tactics Scale; RSQ, Relationship Style Questionnaire PAS-P, Partner Abuse Scale-Physical; PAPS, Physical Abuse of Partner Scale; MSI-R, Marital Satisfaction Inventory-Revised; RQ, Relationship Questionnaire; ECR, Experiences in Close Relationship; AAQ, Adult Attachment Questionnaire; HAI, History of Attachments Interview; CTS2, Conflict Tactics Scale; CIR, Conflict in Relationships; IJS, Intimate Justice Scale; MIMARA, Multi-Item Measure of Adult Romantic Attachment; ECR-S, Experiences in Close Relationships Scale-Short form; CADRI, Conflict in Adolescence Rating Scale.

Securely attached. At a primary analysis, men with secure attachment showed a higher level of social desirability and a lower need for dominance, but apparently the same violence level. Controlling for social desirability, men reporting insecure attachment showed significantly higher scores on CTS than men reporting secure attachment.

According to the study conducted by Buck et al. (2012), who also divided perpetrators in securely and insecurely attached, men that reported attachment insecurity showed higher separation anxiety, higher distrust in partner, higher dependency, lower self-esteem and more impulsivity. They hypothesized that distrust and separation anxiety might explain insecurely attached men proneness to commit IPV.

The following study conducted by Buck et al. (2014) used a different measure to assess attachment, ECR and distinguished even controls in two group according to attachment security. It resulted that securely attached perpetrators do not differ, concerning attachment scores, from securely attached control group; also insecurely attached batterers and insecurely attached controls presented the same attachment scores. So attachment seems to act as a mediator between personality disorder traits and committing violence toward the partner.

Concerning groups distinguished according to violence, Holtzworth-Munroe and Stuart, in 1994, hypothesized a classification based on severity of marital violence, generality of violence and psychopathology or personality disorders. After analyzing previous literature on the subject, they grouped IPV perpetrators in three categories: Family-only batterers (less severe marital violence, lowest level of general violence, lowest psychopathology), Borderline-dysphoric batterers (moderatehigh severity of marital violence, moderate level of general violence, moderate-high psychopathology) and Generally violent-antisocial batterers (very severe marital violence, highest level of general violence, highest psychopathology). Concerning romantic attachment, according to their findings, Family-only batterers tend to have a prevalence of secure or preoccupied attachment, Borderline-dysphoric batterers a prevalence of preoccupied attachment, whereas Generally violent-antisocial batterers a prevalence of Dismissing attachment.

The same team of researchers (Holtzworth-Munroe et al., 2000), conducted a research to test their categorization several years later and added a fourth category: Low level antisocial that presented, compared to Family-only category, more severity in marital violence, a higher level of general violence and same psychopathology levels. The results are consistent with the hypothesis in the previous study (Holtzworth-Munroe and Stuart, 1994) concerning attachment, even though severity of both couple and general violence and level of psychopathology are lower, having been tested on a community sample.

Along with these findings, Waltz et al. (2000) identified three groups, very similar to the classification conceived by Holtzworth-Munroe and Stuart (1994): Generally violent men, TABLE 11 | Studies investigating the relationship between IPV profiles and attachment among perpetrators.


IPV, Intimate Partner Violence; CTS, Conflict Tactics Scale; RSQ, Relationship Style Questionnaire; NA, Not available; AAS, Adult Attachment Style; RQ, Relationship Questionnaire; CTS2, Conflict Tactics Scale - Revised; ECR-R, Experiences in Close Relationship - Revised; ECR, Experiences in Close Relationship.

Pathological violent men and Family only violent men. Men belonging to the first group showed higher scores on attachment dimension of avoidance and lower scores on anxiety, while Pathological violent men showed lower attachment avoidance scores and higher scores on anxiety. Family only batterers, instead, showed attachment anxiety and avoidance scores similar to control group.

In the research conducted by Mauricio and Lopez (2009), the sample has been divided into three categories based on violence level: Low level violence, Moderate level violence, High level violence. It turned out that men belonging to the Low level violence group reported more acts of physical violence (even though it was about non-severe acts) than other groups, even though they presented attachment rates similar to community samples. Men belonging to the Moderate level violence group reported higher scores on the attachment dimension of anxiety; while men belonging to the High level violence group reported high levels on both the dimensions of anxiety and avoidance.

In the study conducted by Chiffriller and Hennessy (2009), the sample of IPV perpetrators was divided in five categories, three of which have been considered in the research conducted by Waltz et al. (2000). Chiffriller and Hennessy (2009) identified two more categories, new in literature, to fit Sexually violent and Psychologically violent perpetrators. Concerning attachment, men belonging to the Pathological violent category resulted to be more preoccupied (low avoidance and high anxiety) and more fearful (high avoidance and high anxiety) than men belonging to other groups. So results are partially concurring with those found by Waltz et al. (2000).

Tweed and Button (1998) made a peculiar classification of IPV perpetrators identifying two categories: Instrumental batterers and Impulsive batterers, based on participant's personality characteristics. So one group included generally violent/antisocial men and the other was made up of men that presented dysphoric/borderline cluster characteristics. Instrumental batterers got higher scores on preoccupied attachment, while Impulsive batterers presented a less secure and more fearful attachment.

# DISCUSSION

The aim of this paper was to review empirical studies investigating the relationship between attachment and IPV and going through two main research questions: Is attachment involved in IPV? How attachment may explain the processes leading to IPV?

# Main Findings

#### Characteristics of the Studies

Studies often adopted a cross-sectional design of research and were more rarely longitudinal. The main and common hypothesis underlying research was the conceptualization of IPV as an outcome determined by the quality of attachment. This hypothesis makes sense when referring to IWM as developing within the very early interpersonal context of an individual and maintaining a stability across life-span (Bowlby, 1988). In this perspective, the investigation of IWM may be recommended for studies examining variables related to childhood maltreatment or past life trauma. However, few studies used AAI to evaluate the quality of attachment in relation to IPV with the most of the them measuring romantic attachment related to a current relationship. Despite the fact that romantic adult attachment styles are thought to mainly depend on early internal working models, it is not the case for all individuals (Fraley, 2002). This approach seems reasonable when the investigators are interested in dyadic variables (e.g., marital satisfaction or conflict levels). But it could be also argued that the quality of romantic attachment is determined by IPV, reversing the initial hypothesis. Indeed, this has been successfully tested in one of the study reviewed in our paper. As a whole, future studies investigating the relationships between the quality of attachment to parents and IPV should prefer longitudinal design of research or the use of valid instrument.

Another central issue is related to the instruments used across the whole literature. Indeed, the prevalent use of self-report questionnaire and the poor (or inexistent) use of interview (e.g., AAI or CRI) did not allow to evaluate sufficiently the role of disorganized attachment in IPV. This issue could be especially problematic because of the role of past life trauma in IPV. Trauma in early childhood has been showed to be often associated with a disorganized quality of attachment, which in turn is associated with a wide range of negative outcomes as such as personality disorders and violence (Rholes et al., 2016). This important gap has to be fulfilled by future research which should select congruent instruments to estimate the role of disorganized attachment in IPV victimization and perpetration.

#### Focus of the Studies: Forms of Violence

Most of the study reviewed in this paper focused on physical and psychological violence and neglected the topic of sexual IPV. Noteworthy, sexuality is considered as the most exclusive domain of romantic relationship and is thought to be tightly related to the development of attachment model (Lichtenberg et al., 2007). In that sense, sexual violence perpetrated by a romantic partner is perhaps the most intimate form of IPV. Interestingly, sexual IPV has been often included in studies examining the relationship between attachment and IPV without differentiating between different forms of violence. As recently stated, research toward sexual IPV is plagued by important difficulties related to the definition of the construct and the lack of instruments available for its assessment (Bagwell-Gray et al., 2015). The few results reviewed in our paper are somewhat contrasting. Despite most of the studies found an association between insecure attachment and sexual IPV victimization, only one research examined the topic among participants with a previous reported history of IPV victimization (Smagur et al., 2018). In relation to sexual IPV perpetration, studies showed that perpetrators were not only anxious (as for the other types of violence) but also avoidant. These preliminary data are in line with the subtyping model of batterers proposed by Holtzworth-Munroe and Stuart (1994). The authors asserted that perpetrators of sexual IPV belong to the most severe group of batterers, often showing antisocial personality traits. However, further studies investigating the role of attachment in sexual IPV should be conduct to increase current understanding of the topic and to offer indications for the tailoring of treatment programs.

Also, another fundamental issue has been neglected by the studies examining the relationship between attachment and IPV: the topic of polivictimization. Indeed, researchers generally decided to create a non-specific index of IPV or to differentiate between several forms of violence. However, studies did not examined differences between single-form or combined-form IPV experiences. As polivictimization has been showed to be associated with worse outcomes, such gap should be fulfilled by future researches following the lines of Ross et al. (2016).

#### IPV and Anxious Attachment

Anxious attachment has been early thought to act as a risk factor for IPV victimization. Indeed, an individual with anxious attachment is usually described as suffering from fear of abandonment and high levels of separation anxiety. They may have difficulty to leave abusive relationships because the loss of the partner is experienced as unbearable and generate so much anxiety that the individual may prefer the least worse option. Similarly, the fact that anxious individuals suffer from low self-esteem (Mikulincer and Shaver, 2005) may lead them to think to not have sufficient resources to front a separation by the abusive partner. These individuals may be especially prone to deceive themselves about the possibility that partner will change. Also, anxious attachment is usually related to a negative self-image, as underserving of love and care. Together with low self-esteem, these characteristics may lead to selfattribution, in terms of responsibility, of IPV. Importantly, a violent partner who is intermittently loving and attending may further reinforce this interpersonal pattern, increasing the value of the relationship the individual fears to loss and favoring illusions about a future change of partner behavior. Unfortunately, these considerations about the mechanisms that operate in the link between anxious attachment and IPV victimization remain speculative, as too few studies included in their research design an examination of other potential variables as for example perception of social support or selfesteem.

In relation to IPV perpetration, results suggest a convergence in literature showing that batterers are prone to be anxiously attached to their partners. Importantly, these findings have been found in relation to every form of IPV. This is in line with the attachment theory of IPV, which assert that an anxious individuals tend to hyperactivate the attachment system, exaggerating protestation signals when attachment needs are not met. For example, they are especially demanding in terms of caregiving and love demonstrations. Also, when attachment needs are not met, they tend to use extreme forms of emotion regulation strategies that generally involve the interpersonal domain (e.g., being reassured by the partner). This hypothesis seems further supported by studies showing the role played by high levels of anger in anxiously attached perpetrators. Violence has been identified as one of these regulation strategies that anxious individuals may use when feeling too much frustration. This consideration may explain results identifying a high prevalence of anxiously attached individual among a specific subtype of batterers, violent only in family relationships (Holtzworth-Munroe et al., 2000). Indeed, they would use violence only to obtain satisfaction of their attachment needs, in intimate context.

The anxious component of attachment has also been involved in the dyadic explanation of IPV. If both partners are anxiously attached, conflict resolution strategies would be probably dominated by engagement by both partners, leading to an escalation of conflict and terminating in episodes of violence (Bonache et al., 2017). Indeed, studies investigating the attachment matching in violent couples showed that couples where both partner have an anxious attachment are more prone to be violent (Bookwala and Zdaniuk, 1998; Bookwala, 2002). Also, anxious interpersonal regulation strategies may lead the partner to retreat from the relationship, further intensifying the frustration of attachment needs and potentially triggering an escalation of intimate conflict, terminating in violent acts. Noteworthy, this pattern could be especially true when one partner has an avoidant attachment, tending to withdraw from conflicts.

#### IPV and Avoidant Attachment

Turning to the avoidance dimension of attachment, our review showed that near half of the studies found a relationship between avoidant attachment and IPV victimization. At an individual level, some consideration can be made in an attempt to understand why an avoidant victim of IPV does not leave a violent relationship. For example, avoidant individuals have typical difficulties in seeking help because of some dysfunctional beliefs. They are generally convinced that showing personal difficulties and vulnerabilities to others is unbearable as they expect that help request would be reject by others, fundamentally unavailable. As such, lack of social support, tightly related to IPV victimization (Zapor et al., 2018), may be very high among avoidant victims of IPV (Davis et al., 2002). Also, avoidant individuals may underestimate the psychological costs of IPV violence, being erroneously convinced to be psychologically immune to emotional threats. Supporting this idea, most of the evidences supporting a link between IPV and avoidant attachment has been brought in relation to psychological IPV victimization.

Turning to the other side of IPV, some studies reviewed in this paper showed that perpetrators are often avoidant in their attachment styles. However, empirical evidences are contrasting. Importantly, models offering a sub-classification of batterers are insightful. Indeed, it has been asserted that a subtype of individuals, being mostly antisocial and highly violent, are especially prone to be avoidant (Waltz et al., 2000). For this subtype, violence may be used as a way to control and manipulate the partner, exerting a politics of fear. This is in line with results showing that gender role stress mediates the relationship between attachment insecurity and controlling behavior among male batterers (Mahalik et al., 2005).

Moreover, these individuals are not only physically aggressive but also use psychological and sexual violence. Interestingly, our review seems to support such hypothesis underlying that the proportion of study finding significant associations between avoidant attachment and IPV are higher in the sexual and psychological sections compared to the physical section. Again, it is highlighted the proficiency to examine the topic of IPV differentiating between forms of violence. Noteworthy, regarding sexual violence, it has been showed that avoidance was significantly associated with IPV only among male and not among female. Following the idea that sexual violence in avoidant individuals may be a way to control the partner, this result makes sense as gender differences in sexuality are often attributed to a different valence in terms of dominance motivation (Malamuth, 1998; Toates et al., 2017).

At a dyadic level, avoidant individuals may elicit in the partner high activation of the attachment system because of a tendency to withdraw and retreat from the relationship. From this perspective, it has been asserted that disengagement during conflicts may be interpreted in the light of an abandonment threat by anxious attached individuals. Such behaviors, in conflictual context, may exasperate the frustration of the anxiously attached partner who in turn would be more prone to use violence as an extreme form of protestation. For example, Bonache et al. (2017) found that among boys, avoidant attachment was related to IPV victimization through self-reported withdrawal strategies and conflict engagement behaviors attributed to their partner. However, there is still lack of studies testing such hypotheses. Future research should include in their investigation other variables, as for example pathological personality and measures of romantic relationship power and dominance.

#### The Role of Gender and Sexual Orientation

As the proportion of men and women being involved in IPV victimization and perpetration is highly unbalanced, it is not surprising that a numbers of researchers explored the role of gender in the association between IPV and attachment. For example, some studies found that the relationship between avoidant attachment and both physical (Bond and Bond, 2004; Karakurt et al., 2013) and sexual (Sommer et al., 2017) IPV victimization was significant only among males. However, some authors found an inverse pattern of results with associations between insecure attachment and IPV victimization being significant only among women (Péloquin et al., 2011; Hellemans et al., 2015). Recently, it has been argued that gender discrepancies may be due to the matching between gender and type of conflict resolution strategies used. Bonache et al. (2017) argued that conflict resolution strategies, which are not in line with gender expectations, might be less accepted and consequently elicit the use of violence by the other partner. In this regard, studies conducted on homosexual population may be a profitable perspective from which observe the role of gender in the relationship between attachment and IPV. For example, Bartholomew et al. (2008) found that avoidance was negatively related to IPV victimization and perpetration among their sample of homosexual men. This contrasts with the studies, illustrated above, conducted among heterosexual population of men suggest that gender expectancies may play a role in the relationship between avoidance and IPV victimization. However, too few studies explore the topic among samples of homosexual individuals and mainly neglected IPV victimization. Indeed, if preliminary results toward IPV perpetration converge with those obtained among heterosexual population, further research is needed to increase the understanding of the topic.

# CONCLUSIONS

Our paper aimed to provide a complete review of empirical evidences investigating the role of attachment in relation to IPV.

Importantly, a great number of studies failed to find significant associations between insecure attachment and IPV victimization or perpetration. However, preliminary results evidenced that victims and perpetrators of IPV are heterogeneous population in relation to attachment. Importantly, IPV is not a deterministic phenomenon and the complex and multidimensional relationships between an individual, her/his resources and the risk factors occurring at different steps of the relationship should be considered. Indeed, a possible explanation is that anxious, avoidant and secure individuals might be at risk of experience or perpetrate IPV but for different reasons. In line with this, the investigation of the relationships between attachment and others central associates of IPV may further shed light on such issue. However, the literature reviewed in the paper often neglected the role of other important correlates of IPV. For example, attachment and IPV have been rarely investigated in relation to poverty or among populations of minority women. Indeed, too few studies include in their design other variables that may interact with attachment styles and explain the heterogeneity of these results.

Some interesting clinical implications might be drawn from this examination. First, attachment theory asserts that IWM could change over time in the context of secure and supportive relationships. As such, increasing social support and reinforcing the development of secure romantic relationships should be encouraged by clinicians. Also, clinicians themselves might provide a secure base to both victims and perpetrators in order to alter insecure IWM and to shift toward a secure one. Also, when working with victims of IPV, clinicians might guide patients toward an increased awareness of how attachment issues have affected their relationship. For instance, anxious attachment may include an awareness of the value of the relationship. The clinician may support the patient to maintain such awareness, framing it in a more positive way. However, clinicians should avoid the reinforcement of the erroneous attribution of internal blames for IPV that anxious victims may show. Instead, this therapeutic process should be promoted by a clinical support in the development of coherent narratives of early attachment experiences. Then, some potential mediating variables explaining the relationship between insecure attachment and IPV victimization and perpetration might be the target of treatment. For instance, emotion regulation capacities (Garofalo and Velotti, 2015; Balzarotti et al., 2016) and especially deficit in the capacity to regulate anger (Garofalo and Velotti, 2017; Velotti et al., 2017) may be a strategic objective in the treatment of perpetrators. Finally, communication capacities related to attachment needs should be improved in both perpetrators and victims. For example, anxious perpetrators should be supported in their capacity to interpret disengagement from partner and to better tolerate and communicate emotions related to interpersonal rejection.

Despite the important contribution provided by this paper, some limitations should be pointed out. For instance, publication bias is a well-documented limitation of systematic reviews and the present work is not an exception. Indeed, we excluded not published studies potentially leading to a misrepresentation of findings in the field. Especially, studies with inconsistent or negative results are hard to publish. Also, most of the studies reviewed in our paper are cross sectional in their design, limiting the possibility to made causal inferences. Specifically, the inability to answer to the question of whether attachment styles precede IPV and to what extent they are the result of psychological changes and specifically changes in interpersonal area, as consequences of IPV, remains a central issue. Finally, in our systematic and qualitative review, we included all studies without important exclusion criteria regarding their methodological quality. However, a further quantitative review should consider this limitation and assess the risk of potential biases deriving from selection of participants, data collection and analysis.

# AUTHOR CONTRIBUTIONS

PV took overall responsibility for the conceptualization and design of the review. She revised it critically for important intellectual content. SB and GR searched for the articles in the review, assessed them for relevance. PV, SB, GR, and RT were involved in the interpretation of data, in writing and editing

# REFERENCES


the final article; final approval of the version to be published; agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

# SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyg. 2018.01166/full#supplementary-material


attachment and relational aggression. J. Marital Fam. Ther. 40, 412–429. doi: 10.1111/jmft.12079


adolescent dating relationships inventory. Psychol. Assess. 13, 277–293. doi: 10.1037/1040-3590.13.2.277


**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 © 2018 Velotti, Beomonte Zobel, Rogier and Tambelli. 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 Intimate Partner Violence Meets Same Sex Couples: A Review of Same Sex Intimate Partner Violence

#### Luca Rollè, Giulia Giardina\*, Angela M. Caldarera, Eva Gerino and Piera Brustia

Department of Psychology, University of Torino, Turin, Italy

Over the past few decades, the causes of and intervention for intimate partner violence (IPV) have been approached and studied. This paper presents a narrative review on IPV occurring in same sex couples, that is, same sex IPV (SSIPV). Despite the myth that IPV is exclusively an issue in heterosexual relationships, many studies have revealed the existence of IPV among lesbian and gay couples, and its incidence is comparable to (Turell, 2000) or higher than that among heterosexual couples (Messinger, 2011; Kelley et al., 2012). While similarities between heterosexual and lesbian, gay, and bisexual (LGB) IPV were found, unique features and dynamics were present in LGB IPV. Such features are mainly related to identification and treatment of SSIPV in the community and to the need of taking into consideration the role of sexual minority stressors. Our findings show there is a lack of studies that address LGB individuals involved in IPV; this is mostly due to the silence that has historically existed around violence in the LGB community, a silence built on fears and myths that have obstructed a public discussion on the phenomenon. We identified the main themes discussed in the published studies that we have reviewed here. The reviews lead us to the conclusion that it is essential to create a place where this subject can be freely discussed and approached, both by LGB and heterosexual people.

Edited by: Pedro Alexandre Costa,

Instituto Universitário de Ciências Psicológicas, Sociais e da Vida, Portugal

#### Reviewed by:

Michelle Lowe, University of Bolton, United Kingdom Magdalena Zadkowska, University of Gdansk, Poland ´

> \*Correspondence: Giulia Giardina giulia.giardina@edu.unito.it

#### Specialty section:

This article was submitted to Gender, Sex and Sexuality Studies, a section of the journal Frontiers in Psychology

> Received: 04 May 2018 Accepted: 30 July 2018 Published: 21 August 2018

#### Citation:

Rollè L, Giardina G, Caldarera AM, Gerino E and Brustia P (2018) When Intimate Partner Violence Meets Same Sex Couples: A Review of Same Sex Intimate Partner Violence. Front. Psychol. 9:1506. doi: 10.3389/fpsyg.2018.01506 Keywords: same sex intimate partner violence, same-sex couple, LGB, domestic violence, IPV, treatment

# INTRODUCTION

Over the past few decades, intimate partner violence (IPV) has received increasing interest from mental health experts. According to the World Health Organization (2012), IPV is related to any behavior between a couple that involves acts of physical and sexual violence, emotional and psychological abuse, and controlling behavior. According to numerous authors, the expression "IPV" represents a form of violence that both men and women can enact, with no regard to age, marital status, or sexual orientations (Capaldi et al., 2007; Ali et al., 2016). The consequences of IPV on mental health and general wellbeing have also been outlined in numerous studies (Campbell, 2002; Anderson et al., 2008; Murray and Mobley, 2009; Giordano et al., 2014; Costa et al., 2015).

The lesbian, gay, and bisexual (LGB) population faces more difficult outcomes compared to the heterosexual population "across different life domains, including mental and physical health, subjective wellbeing, employment, poverty, homelessness, and social exclusion" (Perales and Todd, 2018, p. 190). IPV in the LGB population has not been studied as frequently

as in the heterosexual population: in 2015, research on LGB IPV constituted a mere 3% of the total research on the subject (Edwards et al., 2015). Even though there are a few studies on Same-Sex Intimate Partner Violence (SSIPV), they highlight that the phenomenon occurs at a rate that is comparable (Turell, 2000) or even higher than heterosexual IPV (Messinger, 2011; Kelley et al., 2012; Barrett and St.Pierre, 2013). It can be difficult to identify LGB IPV prevalence rates due to the different methodologies used in the researches. However, according to one of the most recent and representative study reports, almost onethird of sexual minority males and one-half of sexual minority women in the United States affirmed they were victims of physical or psychological abuse in a romantic relationship. In addition, over 50% of gay men and almost 75% of lesbian women reported that they were victims of psychological IPV (Breiding et al., 2013). Breiding et al. (2013) identified that 4.1 million people of the LGB community have experienced IPV in their lifetime in the United States.

Life-time prevalence of IPV in LGB couples appeared to be similar to or higher than in heterosexual ones: 61.1% of bisexual women, 43.8% of lesbian women, 37.3% of bisexual men, and 26.0% of homosexual men experienced IPV during their life, while 5.0% of heterosexual women and 29.0% of heterosexual men experienced IPV. When episodes of severe violence were considered, prevalence was similar or higher for LGB adults (bisexual women: 49.3%; lesbian women: 29.4%; homosexual men: 16.4%) compared to heterosexual adults (heterosexual women: 23.6%; heterosexual men: 13.9%) (Breiding et al., 2013).

Messinger (2011) highlighted that all forms of abuse were more likely to occur in homosexual and bisexual couples than in heterosexual ones. Moreover, he hypothesized that a higher percentage violence was caused by unique risk factors linked to minority stress that is experienced only by LGB people. In addition, the study highlighted that lesbian women were at higher risk of being involved in IPV, followed by heterosexual women, gay men, and heterosexual men. Furthermore, bisexual people appeared to be the most abused group compared to the others; bisexual women, specifically, were more likely to be victims of every type of IPV, excluding psychological IPV.

Most researches on the prevalence of SSIPV have been conducted on a North American population, while some minor studies are focused on Australian (Leonard et al., 2008), Chinese (Chong et al., 2010; Liu et al., 2013), South African (Eaton et al., 2013), and British populations (Guasp, 2012): the results reported similar or even higher IPV rates compared to those for North American populations. Chard et al. (2012), in their transnational research, evidenced the differences in prevalence rates among various countries: participants were recruited through advertisements on Facebook in the United States, Canada, Australia, United Kingdom, Republic of South Africa (RSA), Brazil, Nigeria, Kenya, and India. Their findings showed similar rates between United States and the other nations, while the rate of physical abuse appeared to be similar or more likely to occur in Australia, Brazil, Republic of South Africa, and the United Kingdom than in the United States.

In Italy, two studies were conducted on lesbian IPV—one by Moscati (2016) (as part of a European project) and a survey by Arcilesbica (2011). Moscati (2016) work was mainly focused on the absence of protective laws for lesbian women victims of IPV, and Arcilesbica (2011) attempted to estimate IPV prevalence among Italian lesbian women. The sample comprised 102 lesbian women, mostly Italian (88.2%). Participants answered a questionnaire containing 29 multiple-choice questions. In over one case out of five (20.6% of the total), the interviewee admitted to be afraid of her partner coming back home. Further, 41.2% of women occasionally hid something from their partners because they were afraid of their reactions. In addition, 14.7% of lesbian women declared that they were always afraid of their partners. Almost half of the interviewees identified the damage resulting from a couple fight as psychological; physical damage was reported by 5.9% of the interviewees (Arcilesbica, 2011).

In the light of such findings, it is apparent that LGB IPV needs to be studies further. Nonetheless, public opinion considers LGB abuse as a rare phenomenon: this opinion is particularly strong with regard to bisexual and lesbian women, often idealized as being in peaceful and utopian relationships, far from the violence and aggression that is commonly associated with "typical" male virility (Glass and Hassouneh, 2008; Barnes, 2010). Such a stereotype can be an obstacle to lesbian victims in recognizing that a partner behavior is abusive and not normal (Seelau and Seelau, 2005).

Previous research has suggested the need of further research on the issue: LGB IPV has a double invisible nature that is responsible of the lack of studies on it. In the past, health experts found many obstacles in accessing research and data on SSIPV, a fact that implicated negative consequences in terms of prejudice and misinformation in addition to the more obvious outcomes (Messinger, 2011).

# AIMS

In the light of the background outlined above, this paper presents a narrative review aimed at (1) providing an overview, through a selective narrative review, of the psychological literature on LGB IPV, with a specific focus on treatments and interventions addressed both to victims and perpetrators, and (2) identifying, from the literature, suggestions for future directions in research for LGB-oriented psychological and community services in relation to IPV and the themes outlined by the overview.

# METHOD

A literature research was conducted by using the following databases: PsycINFO, PsycARTICLES, PubMed, and Google Scholar. The search criteria was that eligible studies should have been published in English or Italian, between 1995 and 2017, and focused on the main features of LGB IPV.

The following combinations of keywords were used to conduct the research: (1) Same-sex intimate partner violence OR, SSIPV OR, LGB intimate partner violence OR, LGB IPV; (2) Samesex domestic violence OR, LGB domestic violence; (3) Lesbian domestic violence; (4) Gay domestic violence; (5) Bisexual

domestic violence; (6) Prevalence; (7) Minority stress; (8) Treatment; and (9) Intervention.

**Table 1** presents the selection criteria applied to select the papers.

We created a dataset of the selected papers and conducted a thematic analysis (TA) in order to outline patterns of meaning across the reviewed studies (Braun and Clarke, 2006), using a semantic approach. Braun and Clarke (2006) provided guidelines for conducting the TA, which included a process organized in six phases: (1) Familiarization with the data; (2) coding; (3) searching for themes; (4) reviewing themes; (5) defining and naming themes; and (6) writing up.

Thus, after the Familiarization phase, we assigned each article with a short label that identified the main results that could be relevant to our aims (Coding phase). Thereafter, we identified broader patterns of meaning, each representing a candidate theme to which the papers were allocated. Next, we stepped into the Reviewing themes phase and checked back the candidate themes confronting them with the studies dataset. We attempted to define more inclusive thematic areas by combining specific candidate themes and by selecting a pool of the most frequent ones, which led us to the Defining and naming themes phase. As a result of this process, we identified the six main themes that were focused on in the studies: silence around the phenomenon; association with Sexual Minority Stress; assessment and treatment; couple and group interventions; victims' treatments; and access to services offering help and support.

#### RESULTS

The first outcome of the research included 4700 sources, from which we eliminated duplicates, researches published in languages other than English and Italian, contributions from books or sources other than published articles and surveys. After this selection process, 119 studies met the inclusion criteria for this review.

#### Silence Around Violence

Understanding LGB IPV prevalence and related factors may be difficult because of the silence that has historically existed around violence in the LGB community. Research has revealed that in the LGB community, several common fears became an obstacle for a public discussion on the phenomenon. For example, an aspect frequently claimed was that recognizing IPV in the LGB community may be used to stigmatize the community itself, thereby contributing to building additional oppression and social marginalization (Kaschak, 2001; Ristock, 2003). Similarly, the feminist community was averse to discussing the phenomenon, particularly when it involved lesbian couples: a public discussion on lesbian IPV may increase negative reactions to feminism and female homosexuality; on the other hand, it may minimize the concern regarding male violence against women (McLaughlin and Rozee, 2001; Ristock, 2001, 2003).

Furthermore, culturally created ideologies regarding masculinity and femininity may discourage IPV victims from openly discussing their experience. This happens when the perceived stigma reinforces their own stereotype that homosexual men are less masculine than heterosexual men, or the one that lesbian IPV is harmless (because women are not physically strong and dangerous) (Ristock and Timbang, 2005). Buttell and Cannon (2015) stated that IPV was not about genders, but more about power and control dynamics; thus, to assess and treat IPV, particularly LGB IPV, it is pointless to take into account gender-related stereotypes (Brown, 2008; Little and Terrance, 2010). However, the main resistance from the feminist community came from the risk that discussing lesbian IPV may threaten a feminist belief regarding women's abuse, usually perpetrated by men who are influenced by misogyny and patriarchy. Gender and power were the main factors in this theory; therefore, lesbian victimization was considered both impossible (because of the inconsistency due to the absence of a man in the equation) or explained by the assimilation among lesbian women of misogyny and homophobia, which is subsequently projected on to their partners as women and homosexuals (Ristock and Timbang, 2005).

Many LGB individuals experienced additional victimization and homophobia when they reported the abuse to police (Barnes, 1998; Burke et al., 2002; Bentley et al., 2007; Guadalupe-Diaz and Yglesias, 2013). The LGB community Legal Rights and Protection Laws are intended to protect the LGB community (Moscati, 2016).

Bunker Rohrbaug (2006) indicated that one of the most pervasive and alarming myth was considering violence as a mutual conflict, particularly when the violence occurred in a gay couple, because men "fight equally," as they are assumed to have comparable physical strength. This myth was legitimized by the societal attitude with regard to tolerating violence expressions between men, expressions that were considered admissible and often normalized as a means of dispute resolution or because of greater congruence between violence and male roles (Baker et al., 2013).

This idea implicated serious issues because not only did it created obstacles in providing services for homosexual victims but it also contributed to increasing the tendency to minimize IPV severity (McClennen, 2005). Such an assumption

TABLE 1 | Inclusion and exclusion criteria.


could neglect the study of other types of violence apart from the physical one, such as psychological abuse (Finneran and Stephenson, 2013). One of the reasons for why the "mutual fight myth" was so pervasive is related to researches that reported how common it was for a partner to be violent (Carvalho et al., 2011; Edwards and Sylaska, 2013). This myth was proved to be unfounded when motivations why partners fight back were considered. In this regard, several researches (Merrill and Wolfe, 2000; Bartholomew et al., 2006) attested that self-defense was the most common cause that victims reported to justify their fighting back. Further studies (Bartholomew et al., 2008; Little and Terrance, 2010; Bimbi et al., 2011) also investigated the victim fighting-back phenomenon and suggested that, because of the mutuality, the distinction between survivor and perpetrator might be less clear in LGB communities. Ristock (2001) affirms that fighting back was not only self-defense but also a claim to power and higher position between the couple. A further hypothesis supposed that additional and hidden power dynamics may contribute to the occurrence of IPV. These issues reinforced the illusion that violence was mutual (Ristock and Timbang, 2005). Moreover, the belief that it would be easier for gay men to leave an abusive relationship needs to be considered. This idea arose from another stereotype related to homosexual men being unable to be involved in a stable relationship and often and easily changing partners instead LGB relationship can be as stable as heterosexual ones (Gates, 2015).

Several studies (Austin et al., 2002; Girshick, 2002; Balsam and Szymanski, 2005; Bornstein et al., 2006; Messinger, 2011; Galletly et al., 2012) claimed how bisexual people experienced an additional stress related to IPV because of the lack of support from the LGB community. Bisexual people were doubly marginalized, not being recognized by lesbian and gay people as part of their community and, simultaneously, being stigmatized by heterosexuals. The assumption that bisexual people use the heterosexual privilege leads to the fact that a lot of lesbian and gay people believe that the victimization of bisexual people is not as serious as that of lesbian and gay people. Davidson and Duke (2009)showed that bisexual people were victims of the law system and the services to the same extent. Moreover, studies showed that biphobia within the LGB community increased the risk of IPV between bisexual partners and, simultaneously, reduced help-giving resources (Austin et al., 2002; Girshick, 2002; Balsam and Szymanski, 2005; Bornstein et al., 2006; Messinger, 2011; Galletly et al., 2012).

Without overlooking the peculiar aspects of the LGB community, authors compared the general patterns, types, impact and cycle of violence of LGB IPV and heterosexual IPV (McLaughlin and Rozee, 2001; Hequembourg et al., 2008). Like heterosexual victims, homosexual and bisexual people experienced emotional, physical, and sexual abuse. The outcomes were severe, and included physical injury, social isolation, property destruction and loss, and disruption to work, education, and career development (Buford et al., 2007; Chard et al., 2012; Barrett, 2015). Moreover, victims often reported that the abuse was not mutual and was instead suffered, and the consequences of it made them feel trapped, hopeless, and isolated (Ferraro and Johnson, 2000; McClennen, 2005). There were also similarities with regard to the reasons for remaining with the abusive partner. Both heterosexual and homosexual victims commonly listed the following aspects as reasons to stay: love for the partner, financial and emotional dependency on the partner, (Merrill and Wolfe, 2000). A further resemblance was the connection between stress, violence, and use of substances (Buford et al., 2007; Cain et al., 2008): IPV was related both to depression and substance use among LG people with a previous IPV history, who appeared to have a higher tendency of drug abuse (Kelley et al., 2011).

Gill et al. (2013) highlighted that the higher prevalence rate of HIV in the LGB population also constituted an important difference in their experience of IPV. Merrill and Wolfe (2000) results showed that the main reasons why HIV-positive IPV victims did not leave the relationship were linked to the fear of becoming sick and dying alone or of dating in the context of the disease. On the other hand, HIV-positive partners remained in the relationship because they did not want to abandon their sick partners. The authors argued that IPV increased vulnerability to risks associated with HIV transmission, which in turn affects medical care, mental health, adherence to therapy, frequency of follow-up; in addition, they found that IPV contributed per se to HIV transmission, directly through forced unprotected sex with a partner or indirectly by impairing the victim's ability to negotiate safer sex. Individuals may experience difficulties in negotiating safer sex for several reasons, including the perception of being unable to have control over sex, fear of violence, and unequal power distributions in the relationship (Bowen and Nowinsky, 2012; Gill et al., 2013). In light of these data, it can be said that IPV may be common among people living with HIV. Therefore, it is essential that all service providers screen and provide assistance for issues relating to safer sex, similarly, all HIV service providers should screen for IPV and discuss safety within the context of abusive relationships and helping their clients have safer sex (Heintz and Melendez, 2006).

Even though this fact represented an issue in the heterosexual population, LGB people were more affected by it. In fact, in Merrill and Wolfe (2000) study the lack of knowledge about IPV was the third most commonly reported cause to remain in an abusive relationship. This might be due to the fact that historically, IPV was defined and studied in a heterosexual perspective, excluding any mention of same-gender relationships (Glass and Hassouneh, 2008; Little and Terrance, 2010). There are few existing examples of educational campaigns on LGB IPV, although the research proved how this kind of interventions is effective in encouraging battered people to report the abuse. Consequently, LGB partners involved in violence, and people close to them (services providers, family, friends), evaluated the battering as less dangerous or not harmful at all, and it usually took a longer time to recognize it as an abuse (Dixon and Peterman, 2003; Barrett, 2015).

# SEXUAL MINORITY STRESS

Carvalho et al. (2011) argued that LGB people experience unique stressors related to the condition of being a part of a sexual minority. These stressors, that appear to be

associated to IPV, reflected the experience of Sexual Minority Stress, a model developed by Meyer (2003) with regard to members of a stigmatized group who experienced unique and additional stressors that nobody outside the group could ever experience. This model included internalized stressors (internalized homophobia, disclosure, and stigma consciousness) and externalized stressors (actual experiences of violence, discrimination, and harassment) (Meyer, 2003). Research showed how internalized stressors were positively correlated to physical, sexual, and psychological IPV (Balsam and Szymanski, 2005; Bartholomew et al., 2006; Carvalho et al., 2011; Edwards and Sylaska, 2013); on the contrary, externalized stressors were not related to any form of IPV, particularly when they were considered with internalized minority stressors (Balsam and Szymanski, 2005; Bartholomew et al., 2006; Edwards and Sylaska, 2013).

Thus, studies mainly focused on internalized minority stressors, such as Internalized Homophobia, establishing that IPV perpetrators addressed their negative emotions, originally self-addressed as homosexuals, to their partners. People with internalized homophobia have been deprived by partners of positive emotions with regard to their sexual orientation and reinforced their sense of responsibility in provoking the abuse (Balsam and Szymanski, 2005; Carvalho et al., 2011). Carvalho et al. (2011) reported that internalized homophobia and IPV were related in lesbian couples and influenced by the quality of the relationship. Therefore, both couples' variables and individual experiences were equally fundamental in understanding the homosexual IPV phenomenon (Balsam and Szymanski, 2005; Carvalho et al., 2011). Although the relationship between internalized homophobia and IPV was proven, data suggested that it was not strong (D'Lima et al., 2014). This result might be due to the fact that research participants showed low levels of internalized homophobia, because it is rare that LGB people with high levels of internalized homophobia would cooperate for any LGB study. A further cause could be that the sample comprised highly educated white people (Carvalho et al., 2011).

Two researchers reported that disclosure was positively related to the risk of physical and psychological IPV: Bartholomew et al. (2006) analyzed a sample comprising homosexual and bisexual men, while Carvalho et al. (2011) studied the phenomenon among lesbian women. Such findings may be due to the fact that being openly out implied a longer period of time of being victimized by the partner but also the opposite: a shorter time in LGB relationships could imply lower chances to be involved in an abusive one (Bartholomew et al., 2006; Carvalho et al., 2011). With regard to this last aspect, perpetrators could intimidate the victim by threatening to oust them in front of their family, employer, landlord, former partner, or current guardian of their children (Borne et al., 2007; Carvalho et al., 2011).

The Consciousness Stigma has been the last internalized minority stressor studied in relation to IPV. Carvalho et al. (2011) indicated that stigma consciousness increased the likelihood of IPV. IPV perpetrators and victims reported high stigma consciousness rates; thus, it can be assumed that IPV makes people more worried about stigma consciousness and that it is positively correlated to the tendency to ignore abuse in order to protect IPV victims from the homophobic legal system.

Such results match with high stigma consciousness rates in people who are expected to suffer discrimination and be forced to avoid discriminating situations (Pinel, 1999; Derlega et al., 2003). To what we know, literature offers several evidences regarding the connection between minority stressors and SSIPV. As mentioned earlier, internalized stressors and IPV were strongly correlated. Some studies (Balsam and Szymanski, 2005; Carvalho et al., 2011; Finneran and Stephenson, 2014) showed the existence of a relationship between experienced discrimination and a higher risk of IPV. On the other hand, studies on the relationship between experienced discrimination and risk of SSIPV victimization are contradictory: some indicated the existence of such a relationship (Carvalho et al., 2011; Andrews et al., 2014; Finneran and Stephenson, 2014), while some denied it (Barrett and St.Pierre, 2013; Andrews et al., 2014).

These findings suggest that the connection between discrimination about sexual orientation (based on other people emotions and beliefs) and IPV is not completely clear, but that a relation between victimization and individual feelings regarding one's own sexual orientation (internalized homophobia and stigma consciousness) exists (Edwards et al., 2015). However, it should be noted that such considerations are the result of cross-sectional studies, thereby making it difficult to determine whether a factor developed prior to, during, or after the occurrence of IPV. This implies that it is important to be cautious in generalizing such findings; instead, further research must be conducted on predictors and associated factors (Edwards et al., 2015). Moreover, clinicians should be aware that minority stressors are one of the main obstacles for people who have experienced or are involved in IPV and seeking help, and what could assist them: it was proven that heterosexism exacerbates difficulties in reporting the abuse to the police and in accessing in services for LGB people (Carvalho et al., 2011). IPV victims can be reluctant in seeking legal assistance, fearing discrimination or adequate legal protection. Balsam (2001) observed that over 60% of lesbian women who were interviewed decided not to leave the abusive partner because of lack of resources, and a majority of the sample did not seek help in a women's shelter. In line with Balsam (2001) and Buford et al. (2007) emphasize that services and shelters were often unprepared to support homosexual victims of IPV.

Overstreet and Quinn (2013) created the IPV Stigmatization Model to explain barriers to seeking help. The model described three aspects of the individual experience: "stigma internalization," "anticipated stigma," and "cultural stigma." Stigma internalization referred to the impact of internalized negative beliefs regarding IPV, which can influence individuals' help-seeking behaviors and psychological distress. Surviving IPV can cause guilt, shame, and self-blame, all of which are challenges in seeking help for decreased self-efficacy. Anticipated stigma, that also functions at the interpersonal level, was regarding concerns related to whether others will react with disapproval or rejection toward the survivor when they learn about the IPV, thereby affecting the decision to seek help. Lastly, cultural stigma referred to the notion that IPV victims provoked their own victimization.

# LGB IPV ASSESSMENT AND TREATMENT

fpsyg-09-01506 August 17, 2018 Time: 18:40 # 6

The first program for SSIPV was developed in United States and strictly connected or identical to the ones for heterosexual population (Dixon and Peterman, 2003; Ristock and Timbang, 2005). However, a specific risk was highlighted in considering IPV as a universal experience, since this assumption implicated that the treatment might be the same for each person (Ford et al., 2013). There were similar aspects between heterosexual and homosexual IPV relationships, therefore policies and services tailored for heterosexual may be helpful to design specific interventions for LGB population (Dixon and Peterman, 2003; Ristock and Timbang, 2005). Heterosexual model can be the starting point for treatments addressed to LGB people, who deserve interventions based on their own peculiar experiences and needs (Finneran et al., 2013).

Renzetti (1996) examined the outcomes of the application of an unspecific treatment that did not consider sexual orientation and gender. In that study, 566 North American services were involved, part of the National Directory of Domestic Violence Programs, mostly addressing IPV heterosexual victims. Almost 96% of the workers declared that they were indiscriminately welcoming and responsive to all kind of people seeking help, according to a non-discrimination policy. However, there was discord between the statements made by mental health providers and the victims' reports. In fact, just one out of ten victims received particular care specifically directed to lesbian women, but the remainder claimed that operators did not make any effort to comply with their needs. Other researches (Giorgio, 2002; Helfrich and Simpson, 2006) conducted in the United States confirmed this condition: victims reported heterosexism, discrimination, stigma, ridicule, disbelief, additional abuse, and hostility from services. Cheung et al. (2009) conducted a global Internet-based study on Asian men accessing services as IPV victims. Authors reported that gay men were not perceived as domestic violence service consumers unless they were perpetrators (Cheung et al., 2009). On the other hand, lesbian women highlighted a heterosexist language adopted by emergency, primary care, and other service providers (Dixon and Peterman, 2003). It is considered that services are rarely available for LGB people, and when they are, it is often difficult to access them, particularly in rural areas (Jeffries and Kay, 2010; Ford et al., 2013). Thus, it appears clear how heterosexual IPV, widely studied, can be considered as a starting point to better investigate and address homosexual couple abuse, without overlooking LGBspecific factors (Finneran et al., 2013).

# LGB-Tailored Assessment

Because of the multiple barriers and the invisibility of the problem in the context of IPV services, the role of the victims' health care providers is critical. While it was found that in the United States many emergency departments, shelters, agencies, and clinics had IPV advocacy programs, most of these programs historically failed in responding adequately to abuse in LGB groups (Brown and Groscup, 2009; Hines and Douglas, 2011; Armstrong et al., 2014). Goodman et al. (2015) contended that during initial steps, services providers should recognize the problem, offer empathic support, ensure safety, and help the victim gain distance from a dangerous situation. Healthcare workers should screen for IPV in LGB patients and understand how patterns of IPV are different for these patients (Banks and Fedewa, 2012; Armstrong et al., 2014): standard approaches to IPV screening may be ineffectual for LGB people (Cavacuiti and Chan, 2008). Ard and Makadon (2011) highlighted the need for a sensitive and accurate assessment, which they discussed through clinical, institutional, educational, and research suggestions. The authors indicated that providers must be alert to the possibility of IPV as a cause of distress and illness among their LGB patients. Thus, according to them, clinicians should first inquire about sexual orientation in a sensitive and open manner, rather than simply screening for IPV (Ard and Makadon, 2011). Further, clinicians must use an inclusive language, avoiding any type of homophobic attitude, beginning from the first contact with the client (Eliason and Schope, 2001; Finneran et al., 2013). Ard and Makadon (2011) also highlighted how assessing LGB IPV fulfilled an important educational role for their LGB clients, because of the invisible nature of the phenomenon. Merrill and Wolfe (2000) discussed "recognition failure" as the failure to recognize intimate violent behaviors and, therefore, to seek or offer help such because of widespread ignorance regarding SSIPV. Several authors support public and specialized education believing that it would reduce the incidence of this phenomenon, by promoting earlier help-seeking and strengthening informal and formal support systems for victims (McClennen, 2005; Borne et al., 2007).

Merrill and Wolfe (2000) recommended similar suggestions, considering that SSIPV assessment and treatment should include the following aspects:



# LGB-Tailored Treatments

Even if research testified serious lacks in existing services (Herrmann and Turell, 2008; Brown and Groscup, 2009; Hines and Douglas, 2011), Ristock and Timbang (2005) reported examples of innovative programs developed within LGB communities. They introduced different interventions compared to heterosexual IPV protocols, serving both survivors and perpetrators. For example, they offered batterer intervention programs as well as advocacy programs to help LGB people access the legal justice system (The Los Angeles Gay and Lesbian Center) (Ristock and Timbang, 2005). Further, two approaches focused on the specific needs of queer women in San Francisco were the one promoted by The Queer Asian Women's Shelter (Chung and Lee, 1999) and the one from Queer Asian and Pacific Islander Women (Lee and Utarti, 2003): they attempted to better respond to IPV and address the complexities of being part of a small marginalized community such as the LGB one, teaching how to ask service providers for help. Similarly, the Washington State Coalition Against Domestic Violence developed a protocol for working with friends and family members of IPV victims. As the research highlights, most of the time, victims of violence asked friends and family for help before accessing services, thereby giving them a primary supporting role.

In certain cases, services were associated with communitybased initiatives that involved holding workshops and forums to address healthy relationships (Cronin et al., 2017). Ristock and Timbang (2005) and highlighted how discussion on building healthy relationships appeared to be more welcomed from lesbian victims than support groups for survivors. This fact might be due to victims' concerns regarding their privacy, which was protected during conversations on several topics connected to violence. Such discussion may explore other issues such as expectation in relationships, negotiating differences, power issues, and warning signs of abuse rather than identifying who experienced violence and respecting participants privacy. Another objective was also to shift from organizational interventions to a communitybased prevention to support health relationships and provide information and prevention to lesbian communities (Fonseca et al., 2009; Ford et al., 2013). The variety of approaches presented attempt to better respond to local settings rather than standardizing programs (Hatzenbuehler et al., 2015).

Another attempt to provide adequate services to LGB people was made by The Violence Against Women Act (VAWA) in 2013 that allowed the creation of services in the United States that are specifically designed for LGB victims of IPV and a legislation with regard to their rights. The act explicitly included a nondiscrimination clause that prohibited LGB individuals from being turned away from shelters or other programs funded by The Violence Against Women Act (Armstrong et al., 2014).

Further, several treatments and programs have been developed for individuals who experienced IPV. Some programs focused exclusively on treating the symptoms experienced by the victims, while others attempted to break the cycle of violence through interventions addressed for batterers. The types of interventions ranged from couple and group interventions to individual psychotherapy (Fountain and Skolnik, 2007; Herrmann and Turell, 2008; Dykstra et al., 2013; Armstrong et al., 2014; Quillin and Strickler, 2015).

# Couple and Group Interventions

Lesbian, gay, and bisexual partners often ask for treatment as a couple, and it is only after an initial assessment it becomes evident that the relationship is abusive. Frequently, with the aim of protecting victims, clinicians recommend separate services and refuse to provide couple therapy (Borne et al., 2007). In certain cases, this attitude was damaging and resulted in clients discontinuing treatment or seeking a different therapy (Istar, 1996). Merrill and Wolfe (2000) found that couple therapy was disadvantageous in IPV cases because it made it more difficult for victims to end the relationship and giving violence the label of "couple issues." It also made it particularly difficult for the therapist to guarantee victims' safety after therapy: occasionally, it created additional violence because of certain statements made by the therapist. Moreover, the authors indicated that couple therapy hindered an accurate assessment of the abuse because of victims' fear of repercussions. In certain cases, it damaged partners because of therapist counter-transference, who believed it was right to punish the violent person in the couple in order to protect the victim instead of sticking to therapy (Merrill and Wolfe, 2000).

Dykstra et al. (2013), in their review on IPV treatment effectiveness, found that couple therapy can be an effective treatment and it is occasionally necessary, particularly during the initial phases, to adequately assess the dynamics of the relationship. Moreover, an accurate assessment of the violence and the associated risks should be required in considering couple violence as a treatment option; this would enable the provision of the most suitable assistance for the couple in terms of defining or redefining problems, which can be treated through individual

treatment plans (Borne et al., 2007). Couples therapy can provide a safe space where relationships can be discussed and negotiated (Gilbert et al., 2017). On the other hand, couples therapy can be self-defeating if one or both of the partners presents issues that would best be previously acknowledged through individual counseling (Borne et al., 2007).

The effectiveness of couple therapy increased when combined with either individual or group therapy (Ristock and Timbang, 2005; Gilbert et al., 2017). Coleman (2003) highlighted that the optimal treatment for perpetrators is group therapy combined with long-term psychoanalytic psychotherapy or psychoanalysis.

Dykstra et al. (2013) evidenced that group therapy can be extremely useful in treating IPV and create room for improving emotional and social functioning. Group therapy made it possible to experience support and confrontation in a safe space, thereby avoiding isolation—a common consequence of victimization. The peer group assisted individuals with reliability by challenging unhealthy conduct and encouraging healthy behaviors. On the other hand, perpetrators too had the opportunity to learn new cognitive and behavioral strategies for managing their abusive impulses and express their emotions in a functional and structured manner (Buttell and Cannon, 2015). Occasionally, in case patients refuse to participate in group therapy, group therapy activities can be adapted to individual cases. Coleman (2003) listed some specific techniques: time-outs, control logs, and the Iceberg Exercise (that helped patients to identify emotions underlying their anger).

# Victims' Treatments

A few studies on treatment for LGB IPV victims were conducted in the United States (Browning et al., 1991; McClennen et al., 2002; Dixon and Peterman, 2003; Buford et al., 2007; Fountain and Skolnik, 2007; Ard and Makadon, 2011; Franklin and Jin, 2016). Studies showed that individual mental health counseling can result in good outcomes for SSIPV victims. Couple counseling with victim and abuser was found to be less effective because victims may fear repercussions from the information given during the session (such as details of the victimization) (Buford et al., 2007; Winstead et al., 2017). In spite of these findings, research has indicated that psychology graduate students and clinicians have the inclination to suggest couples counseling instead of individual counseling for LGB IPV victims more often than for differentgender victims (Wise and Bowman, 1997; Poorman et al., 2003).

Two types of counseling proposed as ideal for SSIPV victims were the person-centered approach and Gestalt therapy. These approaches allowed victims to gradually feel more trustful toward therapists and thus become aware of their status, the suffered abuse, and the associated consequences to it (Dietz, 2002). Moreover, it encourages therapists to enable victims to direct the session, thereby learning, in this manner, how to effectively direct their lives. Dixon and Peterman (2003) found that because of the strong motivation to accept help, victims' awareness about the abuse was believed to be longer-lasting. This fact granted victims to gain and adopt useful resources to bring an end to the abusive condition and obtain independence from the partner.

#### Interventions Addressed to the Abusers

In the United States, it is not unusual for abusers to participate in psycho-educative programs finalized to reduce the risk of committing violence on partners in the future. These programs are called "batterer intervention programs" and are based on the following two models (Price and Rosenbaum, 2009; Buttell and Cannon, 2015):


Dykstra et al. (2013) highlight that the Duluth model, alone or combined with CBT techniques, was the most frequently used program in the treatment of abusers. Both approaches do not consider the peculiarities of LGB couples and the role played by factors such as homophobia (Buttell and Cannon, 2015).

Moreover, the Duluth model, based on the patriarchal ideology, was originally designed just for heterosexual couples; however, it was subsequently applied to LGB perpetrators (although in the United States the groups, during the treatment, were often separated according to sexual orientation, even if the programs were mostly the same for both groups) (Price and Rosenbaum, 2009; Buttell and Cannon, 2015). This feminist psycho-educational approach is focused on re-education toward the development of more adaptive attitudes, improving communication proficiency, and ultimately eliminating violent behaviors (Buttell and Cannon, 2015). To the best of our knowledge, there are no studies to test the impact of such treatment on the LGB population (Stith et al., 2012) and the few researches on heterosexual population show limited positive effects (Babcock et al., 2004; Stith et al., 2012). Buttell and Cannon (2015) stated that scholars applying a post-structuralist feminist framework to IPV argued that a one-size-fits-all treatment model for IPV perpetrators (e.g., the Duluth model) should be replaced by culturally relevant and specific treatment options for LGB perpetrators. In their opinion, treatment interventions should address issues of sexism, homophobia, racism, and classism in order to address the ways society materially disadvantages some while privileging others (Buttell and Cannon, 2015).

Cannon et al. (2016) analyzed 3,246 questionnaire sent to directors of domestic violence perpetrator programs in the North American Domestic Violence Intervention Program Survey, in the United States and Canada. The results highlight that the most common approach to LGB batterers was a one-to-one approach instead of a group therapy, due to the difficulties for LGB people to express openly express themselves in heterosexual groups, two programs were projected for the LGB population.

# Cross-National/Cross-Cultural Differences

fpsyg-09-01506 August 17, 2018 Time: 18:40 # 9

Many interventions were developed in the North American context (Istar, 1996; Merrill and Wolfe, 2000; Dixon and Peterman, 2003; Lee and Utarti, 2003; Ristock and Timbang, 2005; Borne et al., 2007; Fountain and Skolnik, 2007; Herrmann and Turell, 2008; Price and Rosenbaum, 2009; Hines and Douglas, 2011; Dykstra et al., 2013; Armstrong et al., 2014; Buttell and Cannon, 2015; Quillin and Strickler, 2015), while a few existed in Canada (Senn and St.Pierre, 2010; Cannon et al., 2016; Barata et al., 2017) and Australia (Leonard et al., 2008; Jeffries and Kay, 2010). Some interventions were addressed to a specific ethnic group, such as Asians (Chung and Lee, 1999; Lee and Utarti, 2003; Cheung et al., 2009), or black people (Helfrich and Simpson, 2014). Moreover, IPV services where more accessible in urban centers where the LGB community was well developed and rooted than in rural areas (Jeffries and Kay, 2010; Ford et al., 2013). To the best of our knowledge, specific researches have addressed to IPV assessment/treatment for the LGB population in other countries.

# ACCESS TO SERVICES OFFERING HELP AND SUPPORT

Because of the impact of homophobia, homosexual and bisexual people may have a significantly more difficult time finding and receiving appropriate help than heterosexual ones, particularly when other variables such as income, ethnicity, and immigration status were held constant (Ard and Makadon, 2011; Barata et al., 2017).

Lesbian, gay, and bisexual victims of IPV access treatments through a wide range of help-giving resources, which can be distinguished into informal (family, friends, acquaintances) and formal resources (support groups, LGB community agencies, hotlines and shelters for IPV victims, medical health-care providers, and the criminal justice system). LGB victims of IPV were prone to seek help from informal resources (particularly friends) (Scherzer, 1998; Merrill and Wolfe, 2000; Turell, 2000), although there was a rather high percentage of people who turned to health care providers and family (Scherzer, 1998; Merrill and Wolfe, 2000; Turell, 2000); on the contrary, organizations specifically designed with the purpose of addressing IPV seemed to have the lowest utilization rates (Lanzerotti, 2006). In terms of the gender of the victim, it emerged that lesbian women had the tendency to seek help from all types of resources equally, while gay men were more prone to turn to the police to report victimizations (Cornell-Swanson and Turell, 2006; Senn and St.Pierre, 2010).

These results confirmed the need for specific interventions for LGB people, particularly considering that the health system offered low quality support, beginning from the fact that health professionals who assessed heterosexual female patients for IPV typically did not similarly screen lesbian or bisexual female patients or male patients of any sexual orientation in the same manner (Jeffries and Kay, 2010; O'Neal and Parry, 2015; Barata et al., 2017). McClennen et al. (2002) identified that a 7–33% of the victims evaluated the health system support as valid. Several studies highlighted that many interventions were perceived as unsatisfying because of homophobic (Tigert, 2001; Helfrich and Simpson, 2006, 2014) or superficial attitudes, denying the seriousness of the violence—"women are not as violent to one another" and "men can protect themselves" (Chung et al., 2008; Fonseca et al., 2010). These findings are consistent with Seelau and Seelau (2005) that considers perpetrators as more aggressive if the victim was a woman instead of a man. Male perpetrators were judged more blame-worthy than female perpetrators. Overall, male–female IPV was considered more dangerous than female–male, male–male, or female–female abuse. Significantly, the gender of the survivor, not sexual identity, was the most prominent factor in predicting witness response. In accordance with this, Arnocky and Vaillancourt (2014) work suggested that men, regardless of sexual identity, were less likely to recognize that they were being abused than women. To date, trainings on LGB IPV received by operators appear to be lacking, while the operators often believe to have an appropriate competence regarding heterosexual IPV (Senn and St.Pierre, 2010; Hancock et al., 2014).

# CONCLUSION

The literature on LGB IPV is recent and limited compared to the one on heterosexual IPV. However, a growing body of empirical research does exist, thereby offering important observations and considerations regarding LGB IPV. Previous studies primarily examined the prevalence of IPV in the homosexual and bisexual population (Turell, 2000; Messinger, 2011; Barrett and St.Pierre, 2013; Breiding et al., 2013), LGB specific features in IPV (Merrill and Wolfe, 2000; Balsam and Szymanski, 2005; Bartholomew et al., 2006; Carvalho et al., 2011; Edwards and Sylaska, 2013; Gill et al., 2013) and barriers to treatment (McClennen et al., 2002; Ard and Makadon, 2011). There are only a few publications on treatments and interventions for LGB IPV (Browning et al., 1991; McClennen et al., 2002; Dixon and Peterman, 2003; Ristock and Timbang, 2005; Buford et al., 2007; Fountain and Skolnik, 2007; Herrmann and Turell, 2008; Ard and Makadon, 2011; Quillin and Strickler, 2015). They can be classified into counseling interventions, particularly for victims (Dietz, 2002; Dixon and Peterman, 2003; Poorman et al., 2003; Buford et al., 2007; Franklin and Jin, 2016), and therapy: couple (Istar, 1996; Borne et al., 2007; Dykstra et al., 2013; Buttell and Cannon, 2015), group (Coleman, 2003; Ristock and Timbang, 2005; Buttell and Cannon, 2015), and for perpetrators (Babcock et al., 2004; Dykstra et al., 2013; Buttell and Cannon, 2015).

Despite the myth that IPV is only an issue in heterosexual relationships, its occurrence among LGB couples was demonstrated to be comparable to or higher than heterosexual cases (Messinger, 2011; Kelley et al., 2012; Barrett and St.Pierre, 2013; Breiding et al., 2013). While similarities between heterosexual and LGB IPV (such as general patterns, types, outcomes, cycle of violence and use of substances) were found (McLaughlin and Rozee, 2001; Buford et al., 2007;

Cain et al., 2008; Hequembourg et al., 2008), unique features and dynamics were present in LGB IPV, which were implicated in identifying and treating IPV among the community (Merrill and Wolfe, 2000; Carvalho et al., 2011; Bowen and Nowinsky, 2012; Gill et al., 2013).

Even though literature on LGB IPV is lacking in general, there is a need for research specifically on treatment (Dupont and Sokoloff, 2005). Results suggested that several obstacles prevent LGB people from getting help in case of IPV (Alhusen et al., 2010; O'Neal and Parry, 2015), heterosexism above all. IPV victims can be reluctant in seeking assistance, fearing discrimination (Giorgio, 2002; Helfrich and Simpson, 2006; Carvalho et al., 2011). Rarely a solution was offered to help LGB people in accessing treatment for IPV, and authors recommended modifications to standard treatments or programs (Calton et al., 2015; O'Neal and Parry, 2015). Studies showed that services and shelters were often unprepared to support IPV homosexual and bisexual victims (Buford et al., 2007; Cannon et al., 2016; Barata et al., 2017). In the United States, many emergency departments, shelters, agencies, and clinics had IPV advocacy programs; most of these programs historically failed in responding adequately to abuse in LGB groups (Brown and Groscup, 2009; Ford et al., 2013; Armstrong et al., 2014). The majority of the researches takes into consideration only North American services and programs existing in urban centers, while rural areas or other countries were not investigated (Jeffries and Kay, 2010; Ford et al., 2013). Comparing the few programs specializing in SSIPV treatment to traditional protocol, they were modified in assessing processes for sexual identity, in helping SSIPV victims in accessing the legal justice system, and in avoiding stigmatization (Merrill and Wolfe, 2000; Ristock and Timbang, 2005; Armstrong et al., 2014; Cannon et al., 2016). However, studies did include recommendations in order to focus on LGB-specific treatment. While many researchers recommended modified versions of IPV treatment, no one empirically studied whether LGB people benefit more from modified versions of treatment than standard treatments (Stith et al., 2012).

It is crucial to address an additional issue related to the cultural and social context: the fact that we found studies on treatment only in the North-American context indicates a lack of research in this field in other countries; however, it is possible that some studies were not present in international databases. The reviewed literature suggested the need of a psychological treatment designed on specific LGB necessities and finalized to

# REFERENCES


guarantee new useful resources and develop self-determination (Merrill and Wolfe, 2000; Calton et al., 2015; O'Neal and Parry, 2015). Intervention for LGB IPV victims and perpetrators should be part of an integrate and complete treatment plan that can involve couples or individual treatment but, in any case, that should be adapted to each specific situation. In line with such considerations, adequate training for mental health providers and standard guidelines for assessment and treatment may lead to more positive outcomes. Improvements should concern victims' well-being and satisfaction and treatment features, such as the durability of the treatment effects; moreover, a new approach may define an easier accessibility to services (Alhusen et al., 2010; Ard and Makadon, 2011; Banks and Fedewa, 2012). Since IPV appears to be an issue as common and serious in same-gender relationships as in heterosexual ones, policies and practices should update to guarantee the same degree of protection (Brown, 2008).

Because of the lack of program specialized in addressing SSIPV it would be important that the emerging IPV programs should provide outreach and educational services by cooperating with the community and offering several services, beginning from direct and physical resources such as shelters, food and clothing, transportation, financial and legal assistance, 24-h hotlines and individual and group therapy. Although traditional battered women's shelters can be recognized as a model for LGB agencies, some changes should be made: for example, a more inclusive language and a focus on experiences of individuals rather than gender, which can make LGB people more comfortable in disclosing abuse. IPV is still a partially unknown issue in the LGB community, which may minimize warning signs and this is why the LGB community needs to be specifically targeted for education regarding IPV and recognize its signs (Coleman, 2003; Dixon and Peterman, 2003; Dutton et al., 2009; Ard and Makadon, 2011; Bowen and Nowinsky, 2012; Calton et al., 2015; O'Neal and Parry, 2015; Cannon et al., 2016).

# AUTHOR CONTRIBUTIONS

GG and LR took overall responsibility for the creation of the frame used in this review and the selection of the papers. GG, AC, and EG searched for the articles discussed in the review. LR and PB supervised the entire work. All the authors were involved in the discussion, writing and revision of the manuscript, and approved the final version of the manuscript to be published.





**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 © 2018 Rollè, Giardina, Caldarera, Gerino and Brustia. 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.

# The Language of Ethics and Double Standards in the Affective and Sexual Socialization of Youth. Communicative Acts in the Family Environment as Protective or Risk Factors of Intimate Partner Violence

Oriol Rios-González <sup>1</sup> \*, Juan C. Peña Axt <sup>2</sup> , Elena Duque Sánchez 3,4 and Lena De Botton Fernández <sup>4</sup>

<sup>1</sup> Pedagogy, Universidad Rovira i Virgili, Tarragona, Spain, <sup>2</sup> Instituto de Estudios Sociales y Humanísticos IDESH, Universidad Autónoma de Chile, Santiago, Chile, <sup>3</sup> Theory and History of Education, University of Barcelona, Barcelona, Spain, <sup>4</sup> Sociology, University of Barcelona, Barcelona, Spain

#### Edited by:

Luca Rollè, Università degli Studi di Torino, Italy

#### Reviewed by:

Vincenzo Bochicchio, University of Calabria, Italy Chloe Rogers, University of Hertfordshire, United Kingdom

> \*Correspondence: Oriol Rios-González joseoriol.rios@urv.cat

#### Specialty section:

This article was submitted to Gender, Sex and Sexuality Studies, a section of the journal Frontiers in Sociology

> Received: 17 April 2018 Accepted: 15 June 2018 Published: 04 July 2018

#### Citation:

Rios-González O, Peña Axt JC, Duque Sánchez E and De Botton Fernández L (2018) The Language of Ethics and Double Standards in the Affective and Sexual Socialization of Youth. Communicative Acts in the Family Environment as Protective or Risk Factors of Intimate Partner Violence. Front. Sociol. 3:19. doi: 10.3389/fsoc.2018.00019 One of the explanatory factors for the perpetuation of intimate partner violence (IPV) is the socialization process. There is broad literature on the role of family in socializing gender roles and the influence on reproducing IPV. However, less research has been developed on the effects of communicative acts in the family environment as a protective or risk factor in front of IPV. This article presents sound evidence confirming the presence of language of ethics that is reproducing stereotyped models of attraction in heterosexual relationships, which empties good people of attractiveness and indirectly contributes to maintain the link between desire and aggressiveness. The language of ethics is characterized by integrating speech acts that exclusively include ethics and exclude desire when talking about egalitarian boys or men. To analyze this reality, a qualitative study has been conducted framed in the communicative methodology. This methodology has been recommended by the European Commission to conduct research on vulnerable groups and social inequalities, which has the aim to advance knowledge on social transformation. Drawing on this approach, three different data-collection techniques have been implemented: in-depth interviews, daily-life stories and focus groups. The fieldwork includes a sample of 52 young men and women between the ages of 18 and 23 from a vocational training high school, and 4 fathers and 4 mothers of some of these young people. The findings confirm the existence of a model of socialization that replicates family relations based on the maintenance of the double standards. Thus, mothers used to employ the language of ethics with their daughters fostering a controversial effect, that is, the latter prefer to start affective and sexual relationships with boys who are aggressive and not egalitarian. On the other hand, fathers used to employ language of desire with their sons stimulating the performance of chauvinist behaviors that denigrate women and girls.

Keywords: language of ethics, language of desire, communicative acts, family relations, socialization, intimate partner violence

**63**

# INTRODUCTION

Research on family and gender issues has analyzed intimate partner violence (IPV) reaching different conclusions (Teaster et al., 2006; Valls et al., 2008, 2016; Yount and Li, 2009; Martin et al., 2013; Medrano et al., 2017; Vidu et al., 2017). One of these conclusions concerns on the central role that socialization process has in the reproduction of this problem. The research presented in this article coincides in some points with these analyses but get new explanations on how communication established within the family context helps to perpetuate counter-productive socialization processes that could lead toward violent relationships. Thus, the main hypothesis that we start from is: Communicative acts settled in daily family interactions, including verbal and non-verbal language, are fostering the maintenance of a double-standard discourse in relation to young people's affective and sexual relationships. Hence, despite the broad knowledge about affective and sexual relationships in young generations, little research has examined how socialization on double-standards discourses is articulated in the interactions and communicative acts established between parents and teenagers. In the present study we will pay attention to the definition! that Berger and Luckmann (1991) established about socialization where primary and secondary processes are differentiated. Following these authors, primary socialization implies to externalize the individual's being into the social world and internalize it as an objective reality. On the other hand, secondary socialization implies also an internalization of institutional or institution-based "sub-worlds," such as the division of labor (Berger and Luckmann, 1991, p. 158). Additionally, we also consider the conceptualization of communicative acts understood as verbal and nonverbal language that daily influence people's actions, decisions and desires (Habermas, 1985). This article will deepen on family interactions aimed at showing how heterosexual girls and boys start conversations with their parents that have an impact on their preferences and decisions concerning sexual-affective relationships which are closely linked with IPV. This analysis provides new insights that help to comprehend the reasons for the persistence of this phenomena among youth. This article is comprised by four sections. The first part presents a literature review regarding previous research on family socialization and language, socialization of emotions, socialization of attractiveness and socialization of gender stereotypes and masculinity. The second part introduces the methodological paradigm and datacollection instruments used. In the third section, the findings are detailed, and finally, in the fourth section the main conclusions of the research are summarized.

The research on family socialization makes several contributions in referring to models of attractiveness and sexual-affective relationships, in this article we present four key themes on this line that have been identified throughout a wide literature review. Firstly, there are a set of studies that go deep into the interactions and social meanings that are expressed through language and how these constitute the main form of socialization in the family environment. Secondly, there are analyses based on the emotions defined within the family environment which are constructed through language. Thirdly, there is research that stresses how socialization framed on the family environment promotes specific models of attractiveness. Lastly, there are contributions that provide evidence on the effects of language use on the definition of gender stereotypes and masculinity models.

Regarding the block of studies which pays attention on interactions, social meanings and language, it is important to mention again the analysis conducted by Berger and Luckmann (1991). They underscore the relevance of family environment in the socialization process starting from the premise that is during childhood, and through interactions with family members, that a person learns how to become part of society. This process happens through children's identification with others that makes them accept certain roles and attitudes. Other key social theorists such as Parsons and Bales (1955) put family as the first socializing institution in industrial societies. They highlighted that nuclear family, and particularly mothers, was the social system that guarantees the proper internalization of social life. Along these lines, Schutz (1967) stated that the lifeworld is the world in which the experience of others constitutes a fundamental element in the formation of self-perceptions. Therefore, the lifeworld is inter-subjective and culturally shared through symbols such as language. Socialization, then, is mainly a social learning which implies the acquisition of structures, behaviors and tastes; and language has a lead role in consolidating that process.

Later, Habermas (1985) goes beyond and argued that the process of rationalization of the lifeworld creates more egalitarian patterns of relationships which are changing the socialization process inside the family environment. In fact, this change on the vision of the role that nuclear family has in modern societies began when this family model started to be considered as the basis of a decadent society (Beck-Gernsheim, 2002). Hence, from mid of 1950's to nowadays, family's functions and forms have been modified, however, and despite of these changes, research shows that family continues having a fundamental role in children's socialization (Mitchell, 2010; Rollins and Hunter, 2013; Höppner, 2017).

Regarding the group of analyses which deepen on the socialization of emotions, research indicates that the formation of emotions in individuals is established mainly within the family (Garner et al., 1997; Elster, 2007; Hunter et al., 2011; Mandara et al., 2012; Shaffer et al., 2012). For instance, Elster (2007) affirms that emotions are based on beliefs and these are defined in socialization processes, so they are transmitted. He indicates that family contributes to inform children about the meaning of these emotions and this helps the latter to properly understand their feelings. Other authors, such as Hunter et al. (2011) and Shaffer et al. (2012) go further into explaining how these emotions are socialized in the family environment. Shaffer et al. (2012) shows that there is a direct influence between emotional development and the kind of family to which the individuals belong. Families shape emotions according to certain risk factors in relation to housing and socio-economic status and these issues directly affect children's feelings (Shaffer et al., 2012). In the same vein, Hunter et al. (2011) focus on how children's emotions are directly related to their parents' emotions, finding evidence that parents' strategies influence young people's emotional beliefs. This research also maintains that there are socializing differences between fathers and mothers; therefore, how young men and women develop their personality depends directly on how parents interact with their children. Consequently, children whose mothers participate more in emotional socialization have more capacity for emotional regulation than those whose mothers are less present in that process. Finally, on this group of studies focused on emotions and similarly than the above mentioned analyses, there are researches which conclude that this emotional socialization within family environment leads to reproduce gender stereotypes, for example the promotion of gender identities based on the distinction between a tougher boy or a fragile girl (Garner et al., 1997; Botello, 2017). Mandara et al. (2012), who performed a research with African–American mothers, also illustrate that stereotyping process, finding that those mothers who take care of daughters are more relaxed and less negative than those who take care of sons. Drawing on a psychological perspective, Brown (2011) deepens on this regard exploring how prejudice is constructed during childhood. He concluded that social behaviors are shaped on these tendencies established throughout individual personality, but insisting on the fact that attitudes and actions are also influenced by the social groups that each person belongs.

The third section of this literature review is focused on the socialization of attractiveness and how language is a key element on this regard. There is an important amount of analyses that stress the fact that models of attractiveness are socially constructed and this becomes a procedure which is influencing people's choices on relation to sexual and affective relationships (Valls et al., 2008; Díez-Palomar et al., 2014; Gomez, 2015; Puigvert, 2016). This group of researches starts from a conceptualization of models of attractiveness that understand them as social patterns which provide of desire or valorization particular types of masculinities and feminities (Padrós, 2012). Therefore, every-day interaction spaces, like the ones established in family, contribute to foster or reject certain models of attractiveness, so these spaces become very important at early ages. Recent studies on this line illustrate an alarming problem concerning the existence of models of attractiveness which are connected to violent behaviors, this means that young people and adolescents are being socialized on attraction toward violence and this can drive them to toxic relationships marked by IPV (Valls et al., 2008). Valls et al. (2008) confirmed that issue and they discovered that this link between aggressiveness and attractiveness is due a chauvinist socialization process which promotes desire toward masculine models that are dominant and violent.

The influence of language in the abovementioned process is strongly important in people, but particularly in teenagers, because it can associate beauty with ethical or non-ethical elements (Ríos and Christou, 2010). Thus, research differentiates between language of ethics and language of desire in order to explain the types of languages that people employ to promote one thing or another. Accordingly, it is quite common to use the language of desire to foster desire and admiration for dominant traditional males, and the language of ethics to talk about egalitarian males (Castro and Mara, 2014; Schubert and Valls-Carol, 2015). As a consequence of this common practice the reproduction of a double-standards scheme is perpetuated (Díez-Palomar et al., 2014). Double standards are understood in that case as the persistence of a desire toward men who have power but not ethical values, and, on the other hand, the maintenance of a feeling of friendship toward men that have egalitarian and solidary attitudes but without power positions (Gomez, 2015). McCarthy and Casey (2008) coincide with this analysis and they indicate an attraction toward violence in young cohorts. These authors also pay attention on the role that family has on this link and they argued that some young people feel their relationship with their parents is weakening, so they seek to fill this emotional void with partners associated with violence. Thus, one of the most relevant conclusions of this research is that many young people separate passionate attraction and non-passionate love, linking the former with violence and the latter with stability.

Finally, the last part of this literature review refers to gender socialization and the construction of masculinity. The analyses on this field are centered mainly on the study of how hegemonic gender models are socialized and reproduced (Kimmel, 2000; Connell, 2005; Javaid, 2017). Concerning the study of masculine gender models, research has especially highlighted the perpetuation of a traditional and hegemonic masculinity model through cultural dominance and violence (Connell, 2005; Shumka et al., 2017). From that position, the definition of two central gender models has been conceptualized from the studies of Connell et al. (1985): emphasized femininity and hegemonic masculinity. Both arise from the definition of hegemony provided by Gramsci and refer to cultural practices that have been maintained as central in gender socialization. Hence, hegemonic masculinity is understood as this model of masculinity that becomes predominant excluding other models to be successful or more visible.

Research that pays attention to this matter also highlights the role that communicative acts performed in the family setting has in the shaping of hegemonic masculinity as a successful model. For instance, Schrock and Schwalbe (2009) show that men who principally define themselves as egalitarian because share domestic chores sometimes carry out a series of communicative acts that reinforce hegemonic masculinity. In this way, through acts such as showing disdain for the tasks carried out by their female partners, men who are apparently egalitarian perform communicative acts of dominance, thereby reproducing gender inequalities. Along the same lines, Hughey (2011) notes that in the United States of America, chauvinist and racist speech continues to be reproduced by some white men in the intimacy of their homes. This author maintains that there are many white men who publicly show tolerance with pro-feminist and anti-racist speech but that in the intimacy of their homes they reproduce speech acts that encourages gender and race inequality. These two studies show that the family environment, in some cases, reproduces gender stereotypes where the traditional model of masculinity is promoted. Then, research framed on men's studies clarify the distinction between traditional and egalitarian masculinities that help to comprehend the reproduction of these gender disparities (Flecha et al., 2013; Castro and Mara, 2014). This distinction is based on the conceptualization of three ideal-types: Dominant Traditional Masculinity (DTM), Oppressed Traditional Masculinity (OTM), and New Alternative Masculinities (NAM). DTM perpetuates chauvinism and IPV, OTM is not violent but can act reproducing chauvinism and double-standards, and contrarily NAM goes beyond and these are the egalitarian men who are neither violent nor chauvinist, overcoming double standards practices. The review on masculinities carried out by Bridges and Pascoe (2014) also pays attention on the emergence of a "hybrid masculinity" which distances from traditional models of masculinity. This typology of masculinity combines toughness and tenderness and, in spite of this alternative gender performance, is not understood as a profound challenge to hegemonic masculinity. Contrarily, these hybrid masculinities are perceived as a contemporary interpretation of the existing gender and sexual inequalities. In a similar vein, Connell (2012), in her reformulation of hegemonic masculinity in the globalization era, realizes an important distinction between violent and non-violent hegemonic masculinities. She affirms that there are men who perform chauvinist behaviors and practices but not being violent, on the other hand she also maintains there are men who portray this hegemonic masculinity being violent and chauvinistic at the same time. All of these studies on the influence of socialization on people's subjectivity and on the role of language in that process, particularly in the shaping of attractiveness or gender stereotypes, highlight the relevant role of family for understanding the mechanisms that reconfigure people's identity and behaviors. Nevertheless, there is a gap in the research that examines how the language employed in the family environment, especially by parents, influences young people's attraction patterns and the reproduction of the double standards. This article will provide data on all these aspects.

# MATERIALS AND METHODS

#### Methodological Paradigm

To gather the evidence on the influence of language employed in the family environment in the socialization process of youth, a qualitative methodology, that took the communicative perspective into account, was employed (Gómez et al., 2011). The main characteristic of research that adopts a communicative perspective is that subjects participate in the research with a horizontal relationship with the researchers. This egalitarian approach is established at the beginning of the research process, where subjects discuss key aspects such as the design of the datacollection instruments and the conclusions reached through the field work analysis (Flecha and Gómez, 2004). In the present study, an advisory board formed by young people, parents, and people involved in the struggle against gender violence was created to satisfy this communicative premise.

# Study Design and Sample Description

The research was based on a case study carried out in a vocational training school in Barcelona (Spain), which was selected because of its social and cultural diversity that is quite representative of the socio-demographic reality of the city. Thus, in this school there are students and families which come from Latin-America and North-African countries as well as from Spain. They also come from different socio-economic backgrounds, but the school is mainly attended by students of middle class, low-middle class, and working class. The sample (n = 60) includes young people, heterosexual men and women ranging from 18 to 23 years old that are attending vocational training, particularly who are registered in courses of personal image, aesthetics and beauty, and telecommunications. The sample is completed with eight mothers and fathers of the young interviewees. In this regard, the group of young people who were involved in the field work were selected discussing with the principal of the school its appropriateness, in particular with the objective to guarantee the criteria of socio-cultural diversity mentioned above. Furthermore, only those parents who accepted to be interviewed were included in the sample.

Three different data collection instruments were employed: life stories, in-depth interviews, and focus groups. Life-stories were conducted with the objective of deepening on specific moments of students' life. In fact, the nature of this instrument helped creating an atmosphere where young people openly and sincerely explained dialogues and interactions with their parents. The objective of this instrument is not to carry out a biography but to construct a reflexive narration about subjects' daily life in order to deepen on their present, past and future expectations. This is an instrument that allows researchers to identify how barriers faced in subjects' life are overcome. To complement this data, the research team decided to perform focus groups with young people who were already friends and had enough confidence to expound their family relations in public. Lastly, in-depth interviews addressed to mothers and fathers were developed in order to consider parents' perspective that could be contrasted with young people's visions. More detailed information on these data-collection techniques is presented in **Table 1**.

# Data Analysis

The analysis of the information was carried out aimed at obtaining knowledge which would be useful to understand the


<sup>a</sup>The interviews were applied to father and mother simultaneously. This allowed us to have much more relevant information regarding the interactions that happen within the intimacy of the family.

<sup>b</sup>Communicational discussion groups applied are: two coed groups of communicational discussion, to communicational discussion groups of women and two communicational discussion groups of men.

reproduction of the double-standards discourse. Thus, the data analysis was focused on identifying communicative acts which maintain this discourse paying particular attention on how the language of desire and the language of ethics are used by young and adult people (parents). Drawing on these premises, all the data collection instruments employed were verbatim transcribed and the quotes emerging were selected to respond the research hypothesis: Communicative acts settled in daily family interactions, including verbal and non-verbal language, are fostering the maintenance of a double-standard discourse in relation to young people's affective and sexual relationships.

Later, we categorized the information considering three main criteria: (a) how communicative acts are used to reproduce the double standards discourse (which kind of language is employed—ethics or desire); (b) who perform the communicative acts (parents or young people); and (c) which implications these communicative acts have in young people's decisions and interests regarding their sexual and affective relationships. The findings obtained were widely discussed with the advisory board that validate their appropriateness.

# RESULTS

Drawing on the previously presented analytical scope, four key issues emerge from our analysis which illustrate these elements that perpetuate the double standards discourse in family relations. First, the reproduction of the language of ethics to speak about boys in the family environment is discussed. Second, evidence on mothers as the main actors who employ the language of ethics is presented. Third, insights are provided about fathers involved in discussions where the double standards about boys' sexual and affective relationships are reproduced. Lastly, the consequences that the use of the language of ethics in the family environment have on young people are explained.

# Language of Ethics in the Family Environment

Among the families who participated in the research, the way that attractiveness toward alternative models of masculinity is promoted is highly important and mainly occurs through the use of the language of ethics. The language of ethics is used focusing on men who are considered morally appropriate to maintain sexual-affective relationships. All participants are able to clearly define what characteristics men should have to become a successful boyfriend from relatives' point of view, particularly incorporating aspects that are "ethically" highly valued. However, it can be observed that desire is not discussed when people talk about these boys, as shown in the following quote, where a mother expresses her desire that her daughter relates with a "formal" kind of man. The verbal language she uses clearly differentiates between two kinds of men, one forged in egalitarian values and another branded by completely contrary values.

Mother: "I always worry that she doesn't just go out with anyone, that the guys that she is with should be nice. That's why I always tell her that she has to look for guys that are worth it, not bums." (In-depth interviews, family members)

The research also shows that the interactions in this kind of language have a contrary effect and it is precisely socializing heterosexual women to develop an attraction toward men who may cause them trouble. The following sentence confirms this kind of interaction that is mainly generated between the two generations. Here the language that is linked to kindness, but lack of desire, is used to describe these "good guys" but generates an evident rejection.

Mother: I remember when Esteban used to come here. He came every day to walk her to school, but she never paid attention to him.

Father: But he has always been in love with her, since they were little.

Mother: But I tell you, never, and the guy didn't lack for trying. I told her that he's a good guy, that he looked nice and things like that. But she didn't like him. At the end, the guy just got bored and found himself a girlfriend. But I think that if she tells him to come back, he'll come running.

#### **I: And when you spoke to her about him, what did she tell you?**

Mother: That I was right, that he was very good, but she didn't like him. (In-depth interviews, family members)

This dialogue between the researcher and the interviewee shows how the mother sees in the boy a good and proper man for her daughter, that is, ethically well valued. However, the daughter does not find him attractive. At the same time, this quote exemplifies the fact that the mother is the person in the family who predominantly talks about these issues with her children. The following lines reveal more details about this issue.

Mothers and the Language of Ethics

The thing is that my dad doesn't care about these issues. The one who does is my mother.

#### **I: And what did she tell you?**

That I should be careful, I should be careful with boys who are too aggressive because they can drive you on the wrong track. (Caro, 19 years, life stories)

As shown in the above quote, girls sometimes choose to talk about their intimate life with their mothers. In this sense, the field work shows that girls feel much more open to discuss the topic of love and relationships with relatives of their same gender, in this case with their mothers. The following quote is extracted from an interview with a father and a mother, where the aforementioned tendency is confirmed.

Father: I don't talk about those things with her.

Mother: I am the one who speaks about boyfriends with her; she's always telling me stuff. Every day I ask her about it. Not him; he tells her that he doesn't want to talk about those topics with her. (In-depth interviews, family member)

These interactions happen when comfortable communication facilitates intimate discussion, for example, when mothers try to approach and get to know their daughters better. Likewise, one of the interviewed mothers explains how she maintains certain dialogue and communication strategies with her daughter to keep her away from men who she, as a mother, does not consider adequate for her daughter. These spaces of communication and intimacy allow interactions that link formality and equality through the language of ethics. In the following quote, this type of interaction is confirmed when a mother refers to the man she considers adequate for her daughter. She does not use a language of desire to describe him; on the contrary, she talks about him employing words linked to kindness.

Mother: I am always telling her if somebody is no good for her. I tell her that she has to think about the future, that she should think if that guy she likes has a future. In the beginning, she always gets angry, but when she meets another guy, she tells me. (In-depth interviews, family members)

The previous quotes also illustrate how the mother makes direct use of the language of ethics to socialize her daughter toward a model of man, ruling out the man who she does not consider satisfactory for her daughter. Talking in this way has the objective of influencing her daughter's choice but utilizing adjectives that are fundamentally based on ethical issues. For instance, as shown in the next quotation, adjectives full of ethical connotations are used. The first one refers to boys as "studious" and the second one as "normal," but none of them have elements that connect with young people's conceptions of attractiveness and desire.

"My mother talks to me more like a friend. My mother is always telling me that I should look at normal boys that are not too lazy, so if tomorrow I can be with them, then it should be all right. That they don't treat me bad or that I have to support them and things like that. That they study, normal, not dumb, but normal". (Cris, 18 years, life stories)

In contrast, the language used by fathers to address attraction and relationships is radically different. Mothers' words are more connected with ethics, while fathers use language, mostly when speaking with boys that is not ethically constructed. This topic will be analyzed more in depth in the next section, where different elements that reproduce the double standards will be described.

# Fathers and Double Standards

In this study, fathers' involvement in socialization is different from that of mothers. Fathers foster double standards when interacting with their children, but only with boys. In this sense, the fathers who participated in this research have different attitudes depending on the gender of their children. The evidence gathered shows the differences in communication between fathers and their boys and girls in similar circumstances. Their language generates a reproduction of dominant traditional masculinity, mainly through the use of words that compare girls with objects, as shown in the following quote.

"My father tells my twin brother that he has to screw all the girls he can. But he doesn't tell me anything, and I can't tell him anything because I am his little girl" (Cris, 18 years, life stories)

This kind of language is disrespectful toward women; it separates itself from ethics and incorporates a sexual component. This component does not exist in the language used by mothers, who, as we have previously witnessed, are those who are most involved in discussions of attraction based on ethical issues. However, in the fathers' communicative acts, there are no transformative elements; on the contrary, they are reproducing elements that foster socialization based on double standards. In the following quote, we can observe this reality from the perspective of a young interviewee:

Yes, my father is a bit old-fashioned; he tells me to do the things that he used to do. Chauvinist-based things. For example, he tells me that I should just screw girls. (Xavi, 18 years, life stories)

Ultimately, as we have shown in previous quotes, fathers' and often mothers' language use contribute to the perpetuation of double standards. This does not help encourage alternative affective and sexual relationships in the next generations; on the contrary, a conservative conception is reinforced. The next quote exemplifies this reality: "My father always says. . . As long as she is hot, that's enough" (Adam, 20 years, life stories). In addition, this kind of language is also socializing boys into chauvinist values as it stated in the next quote where a homophobic statement is expressed. In that case, as Adam said, a father does not accept his son's homosexuality because will not be able to respond to his ideal of boy succeeding with girls: "His father is the one who says 'fuck', the only son that I have and that he is fagot" (Adam, 20 years, life stories).Therefore, consequences of this kind of language use with young generations must be widely explored, and in the next section, the thorough analysis carried out in this study will shed light on this issue.

#### Consequences

Two main consequences have been identified through the analysis of language use: (a) the reproduction of double standards—but paying attention in this case to the attractiveness of masculine dominant models—and (b) the lack of attractiveness of young males who have values but are considered only as friends and not as prospective partners in a sexual and affective relationship. Regarding double standards, several of the interviewed young women maintained that there are two kinds of men, some to have fun with and others to be boyfriends. These notions are reflected through the contradictions manifested by the interviewed girls at the time of choosing a partner. Influenced by the interactions they have had within their family environment, they end up making radically different choices regarding boys. This situation is reflected in the following quote, where the interviewee defines which characteristics a guy should have to capture her interest: "To go out with those that are bastards, that they make you laugh and such. And for a boyfriend, one that understands you, that is sincere". (Cris, 18 years, life stories).

Thus, despite the attempts of families to socialize their daughters into relationships based on egalitarian values, the language of ethics that parents employ is not able to change or socialize their attractiveness toward alternative male models. In fact, they have the opposite effect by encouraging the attractiveness toward violence and even justifying it.

Man, we like him to be like that, aggressive, because you know that whatever happens, he can defend you or they are going to respect you as well because you are his girlfriend. And nobody's going to say anything to you. (Paola, 19 years, life stories).

The second consequence is removing attractiveness from men of masculinity associated with egalitarian values. In this way, interactions within the family environment suggest the complete absence of attractiveness of egalitarian young men. The use of the language of ethics impacts the attractiveness of a model of men that their parents consider adequate for their daughters. This is known by women who see men whom their parents consider good for them directly as weak men or, as they call them, "mama's boys." This is reflected in the next quotation, where a young interviewee is asked about the attitude of her classmates in her high school, and she describes good men as weak: "You know what happens? When you are good, they tell them they are mama's boys." (Paola, 19 years, life stories). These "mama boys" do not generate any kind of desire, in fact their goodness is an explanatory element of this lack of desire, because goodness and attractiveness are separated: "there was a girl that said to me: I don't like him because he is too good and this doesn't turn me on, he isn't hot for me" (Lorena, 20 years, life stories).

These kinds of interactions, in which attractiveness is completely removed from young men with egalitarian values or well evaluated from an ethical point of view, is not an isolated incident. On the contrary, it is a situation in which the interactions between young people make evident the attractiveness toward violent masculine models. In the following quote, it is observed how a boy with values is questioned regarding his ability to be with a girl who, in the opinion of others, is much prettier.

I don't know, I have a friend who, let's see, is not good looking, sort of, and he has a good heart and. . . I don't know. And sure, he is with a really good looking girl and people tell her: How can you be with him? (Lorena, 20 years, life stories)

Consequently, it does not matter what values boys have; girls simply do not perceive these boys as attractive, and that is reason enough to question their relationships. As noted, emotions and attractiveness are socialized; therefore, these kind of interactions are not more than a product of socialization processes, in which families have an important role.

# DISCUSSION

Drawing on the conclusions of the previous analyses collected in the literature review, there are several elements that contribute to understand the influence of socialization in the shaping of models of attractiveness and gender stereotypes. In addition, in those analyses the role that family environment and language have in these processes is evidenced. For instance, the literature helps reveal how conceptions of attractiveness are defined through daily interactions in different socializing spaces like family (Duque, 2006; Urpí and Naval, 2006; McCarthy and Casey, 2008; Gomez, 2015). Similarly, it also helps to comprehend how the construction of gender identity and people's emotions perpetuate affective and sexual relationships marked by the attraction toward violence (Kimmel, 2000; Connell, 2005; Schrock and Schwalbe, 2009; Hughey, 2011; Hunter et al., 2011; Shaffer et al., 2012). However, all these studies lack an explanation of the families' socializing role on young people's attraction patterns and the perpetuation of a double standard discourse.

In this research we try to respond this gap starting from the hypothesis: communicative acts settled in daily family interactions, including verbal and non-verbal language, are fostering the maintenance of a double-standard discourse in relation to young people's affective and sexual relationships. In this regard, we have collected evidences on how young people, particularly young girls, choose bad guys for their initial sexual or affective relationships and how these choices make them more vulnerable to suffer IPV. This last effect is an issue which has been widely explored by previous research (Bukowski et al., 2000) making visible how heterosexual girls who desire this typology of guys are more likely to have abusive dating or abusive marital relations. In fact, research has also demonstrated that the existence of a socialization process that links dominance and attractiveness is an important explanatory factor of IPV in teenagers (Valls et al., 2008). However, present investigation goes beyond these analyses and illustrates how girls' choices in their affective and sexual relationships are conditioned by the interactions and the language used within the family environment. These findings also illustrate that this language is centered on ethics and consequently in double standards. Thus, parents, especially mothers, used to perform a language of ethics with their daughters trying to promote egalitarian masculine models although they reach a controversial impact and finally young girls choose bad boys. On the other hand, research also shows how fathers, employing a language of desire with their sons, reproduce chauvinist and double standards discourses that imply maintaining traditional schemes on young people's sexual and affective interests.

Although previous research already identified the impact of communicative acts to favor attraction toward violence and the reproduction of double standards (Castro and Mara, 2014; Gomez, 2015), there are less analyses focused on how family relations could interfere on this process. Therefore findings presented here give new arguments about what interactions and what kind of language maintain these exclusionary dynamics in the family environment. Henceforth, to continue working on this line it is highly necessary to explore the mechanisms of constructing an alternative language in family relations, which would be based on desire and reject traditional and violent relationships as well as aggressive models of attractiveness. Beck (1992); Giddens (1994), and Beck-Gernsheim (2002) insist that reflexive modernity offers opportunities to reach this objective, to de-monopolize expert knowledge, to create deep revolutions in family intimacy and social movements. This modernity is characterized by a reformulation of subjects' personal relationships because more opportunities to establish an egalitarian dialogue, based on validity claims (Habermas, 1985), are settled. Hence, in current societies families and educational organizations are increasingly promoting interventions based on this constructive dialogue which are providing relevant knowledge to children and teenager for their choices in terms of sexual and affective relationships (Soler, 2017).

The results described in this article encompass many of these elements, meaning that their objective is to have a social impact on overcoming negative choices that young people take (Flecha et al., 2015; Reale et al., 2017). In short, considering all these elements, it can be stated that family relations and discussions can play a fundamental role in preventing IPV because they can position themselves as protagonists of a transformation in the socialization of attractiveness through linking the language of ethics with the language of desire.

#### ETHICS STATEMENT

The research conducted for the elaboration of this article followed the scientific and ethical procedures defined by the EU's Charter of Fundamental Rights and the UNESCO Universal

#### REFERENCES


Declaration of Human Rights. The ethical standards of the investigation are taking into account the recommendations established in the field of Social Sciences and Humanities research of Horizon 2020. In this regard, informed consent forms were distributed among young and adult people involved in the study according the recommendations of the Ethical Regulations of the University of Barcelona where the study was conducted. These informed consent forms were properly responded and returned by students and parents of the students. Finally, it is also important to mention that research and its data collection instruments where validated by the Clinical Research Ethics Committee of the Bellvitge Hospital (Barcelona, Spain).

# AUTHOR CONTRIBUTIONS

OR-G and LD contributed with the analysis of the field work and the elaboration of Results and Discussion sections. JP designed the investigation, conducted the data collection techniques and worked on the stated of the art section. ED participated in the elaboration and review of the state of the art of the article.

# ACKNOWLEDGMENTS

This article is supported by the competitive research project: Uni4Freedom. Violències per raó d'orientació sexual, identitat o expressió de gènere a les universitats catalanes funded by the Programme Recercaixa (Caixa Bank).


families of depressed and healthy adolescents. J. Youth Adolesc. 40, 428–441. doi: 10.1007/s10964-010-9545-2


**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 © 2018 Rios-González, Peña Axt, Duque Sánchez and De Botton Fernández. 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.

# When Violence Can Appear With Different Male Partners: Identification of Resilient and Non-resilient Women in the European Union

#### Juan Herrero<sup>1</sup> \*, Pep Vivas<sup>2</sup> , Andrea Torres<sup>1</sup> and Francisco J. Rodríguez<sup>1</sup>

<sup>1</sup> Department of Psychology, University of Oviedo, Oviedo, Spain, <sup>2</sup> Department of Psychology, Universitat Oberta de Catalunya, Barcelona, Spain

#### Edited by:

Luca Rollè, Università degli Studi di Torino, Italy

#### Reviewed by:

Paola Miano, Università degli Studi di Palermo, Italy Masoumeh Dejman, Johns Hopkins University, United States

> \*Correspondence: Juan Herrero olaizola@uniovi.es

#### Specialty section:

This article was submitted to Gender, Sex and Sexuality Studies, a section of the journal Frontiers in Psychology

> Received: 10 April 2018 Accepted: 15 May 2018 Published: 01 June 2018

#### Citation:

Herrero J, Vivas P, Torres A and Rodríguez FJ (2018) When Violence Can Appear With Different Male Partners: Identification of Resilient and Non-resilient Women in the European Union. Front. Psychol. 9:877. doi: 10.3389/fpsyg.2018.00877 Introduction: Little scholarly attention has been paid to the analysis of the history of intimate partner violence (IPV) against women with different male partners and how it could be related to levels of IPV with the current male partner. From this point of view, been a victim of IPV could increase the vulnerability of women and, therefore, exert a negative influence on the selection of partners over time, thus increasing the odds of potentially mating with abusive male partners. Alternatively, for some women victims of IPV in previous relationships, there may be additional resources that reduce their vulnerability to victimization by new partners.

Methodology: The present study analyzes levels of IPV in different partners of 2376 heterosexual women from the 28 countries of the European Union living together as a couple who had previously lived with a different male partner.

Analysis/Discussion: Multilevel regression results indicated that resilient women were younger, more satisfied with household income, and were involved in shorter relationships. As for their previous levels of victimization, they scored lower on child abuse and non-partner adult victimization. Also, their levels of victimization from previous partners were the same as those of the non-resilient women, with the exception of physical IPV victimization where resilient women scored higher than non-resilient women. Resilient women also informed the interviewer to have ended the abusive relationship because of the violence to a greater extent than non-resilient women and seemed to suffer fewer psychological difficulties due to previous violent relationships. Finally, countries scoring higher on human development index (HDI) showed a larger proportion of resilient women.

Conclusion: Resilient women are mostly characterized by fewer psychological difficulties and lower frequency of adverse situations (in childhood or in adulthood) when compared to non-resilient women. Although resilient women reported a higher

**72**

physical IPV, they nevertheless show fewer psychological sequelae and a greater ability to end abusive relationships. In addition, the human development of the countries in which they live also seems to reinforce their resilience, which suggests combining intervention policies at the individual and contextual levels.

Keywords: intimate partner violence, resiliency, European Union, multilevel, child abuse, adult victimization, victimization by multiple male partners

# INTRODUCTION

Research on intimate partner violence (IPV) against women has provided abundant empirical evidence on its main risk factors. However, less scholarly attention has been paid to the analysis of the history of IPV with different male partners and how it could be related to levels of IPV with the current male partner. Using data from 2376 women of the European Union that had suffered IPV from previous partners and were currently living with a new partner, we analyzed the characteristics of those women who informed of no IPV with their current partner as compared to those women who reported IPV with both previous and current partners.

Despite the growing interest that the study of IPV has generated in the last decades, a minor research effort has been dedicated to the study of the relationship between current IPV and the previous history of IPV. According to a recent review of studies conducted by Ørke et al. (2018) on risk for victimization of IPV by multiple partners the most striking finding was the scarcity of studies on this topic. Improving our understanding of why some women are victimized by multiple partners while other victims seem to be able to create new intimate relationships free of IPV is a promising field of research.

Although a percentage of women are victimized again by different partners (Bybee and Sullivan, 2002; Cattaneo and Goodman, 2005; Stein et al., 2016), some victims of IPV seem to be resilient to the negative effects of abuse (DuMont et al., 2007). This resiliency might be viewed as the ability to achieve good developmental outcomes while experiencing negative circumstances that pose a risk to normal development (Masten, 1994). Among those women who have suffered IPV with previous partners, those who now maintain relationships free of IPV with their current partners can be considered resilient women to the IPV. This invites to deepen the analysis of the characteristics of resilient women. The literature on this topic is very scarce and the available evidence does not yet provide a clear picture of the risk and protective factors of IPV revictimization by different partners. We review the empirical evidence available on this topic.

One of the aspects in which there seems to be greater agreement among researchers is that adverse experiences during childhood can have long-term effects in adult relationships. There is empirical evidence coming from both retrospective and prospective studies that victims of child abuse are somewhat predisposed to also be victims of IPV in their adulthood (see Herrero et al., 2018 for an analysis). As Ørke et al. (2018) have noted in their review of studies, women with IPV by multiple partners use to be exposed to more types of childhood violence and sexual abuse than women exposed to IPV by one partner. According to this, adverse experiences during childhood not only would predict higher IPV levels in adulthood but also a greater propensity to be victimized by different male partners.

Another set of IPV revictimization risk factors includes the type, frequency, and severity of abuse suffered in previous intimate relationships (Kuijpers et al., 2011). There are studies that have found that victims of more severe IPV tend to be victimized in other intimate relationships (Testa et al., 2003; Cole et al., 2008). Kuijpers et al. (2011) have used Foa et al. (2000) model to explain this association: partner violence causes psychological difficulties that, in turn, put women at greater risk of revictimization by hindering the victim's ability to curtail future violence. According to this, women who experience more severe IPV, as well as higher levels of psychological difficulties, become more vulnerable, thus increasing the odds of being victimized again by new partners.

Having been victimized in adulthood by people other than the male partner or the ex-partner has also been linked to IPV revictimization of women (Nishith et al., 2000; Stein et al., 2016). There are studies that suggest that some IPV victimization may arise from previous victimization experiences that cause interpersonal difficulties with, in turn, increase the risk of revictimization (Cole et al., 2008). Although most of this research is based on victims of child abuse (see Herrero et al., 2018 for an analysis), it is also possible than non-victims of abuse might also generate this type of social environment (Cole et al., 2008).

Staying in an abusive relationship has been linked to an increase in IPV (Fleury et al., 2000). Ending the relationship with the abuser may be difficult for the victim (DeKeseredy et al., 2017; Edwards et al., 2018) – if, for example, an increase in violence is anticipated as a consequence. As it has been pointed out, however, the end of the relationship may be a protective factor for future episodes of IPV (Grasley et al., 1999) that has been related to both post-traumatic growth and an increase in psychosocial resources (Senter and Caldwell, 2002; Cobb et al., 2006). The literature on victimization by multiple partners has rarely paid attention to this fact and little is known about the potential effect that ending a violent relationship could have on the creation of new relationships free of IPV. Those women who break their violent relationship can be empowered to create other intimate relationships free of violence, thus reducing the likelihood of IPV with new partners. As a result, breaking the violent relationship would be a protective factor while staying in the violent relationship would be a risk factor for victimization by multiple partners. In relation to this, the presence of children who are witnesses of the IPV could be an important factor in explaining the decisions of the victims of IPV on the end of the

violent relationship and also on their resilience to the IPV. As Rhodes et al. (2010) have suggested, mothers may wish to protect their children from harmful effects of violence but may want to keep the family together, thus avoiding any instability caused by legal system involvement. There is no empirical evidence that one or the other decision is more related to the resilience of women, despite the fact that negative consequences of the continued exposure of children to episodes of IPV have been recognized in the literature (Bogat et al., 2006).

Another set of risk factors identified by Ørke et al. (2018) in their review of studies is the characteristics of women and/or the relationship. Although the evidence is not conclusive, it seems that the youngest women are more at risk of being victimized by multiple partners (Testa et al., 2003; Alexander, 2009). Other sociodemographic characteristics such as income, education, ethnicity, or unemployment, seem to be unrelated to victimization by multiple partners (see Ørke et al., 2018) although there is empirical evidence of their relationships with IPV (Herrero et al., 2016, 2017b). Length of the relationship with a new partner seems also to be positively related to IPV with a new partner (Cole et al., 2008).

While the aspects related to victims and their relationships, although in a limited way, have received the attention of researchers, the study of structural influences in the IPV revictimization process has traditionally been neglected. There is currently a growing body of empirical evidence that links the existence of IPV with some structural conditions whose influence goes beyond the individual characteristics (Falb et al., 2015; Heise and Kotsadam, 2015). This evidence does not seem to have been investigated in studies on IPV revictimization, which could reflect a potential limitation of research in this area. Structural aspects such as the human development of a country have shown to be predictors of the country's IPV levels, after controlling for a wide range of individual factors (Herrero et al., 2017b, 2018). It would be necessary to verify if these structural risk factors also play a relevant role in the IPV revictimization of women, especially in multi-country studies.

# The Present Study

The study of victimization by multiple partners has been a relatively neglected topic in IPV research against women. In recent years, however, there has been a growing interest in knowing why some women who have suffered IPV are able to create new intimate relationships free of IPV. The literature has identified some characteristics of these resilient women: they suffered less child abuse, they have been less victimized in their adulthood (by the partner or ex-partner or by other adults), the consequences of IPV were less severe (fewer psychological difficulties), their new relationships are shorter, and tend to be older. The available evidence, however, is not conclusive (Ørke et al., 2018) and, therefore, a greater research effort is needed to clarify both the risk factors and the protective factors. This lack of conclusive empirical evidence is partially explained by the great diversity of samples used in the investigations, sometimes of a small size and limited representativeness, which limits the generalizability of the results. Also, most of the research in this area explores some of the protective and risk factors, but rarely consider multiple factors in a single research design. An additional research effort should be directed toward the evaluation of integrative models and at the same time to favor the use of large samples with increasing degree of representativeness.

Taking into account all the above, the present study aims to analyze the differences between women who suffered IPV with previous partners and also suffer IPV with their current partner (non-resilient women) and women who have suffered IPV with previous partners and do not experience IPV with her current partner (resilient women). The study was carried out in a large sample of women from the 28 member countries of the European Union (N = 2376) from a probabilistic sample of 42,000 women.

Based on the theory and empirical evidence available, the study includes the evaluation of various aspects potentially related to IPV victimization by multiple partners. It includes not only a set of victimization variables from partners, ex-partners, and other adults, but also experiences of child abuse and other risk factors such as psychological difficulties, length of relationships, children witnessing partner abuse, or being able to end a previous relationship because of the violence. It also incorporates a measure of the human development of the country in which the victim lives. Sociodemographic variables such as age, education, income, and size of locality were also included in the study since there are previous research linking these sociodemographic characteristics with IPV from both single and multiple partners (Kuijpers et al., 2011; Palmetto et al., 2013; Ørke et al., 2018). Finally, to better control for potentially biased responses, lack of sincerity of the respondent as appraised by the interviewer was also taken into account.

# MATERIALS AND METHODS

# Participants

Data from the survey on women's well-being and safety conducted in Europe in 2012 in all the 28 member states of the European Union, were used for this study. The survey was conducted by professional interviewers, trained to guarantee confidentiality. Participation was voluntary and at any time during the interview, the respondent could leave the interview if she did not wish to continue. All the information that allowed to identify the respondents was eliminated from the database. In each Member State, the survey covered all women aged 18– 74 years. Certain populations were excluded from the study as people living in institutions and homeless people. Family members who were away from home for a period of 3 months or more were also excluded from the selection. The interviewers presented the survey as a study on women's welfare and safety. Once the respondent accepted to participate, more details about the content of the survey were provided. This was done to protect the respondent's safety in case she lived with someone who did not want the respondent to participate in a survey on violence against women, including possible perpetrators of violence. The averaged response rates for all countries was 77% (FRA, European Union Agency for Fundamental Rights, 2014). We used data from 2376 heterosexual women living with a male partner who had also previously lived with a different male partner and suffered IPV with a previous partner.

# Measures

#### Outcome Variable

fpsyg-09-00877 May 30, 2018 Time: 18:28 # 4

#### **Resiliency**

Two groups of women were formed based on their IPV scores with previous and current partners (see below). A first group of women who reported to the interviewer having suffered IPV only with previous partners but not with the current partner (n = 1624, 72.7%) (only previous IPV group); and a second group of women who reported to the interviewer having suffered IPV with both previous and current partners (n = 594, 23.7%) (previous and current IPV group). Women in the first group (only previous IPV group) were considered to be resilient (resilient = 1) whereas women in the second group (previous and current IPV group) were considered to be non-resilient (resilient = 0).

#### Previous Relationships

#### **IPV with the previous partner**

Respondents informed to the interviewer whether they had experienced psychological, physical, or sexual violence with a previous partner (different from the actual partner).

#### **Physical IPV**

Respondents were asked how often their previous partner had used physical violence toward them across five items (pushed them, slapped them, thrown hard objects at them, grabbed or pulled their hair, beat them with a fist or a hard object or kicked them). Category responses ranged from 1 (never) to 4 (6 or more times). Average physical violence from previous partner was M = 1.32 (SD = 0.64) (Cronbach's α = 0.88).

#### **Psychological IPV**

Respondents were asked how often their previous partner had used psychological violence across four items (belittled or humiliated them in front of other people or in public, scared or intimidated them on purpose, made them watch pornographic material against their wishes, or threatened to hurt them physically). Category responses were coded 1 (No) to 2 (Yes). Average psychological violence from previous partner was M = 1.22 (SD = 0.31) (Cronbach's α = 0.79).

#### **Sexual IPV**

Respondents were asked how often their previous partner had used sexual violence across four items (forced them to have sexual intercourse by holding them down or hurting them, attempted to force intercourse, made them take part in any kind of unwanted sexual activity or being unable to refuse, consented sexual activity because they were afraid of what the partner might do if they refused). Category responses ranged from 1 (never) to 4 (6 or more times). Average sexual violence from previous partner was M = 1.13 (SD = 0.49) (Cronbach's α = 0.89).

#### **Psychological difficulties**

Foa et al. (2000) operationalized this construct using information about posttraumatic stress disorder (PTSD), depression, anxiety, and substance abuse. Only information about depression and anxiety was available in the FRA survey. Respondents were asked whether they have suffered depression and/or anxiety as a result of the more serious IPV incidents with a previous partner. Category responses were 0 = No, and 1 = Yes. Psychological difficulties scores were calculated summing all the Yes responses (M = 0.42, SD = 0.65).

#### **Children witnessing IPV**

Respondents were asked whether have any children who have been living with her ever been aware of any violent incidents by the previous partner. Category responses were: 1, Yes, 2, No, and 3, No children living with me at the time of the incidents. Original responses were coded as 1, Yes (1836, 77.3%) and 0, No or no children living with her (540, 22.7%).

#### **Ending the relationship because of IPV**

Participants were asked whether they ended the relationship because of violence: Did you end your relationship with any of your previous partners because of violence? Category responses were: (1) Yes, the main reason (55.0%, 962); (2) Yes, but it was not the main reason (20.4%, 356); and (3) No (24.6%, 430). We coded this variable to have the value of 1 for Yes, the main reason, and the value of 0 for all the remaining responses.

#### Current Relationships

#### **IPV with the current partner**

Respondents informed to the interviewer whether they had experienced psychological, physical, or sexual violence with the current partner (different from the actual partner).

#### **Physical IPV**

Respondents were asked how often their current partner had used physical violence toward them across six items (pushed them, slapped them, thrown hard objects at them, grabbed or pulled their hair, beat them with a fist or a hard object, or kicked them). Category responses ranged from 1 (never) to 4 (all the time). Average physical violence from current partner was M = 1.06 (SD = 0.27).

#### **Psychological IPV**

Respondents were asked how often their current partner had used psychological violence across four items (belittled or humiliated them in front of other people or in public, scared or intimidated them on purpose, made them watch pornographic material against their wishes, or threatened to hurt them physically). Category responses ranged from 1 (never) to 4 (6 or more times). Average psychological violence from current partner was M = 1.09 (SD = 0.27).

#### **Sexual IPV**

Respondents were asked how often their current partner had used sexual violence across four items (forced them to have sexual intercourse by holding them down or hurting them, attempted to force intercourse, made them take part in any kind of unwanted sexual activity or being unable to refuse, consented sexual activity because they were afraid of what the partner might do if they refused). Average sexual violence from current partner was M = 1.02 (SD = 0.15). A single measure of IPV with the current partner was calculated averaging the scores of psychological, physical, and sexual IPV with the current partner (Cronbach's α = 0.81; M = 1.08; SD = 0.21). Those respondents scoring 1 (never) were classified as resilient women while those women scoring higher than 1 were classified as non-resilient women.

#### **Length of the current relationship**

fpsyg-09-00877 May 30, 2018 Time: 18:28 # 5

Respondents were asked: How long have you been together in total since you started dating? Category responses ranged from 1, under a year to 7, more than 50 years. Averaged length was close to 11–20 years (M = 2.96, SD = 1.076).

#### Non-partner Victimization

#### **Victimization by non-partners**

Respondents were asked about having experienced physical and sexual violence with other adults than partners or ex-partners. The same scales of physical (M = 1.27, SD = 0.52) and sexual (M = 1.08, SD = 0.29) IPV were used referring to people other than partner or ex-partners (Cronbach's α = 0.79).

#### Child Abuse

#### **Child abuse (CA)**

Respondents were asked how often they had experienced different types of physical, emotional or sexual acts from somebody older than 18 years when they were under 15 years of age.

#### **Physical CA**

The following five items were used to measure physical CA: (1) slap you or pull your hair so that it hurt you, (2) hit you very bad so that it hurt you, (3) kick you very bad so that it hurt you, and (4) beat you very bad with an object like a stick, cane or belt so that it hurt you, and (5) stub or cut you with something so that it hurt you. Category responses ranged from 1 (never) to 3 (more than once).

#### **Emotional CA**

The interview used the following four items to measure emotional CA: (1) say you that you were not loved, (2) say you that they wished you had never been born, (3) threaten to abandon you or throw you out of the family home, and (4) threaten to hurt you badly or kill you. Category responses ranged from 1 (never) to 3 (more than once).

#### **Sexual CA**

The interview used the following five items to measure sexual CA: (1) expose their genitals to you when you did not want them to, (2) make you pose naked in front of any person or in photographs, video or an internet webcam; (3) Touch your private parts – genitals or breasts – when you did not want them to, (4) make you touch their private parts – genitals or breasts – when you did not want to, and (5) Make you have sexual intercourse with them when you did not want to. Category responses ranged from 1(never) to 3 (more than once). A single measure of CA was calculated averaging the scores of physical, emotional, and sexual CA (Cronbach's α = 0.73; M = 1.15; SD = 0.25).

#### Country-Level Variable

#### **Human development index (HDI)**

The HDI measures country development by combining health, education, and wealth. A higher value indicates a higher level of human development (M = 871.54, SD = 35.40). The HDI information for each country was retrieved from international databases (United Nations Development Programme, 2013). Other research has also incorporated HDI as a variable of the country-level and has found that this index outperforms other indexes such as the Gender Inequality Index in its relationship with both IPV and acceptability of IPV (Herrero et al., 2017a).

#### Sociodemographic Characteristics

The survey provided data about respondents age, satisfaction with household income, size of locality, and educational background.

#### **Age**

The ages of respondents were originally coded into seven age groups: 18–24, 25–29, 30–34, 35–39, 40–49, 50–59, and 60+ (M = 5.06, SD = 1.503). The average respondent was in the 40–49 age group.

#### **Dissatisfaction with household income**

The respondents' satisfaction with household income was measured with the following question: "Which one of the descriptions on this card comes closest to how you feel about your household income nowadays?" Responses were coded from 1 (living comfortably on present income) to 4 (finding it very difficult on present income; M = 2.16, SD = 0.95).

#### **Size of locality**

Respondents were asked to describe the type of locality in which they lived: "Which option on this card best describes the area where you live in?" The responses were coded from 1 (a big city or outskirts of a big city) to 4 (a farm or home in the countryside; M = 2.69, SD = 1.21).

#### **Educational background**

Respondents' educational background was coded using a threecategory response scale in the original dataset to make the results comparable across different national educational systems from 1 (primary) to 3 (tertiary; M = 1.93, SD = 0.70).

#### Interviewer Variables

#### **Insincere responses**

Interviewers were asked to assess whether each respondent's responses were insincere overall: "Do you think the respondent was telling the truth in the interview?" Responses ranged from 1 (yes, all the time) to 4 (not at all; M = 1.19, SD = 0.46). Although other research with this same dataset has also controlled for privacy and safety during the interview (Herrero et al., 2017b), we did not include these variables in the analyses because of zero variance in some countries (Germany, Lithuania, Netherlands, Portugal, Sweden, and United Kingdom).

#### Statistical Procedures

We used multilevel regression modeling to take into account the hierarchical structure of the data – individuals (level 1)

nested within countries (level 2). All predictors were centered around the grand mean to ease interpretation of results. Multiple imputations of missing values were performed (Rubin, 1996).

#### RESULTS

**Table 1** shows the results of the multilevel regression model. At the individual level, age, satisfaction with household income, length of the current relationship, psychological difficulties, physical IPV with previous partner, child abuse, victimization with other people than the partner or ex-partner, and having ended a previous relationship because of IPV showed a significant relationship with the groups of non-resilient and resilient women. Age was positively related to being in the group of resilient women: a one-unit increase in age increased 1.11 times the odds of being in the group of resilient women. Dissatisfaction with household income (b = −0.213, p < 0.01) was negatively related to being in the group of resilient women. Given that this variable takes the value 0 for non-resilient women and the value 1 for resilient women, dissatisfaction with household income is predictive of being in the group of non-resilient women. When the OR is lower than 1 and significant, as with dissatisfaction with household income (OR = 0.808), the inverse (1/OR = 1/0.808 = 1.24) measures the association between the predictor and lower values on the grouping variable (being in the non-resilient group of women). Thus, a one-unit increase in dissatisfaction with income increased 1.24 times the odds of being in the group of non-resilient women. Length of the current relationship (OR = 0.759, inverse = 1.32) was also significantly associated with the outcome variable. A one-unit increase in the length of the current relationship increased 1.32 times the odds of being in the group of non-resilient women.

Physical IPV with the previous partner was positively associated with the outcome variable (OR = 1.48): a one-unit increase in physical IPV from the previous partner increased 1.48 times the odds of being in the resilient group of women. Psychological difficulties stemming from IPV was associated with the outcome variable (OR = 0.833, inverse = 1.20): a one-unit increase in psychological difficulties increased 1.20 the odds of being in the non-resilient group of women. A larger statistical relationship was found for child abuse (b = −1.215, p < 0.001) and victimization by non-partners (b = −0.628, p < 0.001). Looking at the inverse of their odds ratios, a one-unit increase in child abuse (OR = 0.297, inverse = 3.37) increased 3.37 times the odds of being in the group of non-resilient women. Transforming the odds ratio (0.297) to probabilities (odds/odds+1 = 0.23) gives an intuitive illustration of the effect of child abuse on IPV victimization from multiple partners. Although by chance women of the sample would have a probability of 50% to be in the resilient group, a one-unit increase in child abuse reduces this probability in 27% (0.50–0.23 = 0.27). A one-unit increase in victimization by non-partners (OR = 0.534, inverse = 1.87) increased 1.87 times the odds of being in the group of nonresilient women. Having ended a previous relationship because of IPV significantly increased the odds of being in the resilient group of women (OR = 1.426).

As for the country-level variable of the study, the unstandardized coefficient expresses the linear relationship between HDI and the ratio resilient/non-resilient women in each country. Note at this point that, while the dependent variable is dichotomous at the individual level (non-resilient vs. resilient), it is no longer dichotomous at the country level. This is so because for each country the ratio of resilient/non-resilient women is estimated and this ratio is no longer dichotomous, but continuous. Higher values in country HDI were statistically related (b = 0.005, p < 0.01) to a higher proportion of resilient women in that country (higher values in the outcome variable). In other words, countries higher on HDI show a tendency to have a greater proportion of resilient women than the average country in terms of HDI.

Overall, when compared to non-resilient women, resilient women were more satisfied with household income, had shorter current relationships, informed to the interviewer to have experienced higher levels of previous physical IPV, as well as fewer child abuse and fewer victimization by non-partners or ex-partners, and experienced less psychological difficulties as a result of previous IPV. They ended the relationship because of the violence to a greater extent and lived in countries ranked higher in HDI. These results take into account the hierarchical structure of the data and are adjusted by the insincerity of respondents as appraised by the interviewer. In fact, this variable turned out to be statistically significant (b = −0.363, p < 0.10, OR = 0.695). If we consider the inverse of the OR ( = 1/0.695 = 1.43) we see that a one-unit increase in insincerity increased 1.43 times the odds of being in the group of non-resilient women. In other words, the women who reported IPV for multiple partners were evaluated by the interviewers as less sincere.

#### DISCUSSION

The study of the multiple IPV victimization by different partners is a relatively neglected area of study. The absence of a greater research effort in this area is surprising considering the frequency with which some women are victimized by different partners (Krause et al., 2006). In the present study, we aimed to analyze the differences between non-resilient (multiple IPV victimization) and resilient (non-multiple IPV victimization) women from the 28 member countries of the European Union (N = 2376). Based on the available literature and empirical evidence we included the study of a number of variables about women past victimization by both partners and non-partners, child abuse, psychological consequences of the abuse, and socio-demographic characteristics. Additionally, we also included variables that have not been traditionally analyzed in this area. Specifically, we also studied the influence of the victim's ability to end the abusive relationship -at the individual level- and the human development of the country in which the women lives -at the country level.

Some of the results obtained in other studies were replicated in our study. With regard to sociodemographic variables, we found, as well as other studies, that victimization by different partners is more frequent among younger women (Testa et al., 2003;

TABLE 1 | Non-resilient vs. resilient women to intimate partner violence (IPV) in the European Union (N = 2376): Unstandardized multilevel regression estimates, robust standard errors, and 95% confidence intervals (C.I.).


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

Alexander, 2009; Tsirigotis and Łuczak, 2018). We also found that non-resilient women had lower satisfaction with income than resilient women. The evidence available at this point in the literature is mixed. In the review of studies by Ørke et al. (2018), only one study reported a positive relationship between low income and victimization by multiple partners (Vatnar and Bjørkly, 2008) while other research on IPV has informed of a negative relationship between satisfaction with income and both physical and psychological IPV (Herrero et al., 2017b).

Length of the relationship was positively related to IPV victimization by multiple partners. As Logan et al. (2006) suggested, during the first stages of a new relationship it might be not only that IPV is absent but also that IPV cues are misinterpreted (e.g., controlling, monitoring, and stalking behaviors). If this is the case, as the relationship progresses with time, women may be able to correctly identify episodes of IPV with their partner.

An important aspect that the present investigation has not been able to confirm is the relationship between previous and actual IPV. There are some studies that have found that higher levels of previous IPV were predictive of higher current IPV (Testa et al., 2003; Cole et al., 2008) while others have not found a significant association (Stein et al., 2016). Our results are partially in the same line as those found by Stein et al. (2016) since both psychological and sexual IPV with the previous partner was unrelated to current IPV. Our study found a negative relationship between previous physical IPV and current IPV, however, those women who had suffered more physical IPV with previous partners were more resilient to future IPV, as they more frequently informed to the interviewer of a free-IPV relationship with their current partner. According to Foa et al's. (2000) model, women who experience more severe IPV, as well as higher levels of psychological difficulties, become more vulnerable, thus reducing their resiliency to future IPV. What we found in our study is that resilient women suffered less psychological difficulties as a consequence of the abuse, as predicted by Foa et al. (2000), but they also presented higher levels of previous physical IPV. In this same line is the work of Cobb et al. (2006) who found greater post-traumatic growth in women with higher rates of previous physical IPV, which suggests that these women with higher rates of previous physical IPV had managed to overcome the violence. It seems, therefore, that some women with higher levels of previous physical IPV reacted to the abuse in a way that reduced their psychological vulnerability, which in turn translated into greater resilience.

One of the possible victim's responses with a detrimental effect on future IPV could be ending the relationship due to the violence suffered (Cobb et al., 2006). The resilient women in our study reported having ended the relationship due to episodes of violence more frequently than non-resilient women. The way these women cope with the abuse seems, therefore, more relevant to their personal adjustment than the levels of IPV suffered (Senter and Caldwell, 2002). This would be especially relevant in the case of previous physical IPV, as our data suggest. These results complement those obtained in the study of repeated violence with the same partner. According to Dichter and Gelles (2012), the evidence on the effect of breaking the abusive relationship in the subsequent rates of IPV could be explained in terms of the reasons underlying that violence, differentiating what is battering from what it is not. When the violence exerted is motivated to gain coercive

control over the victim (battering), victim's decisions about ending the relationship could exacerbate these violent episodes in the aggressor. Thus, leaving would threaten the aggressor's dominance, which may use violence in an attempt to regain power. Alternatively, when the violence exercised is mainly motivated by anger, frustration, retaliation, or self-defense, the end of the relationship is likely to lead to the cessation of violence. As pointed out by Dichter and Gelles (2012), the measurement of IPV from violent incidents, as in the present study, does not allow to differentiate battering from what is not, since the former must also include an evaluation of the dynamic of the violence in the relationship. This should include aspects such as the perpetrator's motive to control the victim or the victim's experience of being dominated and controlled by the violence. Future research on the victimization of IPV by multiple partners should, therefore, include measures on the reasons for violence to identify the existence of battering to further verify this assumption. This might shed light on to what extent the rupture of the violent relationship may or may not remain a potential risk situation for the victim.

The strongest predictor of IPV victimization by multiple partners found in our study was the level of child abuse suffered. This may be because victims of child abuse tend to associate with potentially abusive partners in adulthood. This process, which Herrero et al. (2018) have called conditional partner selection, suggests that victims of abuse in childhood develop a series of psychological deficits that increase the likelihood of ending with potentially abusive partners (see Torres et al., 2013 for analysis of the characteristics of potentially abusive partners), who also find some of these deficits as something attractive in their partners (i.e., anxiety attachment). The empirical link between child abuse and IPV has been consistently found in a number of studies (see an analysis in Herrero et al., 2018). Likewise, in their review of studies on victimization by multiple partners, Ørke et al. (2018) only found one study that did not find any positive relationship between child abuse and victimization by multiple partners (Coolidge and Anderson, 2002) and one study (Alexander, 2009) that only found this link for child sexual victimization but not for abuse and neglect. Our results add to this empirical evidence and help to situate adverse experiences in childhood as one of the main predictors of IPV victimization by different partners in adulthood. This evidence also adds to the empirical evidence already found between child abuse and adulthood victimization.

Adult victimization by adults other than partner and ex-partner was also found to be an important predictor of IPV victimization by different partners. Although its effect is less than that found for child abuse, its influence is relevant when explaining the IPV victimization by different partners. The literature on the effects of lifetime victimization on partner abuse has found mixed results depending on the type of adult victimization measured (Cole et al., 2008; Stein et al., 2016). For instance, Stein et al. (2016) did not find a statistical relationship between having been a victim of both sexual and non-sexual assault and multiple victimizations by different partners in 164 women victims of IPV (Stein et al., 2016). The relationship between non-IPV and multiple IPV victimization by different partners in adulthood has been traditionally explained in terms of the increase in the vulnerability of women victims of victimization and their problems with interpersonal relationships (Nishith et al., 2000; Classen et al., 2005). From this point of view, victimization in adulthood would operate in a similar way to how childhood victimization operates: decreasing the psychological resources of the victim and negatively conditioning their social development (Herrero et al., 2018). Because victimization in childhood has a greater effect on the victimization of IPV by multiple partners than non-IPV adult victimization, according to our findings the sooner victimization occurs, the greater will be its effect on the psychological and social development of the victim.

At the country level, our results suggest that the greater the human development of a country, the higher the proportion of women resilient to IPV in this country. This result points to the importance of the structural factors to understand both the IPV and the IPV victimization by different partners. While the effect of structural factors on IPV is being studied in recent years, this research effort has not seemed to be transferred to the study of IPV victimization by different partners. Herrero et al. (2018) found among more than 20,000 women of 28 countries of the European Union that the country's human development not only negatively influenced the country's IPV rates but it also affected the way in which other predictors of the IPV operated. Other research had already shown a few years earlier that the structural conditions of societies could affect their gender value systems. Thus, citizens from countries with more egalitarian structural conditions (measured through indexes such as the Gender Empowerment Index or the HDI) tend to show more egalitarian gender attitudes and lower IPV acceptability (Gracia and Herrero, 2006; Brandt, 2011). Our results allow extending the influence of the structural conditions not only to women's IPV victimization but to the victimization by different partners. Citizens of countries higher on human development are not only more protected against potential IPV victimization but also tend to be more resilient to this IPV. This circumstance, undoubtedly, suggests extending preventive efforts to levels other than those of the individual. Future research should also incorporate the study of contextual factors since they not only influence the rates of IPV but may also influence other risk factors observed in this study. Contextual risk factors associated to child abuse (Gracia and Herrero, 2008) are particularly relevant at this point, since it has shown to be one of the most important predictors of resilience to multiple IPV victimization by different partners in our study.

The fact that the main antecedent of the multiple victimization found in our study – the child abuse suffered – is distal in nature, should make us think about the importance of global preventive policies throughout the life cycle. Adverse experiences in childhood negatively affect the personal and social development of women, and the structural conditions of the society in which they live exert a notable influence as well. These adverse experiences in childhood not only condition violent relationships with specific partners. Rather, they seem to be linked to a trajectory of sustained vulnerability in intimate partner relationships, characterized by repeated victimization with different partners. In addition, the structural conditions, which are beyond the control of the victim of IPV, also have a distal nature. Again at this point, preventive policies and an orientation of the public administrations toward an improvement of the human development level of the countries are needed.

# Strengths and Limitations

fpsyg-09-00877 May 30, 2018 Time: 18:28 # 9

The present study presents strengths as well as potential limitations. Among the strengths we highlight, on the one hand, the sample used. Having a large sample of IPV resilient and non-resilient women from a probabilistic sample of women from the 28 countries of the European Union is a strength of the study. The lack of large representative samples has probably been a limitation to the development of research in this area. The present study allows an approximation of the real percentage of IPV resilient women in the European Union, which is estimated at around 73% of the female population that has previously suffered IPV with other partners. According to this, a large percentage of the women who suffered IPV were in some way resilient to IPV and able to avoid new violent relationships. The absence of psychological difficulties and the lower frequency of adverse situations (in childhood or in adulthood) are among the characteristics of this group of resilient women. On the other hand, the multi-country nature of the sample has allowed the analysis of structural influences on resilience, which constitutes an innovation in this field of study. Although it is increasingly common to incorporate the study of the influence of structural factors on IPV, to our knowledge this has not yet been applied to the study of IPV resilience. Structural explanations are important since they allow to effectively contextualize the processes under study. If IPV is conditioned by structural factors of society such as the gender value system or gender-related inequalities in health, education, economy, or politics, it is advisable to study them in investigations that incorporate, as the present one, variables in the individual and country levels.

The present investigation, however, is not free of potential limitations. A first limitation would be the fact that the FRA survey does not include information on the behavior of the woman and therefore does not allow an analysis of the bidirectionality of the violence. An alternative explanation for the lack of resiliency of some women in the study could be their aggressive behavior that might elicit aggressive responses in their partners. It would not be exclusively, therefore, a matter of victimization, but also of perpetration of the IPV. Unfortunately, we do not have this information, so further studies should aim to identify this potential group of more aggressive women so as not to confuse them with non-resilient women. Related to this, the interview was the only measure administered and, although controls were carried out on potential response biases, future research would benefit from the inclusion of additional measures beyond the context of the interview. Of particular importance would be the assessment of the mental health status of the participants to potentially exclude women with mental illness whose responses may be distorted, which would pose a threat to the validity of the study. The controls on the response bias performed in the analyses, however, could have alleviated this potential threat. Another possible limitation may lie in the retrospective nature of this work: the victimization rates in the adulthood could have conditioned the recall of victimization episodes suffered in childhood. Although this is a possibility that is always present in retrospective studies that involve remembering past situations that are sometimes very distant in time, the relationships observed between victimization in childhood and in adulthood in this study are consistent with what has been obtained in other investigations (Herrero et al., 2018). It does not seem therefore that the nature of the study has substantially affected the results of the study.

The present investigation has allowed knowing in greater depth some risk factors of women non- resilient to the IPV. Some aspects related to the resilience and that could also exert a notable influence were not considered, however. For example, the empirical evidence indicates that social support is negatively related both to the repeated violence of IPV and to the victimization of IPV by different partners (Kuijpers et al., 2011; Dichter and Gelles, 2012) and that it can be an important characteristic of resilient women (Dutton and Greene, 2010). Further, as most research on social support and resilience has focused primarily on social support from families and friends (Norris and Stevens, 2007), research in this area would benefit from the inclusion of areas of social support other than family and friends that have been linked to resilience stemming from the community (Norris and Stevens, 2007), such as community integration, and both formal and informal community support (Herrero et al., 2004, 2011; Herrero and Gracia, 2007; Juarros-Basterretxea et al., 2018).

# ETHICS STATEMENT

All subjects gave written informed consent in accordance with the Declaration of Helsinki. Ethics approval for the secondary analysis of data was obtained from the Agency for Fundamental Rights of the European Union, who provided a special license for this purpose (Reference No. 96457). Also, the study is compliant with the statements of Principle in Code of Practice for Official Statistics and the specific requirements of the Protocol on Data Access and Confidentiality of the United Kingdom, where the UK Data Archive stores the data.

# AUTHOR CONTRIBUTIONS

All authors contributed equally to this manuscript. It was based on an original draft of JH, further improved by PV, AT, and FR in terms of conceptual framing, data analysis, and discussion of results.

# FUNDING

Support for this research was provided by FEDER funds from the Spanish Ministry of Economy and Competitiveness (MINECO-17-PSI2016-77484-P).

# REFERENCES

fpsyg-09-00877 May 30, 2018 Time: 18:28 # 10



Vatnar, S. K. B., and Bjørkly, S. (2008). An interactional perspective of intimate partner violence: an in-depth semi-structured interview of a representative sample of help-seeking women. J. Fam. Violence 23, 265–279. doi: 10.1007/ s10896-007-9150-7

**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 © 2018 Herrero, Vivas, Torres and Rodríguez. 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 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.

# Intimate Partner Violence and its Escalation Into Femicide. Frailty thy Name Is "Violence Against Women"

#### Georgia Zara<sup>1</sup> \* and Sarah Gino<sup>2</sup>

*<sup>1</sup> Department of Psychology, University of Turin, Turin, Italy, <sup>2</sup> Laboratory of Criminalistic Sciences "Carlo Torre, " Department of Public Health and Pediatrics, University of Turin, Turin, Italy*

Violence against women is a disabler of dignity, liberty, and rights of the person, with murder being its extreme form for silencing the individual. Despite psycho-criminological research providing evidence that violence can happen across cultures, sexes, and societies, other findings show that some forms of violence i.e. Intimate Partner Violence (IPV), which involves more frequently women as victims, is not rare in contemporary society. The aim of this study is to analyze the violence against women, and how it escalates up to the point in which it aggravates into femicide. In order to carry out this study, data from both the Turin Archive of the Institute of Legal Medicine (1970–1997), and the Archive of the Central Morgue (1998–2016) were collected. The interest was to focus on those women who were killed in Turin, between 1970 and 2016, by a male with whom they were involved in a more or less intimate relationship (e.g., matrimonial, sexual, friendship, professional, etc.). Collateral information was also gathered from forensic files that reported sufficient details about the criminal events. The sample was composed of 275 women killed by violence in Turin, Italy, by 260 males. This research was based on two questions: Is murder the worst possible scenario of a long-lasting abusive relationship? Are we witnessing a shift in how violence now happens, becoming perhaps less striking than murder, but not less painful from the victim's point of view? These findings show that escalation into femicide featured more likely within an intimate and affective relationship between victim and perpetrator; they also show that when the perpetrator knew the victim, it was more likely that an overkilling took place. When victims sustained multiple injuries that went beyond those necessary to cause their death, one is in front of an overkilling. These results also suggest that motives behind intimate partner femicide could account for a differential degree of violence, so that the longer and closer the relationship was between victim and perpetrator, the higher the risk of IPV escalating into femicide, and of femicide being executed with extreme and severe force.

Keywords: violence, intimate partner violence, intimate relationships, femicide, overkilling, risk

# INTRODUCTION

Violence is a disabler of dignity, liberty, and rights of the person, with murder being its extreme form for silencing the individual. The focus here is on violence against women that leads to femicide.

Violence against women (VAW) can take many forms and shapes; it can be acted out through behavior or it can be psychological, and therefore difficult to be seen or measured. It can be

#### Edited by:

*Shulamit Ramon, University of Hertfordshire, United Kingdom*

#### Reviewed by:

*Carlos Chiclana, CEU San Pablo University, Spain Michelle Lowe, University of Bolton, United Kingdom*

> \*Correspondence: *Georgia Zara georgia.zara@unito.it*

#### Specialty section:

*This article was submitted to Gender, Sex and Sexuality Studies, a section of the journal Frontiers in Psychology*

Received: *29 May 2018* Accepted: *03 September 2018* Published: *26 September 2018*

#### Citation:

*Zara G and Gino S (2018) Intimate Partner Violence and its Escalation Into Femicide. Frailty thy Name Is "Violence Against Women". Front. Psychol. 9:1777. doi: 10.3389/fpsyg.2018.01777*

**83**

enduring and last for long periods, or it can be brief, but nonetheless intense. VAW is one of the highest concerns globalwise, and much attention, resources and sensitivity are required to put in motion coordinated strategies to intervene to stop its escalation and its worsening.

Though studies are suggesting that violence is not exclusive to men, but that women could also be perpetrators (Sommers, 1994; Ristock, 2002; Belknap et al., 2012), data from the World Health Organization (WHO) are clear in suggesting that women are more likely than men to be the target of violence (Stewart, 2002; ISTAT, 2006), of sexual abuse (Zara, 2018), of psychological abuse (Pico-Alfonso, 2005), of domestic violence (Johnson and Ferraro, 2000; Lloyd et al., 2017), of intimate partner violence (Theobald et al., 2016). This could be also a result of women becoming more open about the problems than men, hence they are more likely to report the abuse and violence than men are (Dutton, 2010), and this may be more so the younger the people involved (Capaldi et al., 2012; Kim et al., 2016).

The most common form of violence suffered by women is intimate partner violence (IPV) (World Health Organization, 2013). IPV is a global concern (Butchart et al., 2010). IPV occurs in different settings, across socio-economic classes, cultures, ages, religious groups (Archer, 2006). In its extreme forms, IPV causes death (Campbell et al., 2007).

Available definitions of IPV are diverse, often either too inclusive or too specific not allowing for a comprehensive perspective. Not all forms of abuse that women suffered from occur within domestic life. Nor can IPV always be explained by gender violence.

Looking at some of the major studies carried out in this field (e.g., Dutton and Goodman, 2005; Stark, 2007; Campbell et al., 2008; Johnson, 2008; Hart and Klein, 2013), Hart and Klein (2013) suggest that IPV endorses physical, sexual, psychological, economic abuse and stalking that are the five multi-faceted methods of violence and abuse that perpetrators utilize to achieve, maintain and regain control of their intimate partners or potential ones. Coercion or terroristic threats, coupled with any of the five methods of abuse above mentioned, constitute IPV (Smith Slep et al., 2016).

Intimate relationships are defined as relationships that involve an intense emotional and/or physical investment (Miller, 2015). Intimacy is a primary human need (Strong et al., 2011), and the need to belong is a fundamental human motivation that fosters desire for interpersonal attachments (Baumeister and Leary, 1995). This is why, Baumeister and Leary (1995) advocate that "much of what human beings do is done in the service of belongingness" (p. 498). Human companionship strongly affects the quality of life (Robles et al., 2014), as well as influencing rates of illnesses, such as cancer, and mental well-being (Whisman et al., 2010). This is no surprise since studies show that people who experience a fulfilling relationship or marriage are generally healthier and have lower mortality rates than divorced, separated, and never-married individuals (Strong et al., 2011).

The intense emotional investment in a relationship could also be imaginary when a person thinks that the other partner is reciprocating and is interested in developing closeness and physical proximity too. According to the Encyclopedia of Human Relationships (Reis and Sprecher, 2009) relationships are fundamental to nearly all domains of human activity along the life-course. In the broad concept of human relationships are included all types of human associations, friends, lovers, spouses, room-mates, work colleagues, team-mates, parents and children, relatives, neighbors, business partners, and so forth. Although each of these connections is unique in some respect, they share a common core of principles and processes. When people are involved in healthy, satisfying relationships, they live, work, and learn more proficiently. When relationships are distressing or frustrating, or too asymmetric in the types and quality of interaction, people are less satisfied, less healthy, and less constructive. Dysfunctional relationships are often at the basis of abuse, violence, homicide and femicide (Shackelford et al., 2005).

Violence does not always, or immediately, lead women to death, but the consequences of these acts are equally debilitating; the physical, psychological and social effects of violence vary. It has been established that VAW incurs high costs<sup>1</sup> , including both those related to preventing or dealing with this type of violence (such as police, risk assessment practice, law enforcement)<sup>2</sup> , and those costs incurred by its consequences (such as health) (Varcoe et al., 2011).

## The Scientific Focus of IPV and Femicide

Scientific, clinical and social concerns directed at violence against women, and at how relationships between men and women

<sup>1</sup>Calculating the costs of VAW is complicated because studies employ different methods of estimation, gather different data on different types of violence (intimate partner or domestic violence), count various types of costs (to direct and indirect victims, employers, the community, and governments) (Council of Europe, 2012). A report by Badalassi et al. (2013) entitled "Quanto costa il silenzio?" (that means "What is the cost of silence?"), published by the Italian NGO Intervista Onlus (and from now on quoted as Intervista Onlus), attempts to offer accurate estimates of the costs. The study concludes that the overall indirect (non-monetary) and social multiplier costs of violence against women in Italy is about EUR 14.3 billion (approximately USD 16.2 billion) (Intervista Onlus, 2013, p. 15). According to the same study, when the direct costs to the public sector (EUR 1.8 billion) (USD 2.1 billion) and the economic multiplier costs (EUR 604 million) (USD 690 million) were added, the total was estimated at EUR 16.7 billion (USD 19.19 billion). The study also shows that the costs related to the prevention and contrast of violence against women would be around EUR 6.3 million (USD 7.2 million) (Intervsita Onlus, 2013, p. 15).

<sup>2</sup>Taking into account all the costs briefly described, different laws were passed in the world to regulate VAW. In 2011, the Europe Council ratified the Istanbul Convention (2012) that was the first comprehensive tool for opposing all forms of violence. In Italy, a legal enforcement process against discrimination was achieved, passing through different steps (as New Family Law, Divorce Law and Abolition of Crimes of Honor) and in 2008, with the enactment of the law known as "Measures against persecutor acts," some significant measures were set up as the preliminary sign of significant social and legal changes. In 2013, Italy ratified the Istanbul Convention (2012). Moreover, the Italian government passed an array of measures designed to tackle the problem of violence against women, and introduced Law Decree n. 93 of 14 August 2013, which then converted, with amendments, into Act n. 119 of 15 October 2013. This law was enacted by laying down some key issues: complaint irrevocability; legal aid to victims; increased punishment in special cases (e.g., victim's pregnancy, presence of children, etc.); the possibility for the alleged victims of reporting anonymously to the police their claims over IPV incidents; the introduction of a warning procedure in the case of stalking. Legal initiatives were complemented by the activity of anti-violence centers and women's associations, springing up on Italian territory in the last 20 years, with the aim to fill gaps in the welfare body.

escalate into femicide have generated significant world-wide initiatives to attempt to intervene efficiently enough to prevent it from its buds or to stop its persistence, escalation and aggravation.

The term femicide was first used in 1801 by Corry in his—A Satirical View of London at the Commencement of the Nineteenth Century (p. 49)—intended to signify the killing of a woman. His description of this species of delinquency (as he put it) was also influenced by the cultural and moral tone of the period, though not diametrically opposed to many of the motives behind IPV nowadays. This is confirmed by the fact that it was also recognized that women could perform violence and murder (Corry, 1801/1809, p. 168).

Nearly two centuries after, femicide was used to symbolize a gender-based murder (Russell and Harmes, 2001; Russell, 2012). The term includes not only female murders committed by partners or former partners, but also of girls murdered by their fathers or relatives because they rebel against an obsessive control of their lives, identities or sexual choices, or because they refuse a marriage imposed to them (Russell and Harmes, 2001). The term also includes any murder of a woman by a man, independently of the type and intensity of the relationship, because of the exercise of power or dominance, for reason of hate, disdain, passion or sense of ownership over the same woman.

The social and ideological mainstream stance that VAW and femicide research has taken (Bandelli, 2017) fosters the assumption that if one investigates closely the types of perpetrators and victims, and their relationships, one betrays the social and cultural nature of it. Conversely, scientific research should fight against any form of reductionism, if the endeavors are to grasp complexity, and to avoid moral panic. The former requires bringing VAW into focus by offering an integrated perspective in which individual, relational and social dimensions are not artificially separated. The latter looks critically at any delimitation in this field of knowledge that "lay down the rules by which the problem can be talked about" (Critcher, 2003, p. 168).

The separation of the individual from the social setting is not recent, and many eras in history have witnessed this hiatus. Donati (2011) argues that "today matters are different because they are no longer simply instances of dehumanization, but of an irruption of the inhuman into the social, one that progressively displaces what is still human" (p. 21). By doing so, the risk is "that the 'social' is no longer seen or heard [. . . ] as something immediately human" (Donati, 2011, p. 21), and it becomes not anymore related to the individual reality, and to how the person contributes to promote or endures it.

Few studies have examined these aspects and played a significant part in expanding our understanding, and enlightening the processes that can trigger interpersonal and intimate violence into women targeted violence, and from this into femicide. It is interesting to see that some of these studies embrace either a forensic and legal medicine approach or a psycho-criminological and risk assessment approach.

For the former approach, it is worth mentioning three recent studies carried out in Italy. Bonanni et al. (2014) investigated the Italian scenario of femicide by analyzing four cases chosen to profile a specific sub-group of femicide, and comparing it with the international one. Data regarding the type of relationships victims and perpetrators had before the killing took place, what still bound them to the point of the extreme act of killing, and whether the victim's decision to break off the relationship represented the trigger for the rage into femicide, were examined. It emerged that the prior relationship between perpetrator and victim was relevant in the cases investigated, and also influenced the modus operandi by which women were killed.

Moreschi et al. (2016) explored the cases of 34 femicides that occurred in Italy over a 21-year period (1993–2013). Besides the analyses of typical epidemiological aspects of femicide, the focus was on the circumstances and risk factors surroundings the crime, and through the examination of aspects such as perpetrator's motive or specific risk factors (e.g., legal possession of firearms, previous violence and threats, time occurred after the ending of the relationship), their study aimed at profiling some possible preventive strategies.

Another work on femicide in north-west Italy carried out by Trecastagne et al. (2016) integrated a forensic pathology approach with a more social perspective in order to establish to what extent the law and the cultural changes, which took place in Italy between 1970 and 2012, had an impact on the rate of these crimes.

For the psycho-criminological and risk assessment approach, some specific studies look at the risks involved in VAW and in IPV. Owing to the seriousness and prevalence of the problem, Kropp and Cook (2014) believe that it is essential to provide the criminal justice system, the court, the health care, and shelter and protection settings with scientific evidence on how to conduct risk and threat assessment, and which instruments to use, with whom, when, and in which contexts. Most IPV risk assessment tools<sup>3</sup> involve an integrated assessment of criminogenic risks of offenders, along with an evaluation of the victimogenic risks of the victims (Campbell et al., 2001; Baldry and Winkel, 2008). When assessing the risk of IPV, it seems in fact partial not to explore the criminogenic dimension along with the victimogenic one.

Other studies have attempted to explore intimate partner violence by looking longitudinally at the risk processes implicated. Lussier et al. (2009) analyzed data from the prospective longitudinal Cambridge Study in Delinquent Development (CSDD) to examine to what extent IPV in mid-adulthood could be predicted by early childhood risk. Neuropsychological factors (e.g., verbal reasoning, verbal intelligence, etc.) and a criminogenic family environment (e.g., parental criminal record, low income, inadequate parenting, parental conflict, etc.) were measured between ages 8 and 10, while antisocial behavior was measured from age 8 to age 18 (e.g., overt behavior such as aggression and violence; covert

<sup>3</sup>Many risk assessment tools, in the fashion of professional structured instruments are available in the scientific community for professionals (see Zara and Farrington, 2016 for a complete review). For instance, a more general assessment of threat can be performed with the Danger Assessment (DA) (Campbell, 1995). For a more targeted assessment the Spousal Assault Risk Assessment Guide (SARA) (Kropp et al., 1995, 1999) is the most used. For a more specific assessment of obsessive search for intimacy risk, Stalking Assessment and Management (SAM) (Kropp et al., 2008) is currently one of the best instrument available.

behavior such as being deceitful and dishonest; reckless behavior; authority-conflict behavior). IPV was measured at age 48 using a self-report instrument completed by the participants' female partners. Findings suggest that perpetration and victimization rates were relatively high; violence was mostly mutual, and men were more likely to be victims than perpetrators. A criminogenic environment increased the risk of IPV by fostering the development of antisocial behavior and neuropsychological deficits, suggesting that IPV is never a private matter, but that the nature and quality of the relationship, along with individual and familial factors, contributed to IPV. Other studies using CSDD data advocated that processes of discontinuity and continuity between childhood and adolescent risk factors seem to increase the likelihood of future involvement in IPV by male partners (Theobald and Farrington, 2012), and between generations, though with some differences in risk factors for males and females (Theobald et al., 2016). Hence, the acknowledgment that IPV is not just situational, contextual and cultural is scientifically recognized.

Moving in this direction, the aim of this paper is to examine critically the individual and the social, the psychological and the relational, the cultural and the human aspects of what makes people violent, what encourages men to abuse women, and what fosters male partners to kill their female partners. The assumption is that a one-factor explanation i.e., patriarchal culture (Walker, 1989) and the reinforcement of its values (Ehrensaft et al., 2004; Dutton and Nicholls, 2005), cannot provide for comprehensive and exhaustive explanations (Noller, 2007) of what triggers IPV. IPV does not occur in a social, relational, and psychological vacuum: it is likely that the type and quality of the relationship might play a significant role in setting up opportunities to exercise aggression, and direct violence. It follows that an integrated approach might foster a broader and deeper understanding of what makes people violent, and what makes men abusive and aggressive toward women.

# The Present Study

This paper addresses intimate personal violence, and differentiates between risk factors that are at the basis of what triggers violence against women, and what sustains it in time, by an integrated and interdisciplinary perspective. The type, intensity, and length of relationships between female victims and male perpetrators will also be looked at, so as to be able to explore the extent to which IPV could contribute to femicide.

The focus of this study is on IPV, and whether its forms have changed over the years, and if yes, how. The temporal period of investigation is 46 years (1970–2016), which is a sufficiently long period to allow for exploring possible changes in the type of victims targeted; the type of relationships and the affective intensity between victim-offender; the dynamics involved in the perpetration of violence up to the escalation into femicide. Understanding VAW during this temporal period could offer insights into those early risk factors that influence and alter the quality of interpersonal and intimate relationships, and that may be informative to endorse preventive interventions before it would be too late.

# MATERIALS AND METHODS

## Hypotheses

IPV is likely to occur within a relationship between victim and perpetrator, and not independently of it.

It is assumed that the heterogeneity that featured in how violence is perpetrated depends at least on two aspects: the type of relationships between female victims and male perpetrators, and motives for being violent. These aspects affects how violence is performed, and the extent to which it escalates. While it is accepted that violence is a matter of individual (criminal) responsibility, it is also important to analyze the context in which VAW occurs, so as to be able to identify those risk factors, and relational and social conditions that make it possible for perpetrators to abuse and kill their victims.

It is assumed that not all VAW are gendered targeted. It is assumed that IPV is likely to be addressed toward specific victims with whom the perpetrator has had, or has, or wished to have (or have had) a relationship. The "relationship component" is likely to confer to the violence a particular overtone in the dynamic of killing, in the weapons used, and in the setting in which it occurs.

It is assumed that the closer and more intense the relationship between perpetrator and victim is, the more brutal the violence is, and the more likely that it will escalate into overkilling.

# Procedures

In order to meet all the ethics standards, the researchers in this study followed all possible procedures to ensure confidentiality, fair treatment of data and information, and to guarantee, at each stage of the research, that the material was treated with respect and discretion. The research protocol was organized according to The Italian Data Protection Authority (Garante per la protezione dei dati personali), nr. 9/2016, artt. 1 and 2 (application and scientific research purposes), and art. 4 (cases of impossibility to inform the participants, e.g., deceased people), and in line with the Italian and the EU code of human research ethics and conduct in psychology, forensic pathology and legal medicine.

While valid consent is a paramount requirement in scientific research with human participants, there in fact can be an exception to exempt researchers from obtaining consent. This is the case in which participants are deceased. This research lies in this specific situation because the sample comprises femicide victims. The data were archived both at the Institute of Legal Medicine, which signed a letter of intent with the Department of Psychology (University of Turin) to support this research, and at the Archive of the Central Morgue of Turin whose Director authorized data collection.

The research was assessed and approved by the Bioethics Committee of the University of Turin (protocol nr. 191414/2018).

All data were anonymized and made unidentifiable, and were also numerically coded for statistical purposes. The software package IBM SPSS Statistics version 25 was used.

Through a retrospective review of database of records from the Institute of Legal Medicine, and the Archive of the Central Morgue in Turin, this study identified all cases of women killed in Turin city and its outskirts from 1970 to 2016.

The sample comprises all female victims by male perpetrators. Features and characteristics of female victims and male perpetrators were taken into consideration. Specifically the types of relationships between victims and perpetrators, whether the relationship was abusive, the dynamics of the killing, and the motives that fostered the escalation into femicide were explored. Injury data were extracted by study investigators directly from the medical examiner records; when available the research team reviewed final autopsy reports and juridical records. Most data gathered contain information about age, race, profession, type and length of relationship between victim and perpetrator, social status, injuries, place of killing, reaction of the perpetrator after the murder, and a final determination of manner of death (natural, injury, murder, suicide), and cause of death. The cases of natural or suicidal deaths of women were excluded from the analysis.

To define injuries as excessive killing or overkilling, conservative scientific criteria were followed, so that if the victims sustained multiple injuries that went beyond those necessary to cause their death, this was accounted as overkilling. Jordan et al. (2010)suggest that overkilling involves multiple injuries resulting in one or more causes of death (i.e., multiple gunshots wounds) or multiple wounds distributed over two or more regions of the body (Salfati, 2003). The body regions were divided into three macro-regions: a. head, neck, face; b. torso and arms; c. legs. The categorization of data into motives of crime i.e., risk factors for intimate partner (IP) femicide, and then assessing whether the femicide comprised overkilling, were completed by two independent raters. When a discrepancy emerged, they discussed the case with the research group, and re-assessed it, until a better level of agreement was reached. The Cohen's Kappa statistic (Cohen, 1960) provides a measure of agreement between raters that takes into account chance levels of agreement, and it is appropriate for this type of data. Kappa for the category 'motives of crime' was =0.782, while for the category 'overkilling' Kappa was =0.778, suggesting, according to Viera and Garrett (2005), a substantial inter-rater agreement coefficient for both variables.

During data collection, every effort was made to record additional information related to the life of victims and of perpetrators so as to be able to reconstruct the events that led to the mortal accident as accurately as possible. Whenever possible data related to previous IPV or domestic violence incidents were recorded. Only a portion of data collected through file reviews are presented here.

#### Sample

The final sample involved in this study was composed of 275 women killed by violence in Turin, Italy. The violence was perpetrated by 260 males, indicating that in 95 % of cases the perpetrators (n = 247) killed only one victim, while in 11 cases they killed two victims, and in two cases three victims.

For some specific variables (e.g., nationality of the perpetrators and victims, criminal records, types of relationship, crime scenes, etc.) some data were missing; hence, some of the percentages do not sum up to 100%.

The victims were mostly Italian (n = 247; 89.8%), while the rest of the victim sample was comprised of foreigners. In 45.4% of cases they were unemployed (n = 104); in 28.8% of cases (n = 66), they were involved in a non-qualified job (e.g., cleaning job), and in 25.8% of cases (n = 59) they held a qualified profession (e.g., nursing, etc.).

While it was not simple to gather complete data on the 260 perpetrators, every possible effort was made to collect sufficient information to profile them. The vast majority of perpetrators were Italian (n = 177; 91.2%), while in 17 cases they were foreigners (8.8%). In 29.8% of cases, they were unemployed, while in 54.2% of cases (n = 97) they had an occupation. When employed, 46.9% (n = 84) of them were involved in a qualified profession (e.g., civil service, teaching, etc.), and 4.7% of them (n = 13) had an unqualified profession (e.g., trading, factory work, etc.).

When data were available, their criminal careers were taken into account by looking at whether they had been involved in any previous crime apart from the index offense of femicide they were convicted for. In 77.7% of cases (n = 202) it was the first time they were officially involved in an offense, with no previous convictions reported; the rest of the sample was made up of 22.3% (n = 58) of individuals who had previous convictions.

#### Analytical Strategy

Descriptive analyses with Chi-square and Odds Ratio (OR) were carried out to explore characteristics of the sample involved. The OR was calculated to identify which factors significantly explained motives for killing and which others predicted the dynamic for killing up to extreme killing, i.e., overkilling. The OR provides information about the existence, direction, and strength of an association between target and comparison groups regarding the likelihood of an event occurring (Farrington and Loeber, 2000). Where ORs are higher than 1, people with that particular attribute have relatively higher odds of offending than those who do not have this attribute.

# RESULTS

**Table 1** synthesizes the historical distribution of women killed in the Turin area every 5 years from 1970 up to 2016. As shown, the numbers of femicides decreased by years, showing a higher concentration of killing incidents up to 1996, and then a decrease. Dividing the crime period at the quartile cut off year (1996), into two macro-temporal categories, these data suggest that 73.5% (n = 202) of the female murders occurred from 1970 to 1996 (dated femicides), while 24.6% (n = 68) took place between 1997 and 2016 (contemporary femicides).

Contrary to the common opinion, which fosters the idea that women are more in danger in isolated places and at night, in 73.9% of cases (n = 198) the killing usually occurred in the house of the victims or of the perpetrators, and in 26.1% of cases (n = 70) the victims were killed either in an isolated place (e.g., country side or outskirt of the city) or in the car. Most of the deadly incidents (61.5%; n = 155) occurred during the day, between 06:00 a.m. and 5:59 p.m., while the rest of the victims (38.5%; n = 97) were killed at night, between 6:00 p.m. and 05:59 a.m.



*For five cases it was not possible to establish the year of crime.*

# Vicims and Perpertrators: Who Were They?

In this study, the victims had an average age of 46.15 (SD = 20.96), while the perpetrators had an average age of 42.96 (SD = 16.60). This difference was near to statistical significance indicating that the victims were slightly older than their offenders, t(448,366) = −1.821, p = 0.069, showing a small effect size, r = 0.17.

In 88.3% of cases (n = 228), the victims knew the perpetrators, while in only 11.6% of cases (n = 30) the perpetrators were strangers. Looking closely at those cases in which the victims knew their perpetrators, in 60.5% of cases (n = 156) they were involved in an intimate relationship; in 27.9% of cases (n = 72) they were acquaintances (e.g., neighbors, customers, etc.).

When victims and perpetrators knew each other, the average length of the relationship was 13 years (SD = 13.36; Min. = 0.01 month - Max = 62 years). Looking closely at these cases in which they had a relationship, in 43.1% of cases (n = 88) they never lived together, while in another 43.1% of cases (n = 88) they were living together and shared a house when the killing occurred. In another 13.7% of cases (n = 28) the killing occurred after their cohabitation was interrupted. In all cases, there was some evidence that an undergoing dysfunctional relationship mortgaged their life.

#### The Dynamic of Crime

In 75.3% of cases (n = 207) the perpetrator responsible for the killing was identified, found guilty, and convicted, while 24.7% of cases (n = 68) comprised unsolved and cold cases.

Information about the dynamic of crime and the reaction of perpetrators after the killing was also gathered. Data suggest that in 46.5% of the cases (n = 107) the perpetrators reacted by disposing of the victim's body, escaping, or denying the event or having any responsibility in it. In 53.5% of cases, perpetrators admitted the crime. In 13.8% of cases (n = 36) the murder was followed by suicide.

The types of lethal weapons used to kill the victims comprised sharp weapons and firearms in 57.5% (n = 150), while 42.5% of cases (n = 111) involved the use of improper weapons such as objects and bare hands. In those 59 cases in which the information was known, the firearms were possessed illegally or were not clearly registered to the authorities in 54.2% of cases (n = 32), though in most of these cases (n = 37; 62.7%) the perpetrators seemed to have handled the firearms with easiness, measured by the ratio between tagging the victim target and spotting successfully the target even at distance, and when in movement.

# Dying by Which Death?

The most common causes of death were gun injuries, stab wounds, and strangulation. The body regions mostly involved in the injuries were in 53.6% of cases (n = 140) the head, neck and face of the victims, in the 23.4% (n = 61) the upper-body of the victims (e.g., torso, chest, and arms), and in 23.0% of cases (n = 60) the injuries were spread over the entire body. In 56.3% of cases (n = 103) the victims did not manifest any defense reaction. In 40.7% of cases (n = 111) the victims were overkilled, implying that the attack caused excessive trauma beyond that necessary to cause death.

Understanding the motives behind the IPV that led to violent death was also explored. As emerged in previous studies (Campbell et al., 2007), motives are heterogeneous, and difficult to be identified precisely, especially because in most cases concurrent causes are at the basis of the crime. The motives behind the killing of the women involved in this study can be divided into six main categories. In 31.1% of cases (n = 80) the motive was crime of passion mostly related to an intense relationship that went terribly wrong; in 18.7% of cases (n = 48) the motive was related to family problems that involved ruminative thinking over difficulties such as insurmontable debts, a piece of news over the diagnosis of an incurable disease or losing the job. These events seemed to have preoccupied intensively and obsessively the perpetrator to the point of desperation: in these cases evidence of rows between the victim and the perpetrator sustains the assumption that the killing was not impulsive, but evolved into a reaction that from bad went worst. In 15.2% of cases (n = 39) the victim was killed as a consequence of another crime, and the motive was purely antisocial; in 14.0% of cases (n = 36) the motive was a predatory crime that involved also some sexual motivated killing. In 13.6% of cases (n = 35) the perpetrator acted under impulsivity and loss of control after a row or a refusal, and in only 7.4% of cases (n = 19) the perpetrator acted because of a mental disorder. In 18 cases it was not possible to identify the motives of crime.

It was significant to explore whether the types of relationship and the emotional intensity involved between victims and perpetrators (stranger vs. intimacy vs. affective vs. family closeness) were risk factors for an extreme act of killing that led to overkilling, and whether the overkilling could be differently explained by the variety of motives involved.

An analysis carried out with Chi-square and OR explored the likelihood of overkilling by types of relationships and motives of crime.

In 45.1% of cases (n = 82), when the victims knew their perpetrator the likelihood of being overkilled, with death being preceded by afflicting and brutalizing acts, was higher in comparison with the likelihood of being overkilled by a stranger (p = 0.002). The OR shows that the risk of overkilling almost quintuplicated when the perpetrator was known to the victim, rather than when he was a stranger. **Table 2** shows the significant results.

The risk of being overkilled continued to be significantly different depending on the types of relationship involved, and the more intense the emotional closeness between victim and perpetrator.

When the victim was killed by a partner or an ex-partner or a family member, with whom there was respectively an intimate or an affective bond, the risk for being overkilled was almost five times higher than when killed by a stranger. The risk of being overkilled was also significantly higher when there was a sort of relationship, albeit superficial, between victim and perpetrator, as in cases of neighbors, or colleagues or acquaintances, than in those cases in which the victim was killed by a complete stranger. No difference in the level of risk was found when comparing the degree of closeness in the relationship (intimate and affective versus acquaintance), indicating that in such cases the women were indifferently killed with a similar violent intensity.

Differences emerged when motives for killing were explored. When comparing crime of passion vs. family problems the likelihood for the woman of being overkilled was significantly higher. Similar risk emerged also when crime of passion was compared with antisociality (e.g., being overkilled as a consequence of another crime being committed) (see **Table 3**): when the perpetrator was emotionally related to the victim and acted out because of passion, it was more likely that he overkilled the victim. No significant differences emerged when the woman was killed because of predatory violence or because of impulsivity or because of a mental problem of the perpetrator.

In order to explore whether, in the course of the 46 years considered, the motives behind violence that led to death changed, crimes committed from 1970 up to 1996 (n = 190) were compared with those committed from 1997 up to 2016 (n = 62). Findings show that while overkilling was not significantly different depending on the period in which the woman was killed, some differences were found when motives were compared by the two macro-categories considered (multi-problematic relationships vs. antisociality). It was more likely that contemporary femicides (those occurred from 1997 onward) were motivated by multi-problematic relationships rather than general antisociality (OR = 0.576; 95% CI =0.315–0.1.054). While this result is only near to statistical significance (p = 0.09), it suggests that interpersonal, intimate or family problems were likely to drive a person to kill especially if the perpetrator was emotionally closer to the victim, and some emotional turmoil was affecting their daily interaction.

No doubt these findings are only preliminary and that further studies are paramount to continue to explore not only why, but also when and how women become the victims of extreme violence. However, their message seems to be insightful, which is that a higher risk of being victimized and killed lies more in intense, though problematic, relationships, rather than in criminality itself.

# DISCUSSION

This study focused its attention on women victims of deadly violence between 1970 and 2016 in Turin. Findings show that over these 46 years of investigation there has been a diminished rate of femicides. Nevertheless, they seem to suggest that most victims were not an indiscriminate and accidental target of violence.

These findings are counter-intuitive. While the common perception indicates that women are thought of as being at greater risk when away from home, with strangers, at night, and in isolated places, in fact most violence seems to happen at home or in familiar surroundings. Research shows that violence does not occur randomly, as shown in the description of the ≪ideal victim≫ and the ≪ideal offender≫ (Christie, 1986): victimogenic factors should be taken into account because they seem to be an essential component in fostering violence, transforming and escalating it into femicide. Understanding these aspects becomes relevant for preventive and intervention purposes.

According with international studies (Bailey et al., 1997), these data show that the victims involved in this study knew their perpetrators, and that the motives behind the killing mostly lie in problematic relationships and in family problems that led to aggressiveness, persecutory thinking, and then killing, and overkilling. They also show that IP femicide was more likely the result of a deteriorating relationship and of fading respect and trust between partners, than an act of pure cold crime; IPV was indeed the recurrent factor for escalating into femicide.

These data seem not to sustain the hypothesis of a gender violence to explain these cases of femicide, because in most of cases the killing was motivated by interpersonal and intimate motives that were at the basis of the escalation into deadly violence.

Current findings should be interpreted in the light of a few limitations that should be taken into account.

It was difficult to identify those cases in which the motive was purely gender - when the woman was killed because she was a woman. This motive might have been behind those cases in which victims and perpetrators did not know each other, but it was impossible, with these data, to reach that conclusion. In those cases in which the perpetrator was a total stranger, it could be assumed that the deadly violence had been motivated by control, power and antisocial motives. This is, in fact, related to the fact that some predatory murders were acted with coldness, detachment, and with some systematic precision, as in those cases in which the perpetrator committed more than a murder.

Because it was not possible to gather first-hand information from family members about the quality of the relationship between victim and perpetrator, and from perpetrators about the motives behind the killing the evidence gathered explain only part of dynamics of the IPV that fostered the femicide.

Furthermore, it was impossible, with these data, to reconstruct exactly whether the violence was mostly unilateral (from man toward woman), and to identify the recurrent victimogenic factors that interacted with other factors to escalate into femicide. TABLE 2 | Overkilling by relationship types and affective intensity.


*Comparing coding* = *0; 1.*

TABLE 3 | Overkilling by motives for deadly violence.


Crime of passion (0) compared with: Mental disorder (1); Impulsivity (1); Predatory killing (1) showed no significant results.

*Comparing coding* = *0; 1.*

In most cases in which this information was present it emerged that violence was a by-product of the relationship, was dyadic, and was the end result of a dysfunctional interaction in which personal, familial, cultural, economic and social aspects contributed to the worsening of the relationship. More studies are needed to explore the interactive combinations of criminogenic and victimogenic factors.

Despite these limitations, studies such as the one presented here are insightful in so far as they acknowledge that there is not a more suitable scientific subject such as IP femicide to see the gap between the ≪law on the books≫ and the ≪law on the streets≫ (Zimring and Hawkins, 1997).

Many years of under prosecuted IPV incidents may have fostered an implicit license that it is somehow tolerable to be abusive, insulting, psychological humiliating or physically coercive toward one's one partner, and this is why in such cases the problem cannot be "simply" identified with the "crime" committed, and solved by the administration of the law and the conviction of the perpetrator. This study shows that only a minority of the perpetrators were involved in a criminal career, and killed the victims because of pure criminality. What especially constituted the problem was the process of deterioration of the relationship, and the building up of a pattern of interpersonal violence that filled up with intolerance, misunderstanding, control and disrespect, the gaps that divided the partners.

Contrary to the media that presents a quite alarming situation (e.g., the availability heuristic) (Kahneman et al., 1982), these data show instead a reduction in the number of femicide over time, which is coherent with other psycho-criminological studies (Puzone et al., 2000; Dutton, 2012), and criminal data in Italy and in the Western world.

However, these findings should not be interpreted as if VAW and IPV were not a problem anymore or did not exist. Rather these findings inform us that the nature of abuse and violence has changed. Over the years, VAW and IPV may have taken a more silent (e.g., psychological and emotional) and a less spectacular (e.g., killing) vest. This study also raises some further questions that deserve a specific investigation over whether changing life or breaking off contacts with the perpetrator does really interrupt victimization or whether it instead increases the risk of it. It may be that previous forms of violence and victimization, especially within dysfunctional relationships, remain a determining factor of continual abuse, though differently manifested. Certainly, many women continued being abused, also after distancing themselves from their perpetrators; this is so because abusive and violent relationships last a long time, even though they change shape, and proximity.

#### CONCLUSION

Despite all the emancipation characterizing Western society, women are still the target of IPV. In so many cases this violence is less physical, with the result that the process of escalation might be transformed, so that women are less likely to be killed, or if killing emerges, it occurs after a longer time of silent, "behind curtains", abuse in which they had been dominated and controlled.

What is the message to take home from this study? VAW and IP femicide require an interdisciplinary perspective. Scientific research advocates a multidimensional approach, which looks at the problem with an equifinality lens. Femicide cannot be seen only as the story of cultural violence against women, because the story of the deadly violence against the women involved in this study, as in many other studies around the world, recounts more a story of IPV; the unfortunate story of a specific victim in a specific relationship who became the attention for violent vent out and destructiveness. This differentiation is essential to avoid reducing or enlarging to culture something, which is instead, according to these data, especially relational, at times intimate, at times familial, and at other times social, though essentially all of these levels together.

# REFERENCES


In addressing this differentiation in this research, the interest was to avoid the Hamletian's error, which is imagining the story of Hamlet without Ophelia, though Ophelia could easily disappear because her story was thought as ancillary or as irrelevant without Hamlet (Edwards, 1979).

These findings demonstrate that most victims (Ophelias) and perpetrators (Hamlets) exist the way they are because of the way they interact, or do not interact, or interact with each other dysfunctionally, disrespectfully, and immaturely.

#### ETHICS STATEMENT

This study was carried out in accordance with the recommendations of the Italian Code of Psychologists and the APA Code of Ethics. The protocol was approved by the Bioethic Committee of the University of Turin. All data gathered and employed in this study were anonymized and made it unidentifiable according to the Italian privacy and data protection law that regulates the use of data for scientific research.

#### AUTHOR CONTRIBUTIONS

GZ and SG conceived, planned, organized, designed the study, and attained data. GZ analyzed the data and interpreted the results. GZ drafted the article. GZ and SG critically revised of the article.


Donati, P. (2011). Relational Sociology: A New Paradigm for the Social Sciences. London: Routledge.


**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 © 2018 Zara and Gino. 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.

# Domestic Violence in Separated Couples in Italian Context: Communalities and Singularities of Women and Men Experiences

Paola Cardinali<sup>1</sup> \*, Laura Migliorini<sup>1</sup> , Fiorenza Giribone<sup>2</sup> , Fabiola Bizzi<sup>1</sup> and Donatella Cavanna<sup>1</sup>

<sup>1</sup> Department of Educational Science, University of Genoa, Genoa, Italy, <sup>2</sup> ASL3, Family Mediation Center, Genoa, Italy

Relationship breakdown and separation represent a critical aspect in domestic violence. Few studies have investigated domestic violence in separated couples. Moreover, there is a need for a more in depth analysis of gender differences that could enhance the comprehension of the phenomenon. The primary aim of this research was to analyze, through a qualitative approach, which kinds of domestic violence are characteristic or major in separated couples in the Italian context, where this phenomenon has not yet been sufficiently investigated. Participants are 60 separated couples (mean age: M = 48; F = 44) who attended a Family Mediation Center. A descriptive study was conducted using grounded theory methodology. A brief narrative task was administered to both ex-partners separately. The transcriptions were analyzed using NVivo 11 software. From data analysis, some themes emerged regarding typology of domestic violence specific of the separation context and shared by both men and women. The analyses of gender differences showed that there is a gender specific experience of domestic violence. Results highlight that women narrate both physical and psychological violence, while men relate only psychological abuse focused on limiting access to children. We discuss these findings in relation to possible appropriate gender specific intervention and prevention efforts.

#### Edited by:

Luca Rollè, Università degli Studi di Torino, Italy

#### Reviewed by:

Alessandra Salerno, Università degli Studi di Palermo, Italy Dagmar Kutsar, University of Tartu, Estonia

> \*Correspondence: Paola Cardinali paola.cardinali@unige.it

#### Specialty section:

This article was submitted to Gender, Sex and Sexuality Studies, a section of the journal Frontiers in Psychology

Received: 05 June 2018 Accepted: 13 August 2018 Published: 03 September 2018

#### Citation:

Cardinali P, Migliorini L, Giribone F, Bizzi F and Cavanna D (2018) Domestic Violence in Separated Couples in Italian Context: Communalities and Singularities of Women and Men Experiences. Front. Psychol. 9:1602. doi: 10.3389/fpsyg.2018.01602 Keywords: domestic violence, intimate partner violence, gender communalities, gender singularities, separated couples

# INTRODUCTION

Domestic violence represents an important concern for society; it is a widespread problem with adverse health consequences for all members of the family system. It has been defined as a range of actions that include physical and psychological aspects. Domestic violence against adults can be divided into three main types: psychological, physical, and sexual violence (World Health Organization [WHO], 2002). Inside the psychological abuse we find intimidation, constant depreciating and humiliating, and some controlling behaviors, such as isolating a person from their family and friends. Other forms of control are about monitoring a person's movements and limiting their access to information or assistance. Physical aggression includes slapping, hitting, kicking, beating, and other violent behaviors. Sexual violence concerns forced intercourse and other forms of sexual coercion.

**94**

In the literature on domestic violence a lot of attention has been dedicated to different forms of intimate partner violence (IPV). IPV has been defined as a set of assaultive and coercive behaviors that includes threats, psychological abuse, physical aggression, and other hostile behaviors (Peisch et al., 2016). It occurs within an intimate relationship and shows consequences at physical, sexual, or psychological level and remains a prevalent global health problem (Catalano, 2000; Garcıa-Moreno et al., 2015).

Several studies have explored prevalence and determinants of IPV; according to Bucheli and Rossi (2017), attitudes toward men's violence and women's violence are correlated and can be due to the same factors. Pollack (2004) proposes a model about the intergenerational transmission of violence that is consistent with social learning theory (Bandura, 1969). The sociocultural perspective emphasizes the role of shared beliefs about gender roles and inequities in explaining differences in domestic violence between countries (Bell and Naugle, 2008).

Recent study shows that dehumanization reported by women represents a significant factor involved with partner abuse (Homa et al., 2017).

Intimate partner violence includes both verbal (e.g., insults, yelling, humiliation) and physical (e.g., pushing, shoving, choking) behaviors, that often tend to co-occur (Pepper and Sand, 2015). Psychological abuse comprises all devaluing or humiliating behaviors and forms of dominance and isolation (Cleak et al., 2018). Longobardi (2017), reporting the Centers for Disease Control and Prevention classification (Centers for Disease Control and Prevention, 2015), defines four main types of IPV: physical, sexual, stalking, and psychological. Physical violence has been defined as the intentional use of physical force to damage someone through behavior like scratching, pushing, shoving, or throwing. Sexual violence concerns sexual acts that are committed by another person without the consent of the victim. A special attention has been devoted to stalking, a "pattern of repeated, unwanted attention and contact that causes fear or concern for one's own safety or the safety of someone else (e.g., repeated, unwanted phone calls, emails, or texts; leaving cards, letters, or flowers, etc.)" (Longobardi, 2017, p. 2039). Finally, psychological aggression includes the use of verbal and non-verbal communication to damage and/or to control another person (e.g., humiliation; limiting access to transport, money, relationships; threats of physical or sexual violence; and control of reproductive or sexual health). Johnson (2008) refined IPV types to reflect dyadic patterns within couples which views one partner's use of violent and controlling behavior in combination with the other partner's behavior.

According to some authors psychological IPV may be more mentally damaging than physical aggression (Crick and Grotpeter, 1995; Coker et al., 2002; Hellemans et al., 2015). Regarding this, Pepper and Sand (2015) found that only the perpetration and victimization of psychological violence were related with the overall feeling of oneself as a problematic person.

# Intimate Partner Violence and Relationship Breakdown

Relationship breakdown and separation represent a critical context for the study of domestic violence. The separation for a couple is a stressful life event and is associated with increased negative mental health and health problems. Therefore, the separation could be considered a risk factor for IPV (Logan and Walker, 2004). Furthermore in addition to the stressors, psychological problems may be experienced during a typical separation, especially women leaving abusive relationships often experience health and psychological problems related to the violence during the relationship. Separated women are more likely to experience violence than married women, and it is most common for women to experience violence from expartner. In the study of family relationship, IPV and parental separation are both considered major potential problems for children's adjustment (Holtzworth-Munroe, 2011). It may be that violence follows separation, or the decision to separate is due to violence. International studies indicate that leaving a violent partner may increase the risk of more severe, or even fatal, violence. Indeed, the risk of violence increases during the process of separation when emotions are intensified (Cleak et al., 2018). In this process, destructive communication, such as throwing insults or bringing up events from the past, breeds strong relationship dissatisfaction. According to Johnston et al. (2005) study, the percentage of parents reporting domestic violence is higher among separating and divorcing parents than in the general population. In Beck et al. (2010) study, 85% of wives and 77% of husbands reported abuse (including emotional abuse and coercive control) during separation. Literature reveals that male partner violence or abuse is a statistically significant predictor of the female partner's decision to separate (Hardesty, 2002). IPV is one of the main reasons given by couples seeking divorce (Amato and Previti, 2003; Gravningen et al., 2017).

It is well established that homicide rates are higher for women who have separated from their partners than for women in ongoing relationships (Hotton, 2001), this heightened risk of homicide following a separation is not found for men (Johnson and Hotton, 2003). However, according to DeKeseredy et al. (2004), separation may prevent or reduce the likelihood of physical assaults and emotional abuse against some women by their former partners. Separation may protect women from control-motivated assaults or from emotional abuse (Babcock et al., 2004).

# Gender Differences in IPV

As already highlighted in separation context, some studies underline some differences in IPV according to gender. In the specific context of relationship breakdown, men could see women's decision about separation as a challenge, which makes them turn to violence as a mechanism to reestablish the culturally prescribed gender domination (Flake and Forste, 2006). According to Straus (2006), the most persistent and controlling forms of violence are perpetrated by men, this seems to confirm that IPV patterns could differ by gender. Also recent studies argue that women are not as violent as men and are

more likely to use resistive or defensive violence (Holtzworth-Munroe et al., 2010; Carney and Barner, 2012). According to Caldwell et al. (2012), gender has a significant role in IPV because it is highly correlated with power. However, past findings have pointed that men and women tend to have different patterns of reporting of IPV; in particular men tend to under-report their own IPV perpetration while women are more likely to underreport their IPV victimization (Ko Ling, 2011). Men were more likely than women to be reported as using violent behavior like pushing, clutching, shoving, dragging, and choking, all fairly serious violent actions (Melton and Belknap, 2003; Ross and Babcock, 2015). In the specific context of separation, literature underlines a significant gender difference in the proportion of men and women citing domestic violence as a reason for the breakdown of their relationship (Gravningen et al., 2017).

Some studies highlight relational nature of IPV: men seem to be engaged in violence perpetration against non-violent partners at higher rates than women. Women more frequently perpetrated violence and control behavior in relationships with violent and/or controlling men (Coker et al., 2000; Mennicke and Kulkarni, 2016). Some studies underline that one area that has yet to be sufficiently explored is whether men and women agree on the acts, behaviors, and attitudes that comprise IPV in general (O'Campo et al., 2017), this is even more significant in case of separation. For these reasons, in the present study, we aim to fill the gap in the literature about similarities and differences in women's and men's experience of domestic violence during the separation process.

#### Domestic Violence and Italian Context

In Italy, domestic violence is a widespread phenomenon. Domestic violence in Italy is a social reality at odds with the national ideology of family unity and cohesion. Perhaps this contradiction accounts for the scarcity of Italian research about IPV (McCloskey et al., 2002).

The National Institute of Statistics conducted a study in 2014 about domestic violence that provides some clues to its prevalence: 6.788 women suffered some form of physical or sexual violence during their lives. 20.2% of these women suffered physical violence, 21% sexual violence, 5.4% more severe forms of sexual violence such as rape and attempted rape. 13.6% suffered physical or sexual violence from partners or former partners (2.8 million), 5.2% (855,000) from current partners and 18.9% (2.44 million) from former partners. Most of the women who had a violent partner in the past left him because of the violence (68.6%). 41.7% of cases this was the main reason for relationship breakdown, for 26.8% domestic violence was an important element in the decision.

Separated or divorced women endured more physical or sexual violence than others (51.4% against 31.5%) (National Institute of Statistics, 2014). However, no data are available about men.

Experts have brought attention to the complexity and specificity of domestic violence associated with divorce. This topic needs to be investigated with particular attention to be contextualized with the mediation practices. Handling mediation cases with a history of domestic violence is one of the most controversial issues in the field of divorce mediation (Ballard et al., 2011; Pokman et al., 2014). However, it is an important topic because a significant number of separated couples, engaged in mediation intervention, report IPV and abuse (Rossi et al., 2015). Currently, there is a great deal of variation in how cases with IPV are handled by mediators. Some programs exclude violent cases from mediation, others simply conduct mediation as usual (Holtzworth-Munroe, 2011).

While in United States, associations provide significant guidance about case of domestic abuse that appears in mediation intervention, in Italy there are no explicit guidelines. However, one of the first laws that introduced the intervention of mediation is the Law number 66 of 1996, which reformed sexual violence and also took into account domestic violence; it suggests the intervention of a family mediator to protect the family relationship. In 2001, the Law number 154 about measures against violence in family relations introduces the express possibility for the judge to suggest mediation to hostile partners. With the Law number 54 of 2006, family mediation has been formally recognized as one of the tools that the judge can indicate in the treatment of cases of separation. This law provides that the judge, with the consent of the parties, can postpone the adoption of measures to allow spouses, using experts, to undertake a process of mediation to reach an agreement, with particular reference to the protection of moral and material interest of the children.

The analysis of family relationships in the Italian context should be made, taking into consideration the transformations of recent decades. The popular portrait of Italy as a country in which "family matters," and the insistence of personalities with high public visibility on the importance of family integrity, are not matched by separation and breakdown rates. The transformations occurring in family relationships, on a psychological and social level, indicate a widespread and pervasive "fragility" of relations and their meaning. With regard to marital instability, National Institute of Statistic's study about separation conducted in 2015 in Italy underlines that there was a substantial increase in the number of divorces that reached 82,469 cases (+57% compared to 2014). It is important to remember that in 2015, for the first time in Italy, two important regulatory changes concerning the dissolution of conjugal unions (law no. 132/2014 and law no. 55/2015) became operational. Much more moderated, and in line with the trends in previous years, is the increase in separations (91.706, +2.7%) compared to 2014. At the time of separation, husbands are an average 48 years old and wives 45 years old (National Institute of Statistics, 2015). However, the fragility of family living appears to constitute an existential condition, strongly connected with the uncertainty of modern society. In most of the Western world, a small number of people characterizes the nucleus of many families, especially in the urban context. This may constitute a risk factor for family isolation implying reduction of cultural, relational, and economic resources (Canvin et al., 2009). Moreover, families are particularly vulnerable to transitions and changes, particularly with respect to instability and precariousness of relationships. Furthermore, gender relations represent a modern challenge for the family, which remains primarily organized according

to cultural determinants that define gender characteristics and differences. Some studies underline the matrifocal element of the Italian family unit. Women are perceived as devoted to family tasks and manage the housework, and men earn the income (Evertsson and Nermo, 2004). Rania et al. (2015) propose an image of paternity in Italian context slowly changing and redefining: new fathers seem to be more involved in the care of the children but mainly in recreational and executive activities, whereas mothers have a more active and organizational role than fathers. The family dynamic, particularly with minors, benefits a more stable structure (Migliorini et al., 2011, 2015) and continuity of relationships, even within the current dynamic of family setups (Garfinkel et al., 2001).

# Aim of the Current Study

No studies were found that reported IPV analyses among separated couples in Italy. The present work aims to increase the knowledge on:


# MATERIALS AND METHODS

# Participants

Participants are 60 separated couples. The average age of men is 48 years old and 44 years for women. The majority of the participants are graduates (55% of male; 43% of female) and employed (40% of men are workmen, 30% of women are office workers). Before breakdown, 70% had been married, while 30% cohabited. The average duration of the union was 12 years. All couples have one or more children.

# Materials

Because of the lack of existing research on this topic, we chose a qualitative study design. In recent years, psychosocial researchers have become increasingly aware of the need to improve qualitative methods in studies to understand the phenomena from the point of view of those who experience the situation. In addition to collecting demographic data (age, gender, educational level, current employment status), during their first meeting with the operator of Family Mediation Center, participants were asked to complete a narrative task. In accordance with the methodology already used in previous research (Tani et al., 2016), participants were requested to think about the history of their relationship, and briefly describe the main characteristics of the relationship with their ex-partner. Researcher through specific questions introduced the narrative: "Could you speak about the relationship with [partner's name] in your own words and without my interrupting you with any questions or comments? What kind of person [partner's name] is? How are you getting along together?"

The task is a stimulus that can facilitate the reflexive function. The participants had to exercise his/her awareness on themselves, on their ex-partner, and on the relationship between them. They also have to operate an integration between the emotional level and the cognitive level; between sensations emerging during the story and memories.

# Procedure

The project has been presented to the couples that began the mediation intervention. Participants were asked to fill out a brief socio-anagraphic schedule and informed consent. We include only couples in which both partners agree to participate. All narratives were audio taped during the first meeting and verbatim transcript. We chose to collect data in a Family Mediation Center in a medium-sized city in the northwest of Italy. The Center provides mediation services to divorcing or separating parties who have been court- or self-referred. All participants took part on a voluntary basis.

# Data Analysis

A grounded theory approach (Glaser and Strauss, 1967) was selected for the present study. We use as prevalent the objectivist approach because of the descriptive and explorative nature of the aims. The transcripts were analyzed with an iterative process of collecting and examining data (Charmaz, 2005). Data were compared from common teams using NVivo11 software. The narrative transcripts were coded privately and independently by two researchers using a codebook, and coding scheme for emerging themes or recurring domains of meanings across the narratives (Lofland and Lofland, 1995; Rossman and Rallis, 1998). All disagreements' were discussed, and a code was agreed on. The software was used to organize the coded statements into nodes containing similar concepts and hierarchies of categories and subcategories. The data analysis generated some graphical representations about the main topics. The quotes inserted in the results were chosen from narratives to best represent the core emerging themes. The quotations were checked carefully to ensure that the meanings were preserved in the form that they were presented by the participants.

# RESULTS

The analysis of the narratives in separated couples underlines some forms of domestic violence. We organized these materials in three main aspects: (1) the domestic violence experienced by both partners, (2) the domestic violence experienced exclusively by women, and (3) the domestic violence experienced exclusively by men.

As regards domestic violence, narratives highlight some characteristics common to the two groups. Both men and women reported domestic violence related to psychological abuse. In particular narratives analysis brings out seven sub-categories: limiting access to friends, oppression, verbal abuse, yelling, threats, slandering, and humiliating, that are briefly described in **Table 1**.

Below we present some selected quotations to illustrate the main categories emerging from narrative analyses. We chose both

TABLE 1 | Categories of domestic violence present in both women and men narratives.


men and women's citations to better underline the conformity in meaning.

Both men and women complain of limiting access to friends, often associated with irritability in the partner:

I could not meet females ... my friend lost her husband and I could not even invite her home, because all my female friends were sluts but all her male friends were perfect (M., man) I have to be careful if I talk to someone, a friend, he understands badly, that is ... he gets nervous (G., woman)

A second element that both reported is the sense of oppression from ex-partner and their family:

I always felt high level of suffocation (P., man)

I did not feel free to make choices because his parents, his father and his mother, they were very pressing ... that is, they gave advice that then they turned into obligations (A., woman)

Verbal abuse includes both insults, both oral violence that affects the ex-partner in her fragility:

He said me that I was a bitch, with statements such as . . . woman of shit, with statements such as bitch ... insults, on insults, on insults, all in front of the child (P., woman) She wrote me some messages "I have a family and you have not" (E., man)

Another common category is yelling:

Her phone calls, her screams and these things make me sick, she was yelling and this hurt me (R., man)

For a stupid thing he raised his voice very strongly, he yelled (L., woman)

Men and women report to be under threat:

He intimidated me, he intimidated also my children, he sent photos (V., woman)

She said to me: "Look . . . your dog has finished eating, or you give me money, or you buy him food or I allow him to starve to death, do you know that?" (M., man)

Slandering comprises false and defamatory statements perceived by partners:

I meet some friends, and they say that he goes around saying that I'm a whore, in front of my daughter (V., woman)

It is been more than ten years since she accuses me that . . . I'm alcoholic, she called all my friends to say this (R., man)

The last category of domestic violence presented in both narratives of men and women is Humiliating:

I woke up one morning that there was this girl to sleep at home, I was in the children's room, when I woke up, he (the ex-partner) was in bed with her, he was comforting her ... but you cannot go to console another woman at home! With your wife! In underwear to console another woman ... humiliating! (I., woman)

She continued to argue that she was not my daughter, but I am the father, so ehm . . . I insisted "Do the DNA test and then you'll find out what will happen!". Why do you humiliate me so much? (L., man)

While men and women may experience common domestic violence behaviors, there are also some important differences. Below we present the results related to the second aspect highlighted by the analysis: the domestic violence experienced exclusively by women. **Figure 1** shows a graphical representation that summarize the Domestic Violence categories present only in women experience; it comprises both psychological and physical abuse.

Psychological abuse narrated by women comprises different categories: limiting access to money, limiting access to work, restricting movement, stalking, dehumanization, constantly criticizing. We describe these issues by quoting some sentences from women transcriptions.

Women narrated violence related to limiting access to money, that created a condition of dependence, as described in the words of this woman:

I was economically dependent on him in the sense that I was not free to do shopping and ... I could not buy things without previous authorization. (L., woman)

Behind this dependence is hiding the request to do something to access the economic resources:

I could not buy even a underwear, all things that I want to buy should be inside an exchange agreement ... If I did housework, if I did something for him and his family, then it was possible that he approved my shopping or that he decided to buy me something. (G., woman)

This limitation contributes to women perception to be not equipped to face the social reality:

He had the management of the woman, that is ... I do not have a contact with real world, I do not know what is a bill; I do not know anything about these things, he has always done everything. (G., woman)

The Limitation of access to work is a strong reason for hostility:

We began to have fights for the money, because I started working and he told me that he brought me here (from my country) and that I had to give him money back. (E., woman)

Women describe the Restriction of the movements as a psychologically violent act of control:

He managed to confiscate even my house keys and my phone because I cannot leave the house". (D., woman)

In the women narratives emerge the story of some episodes of Stalking:

He followed me, controlled what I did, not only controlled me: check at all the people who stood next to me, that is really the impossible. (P., woman)

When he called me twelve times to day, fifteen times a day, not answering the phone meant that he invaded everywhere. (D., woman)

The narratives of women revealed also Dehumanization, in particular objectivation and animalization:

At home I was just used, like ... not like a human being. (L., woman)

It made me feel like a non-person; I didn't have more my personality, I was a zombie. (M., woman)

He took the keys of the house, he locked the door and he told me: "now you stay there for a while!"... honestly I felt like ... a package. (P., woman)

He treated me worse than a dog. (N., woman)

Finally, in women's verbalizations, the Constant criticism is perceived as a violence that threats personal identity and beliefs. The criticisms refer to different aspects of the person, as reported by this woman:

He said me that I was disgusting, I dressed badly, and . . . in his opinion I did not behave well in everything, from silly things to important ones. (R., woman)

These critiques seem to undermine these women, as it clearly emerges from the following sentence:

He was constantly telling me that I did not know how to do things, he told me that I was not pretty enough, that I was not good enough and then . . . there was a moment that . . . this caused me distress. (T., woman)

Some women point out that the criticism appeared associated with behaviors of little relevance:

He, for a trifle, for a light on, for an overturned sugar, for a stupid thing he raised his voice very loudly, he did not say bad words, but had some little phrases that hurt me: "you're brainless!". (L., woman)

This constant judgment does not seem to end with the separation, but rather increases:

"he was always judging me, has always judged me, he has always blamed me, and still now he continues to make it more and more than previously". (C., woman)

The presence of physical violence emerges exclusively in the narratives of women. This kind of violence is articulated in different forms: hitting ("he puts his hands on me three times"),

choking ("he puts his hands on my neck"), spitting ("he spits in my face, I consider it an extreme cowardice"). Physical violence is often narrated in association with the verbal violence as evidenced by this phrase:

He kicked me and he started insulting me (R., woman).

Physical violence generates a sense of fear in the women who have suffered it:

He is a violent person who has put his hands on me so many times and, therefore, he frightened me (S., woman).

Fear is so much ... fear of even yes to be . . . to suffer certain things again ... (D., woman)

Domestic violence present only in men experience is represented in **Figure 2**.

For men, the main violence is related to limiting access to child. They attribute this to a sudden decision:

Suddenly she decided that I couldn't see the baby. (G. man)

In the verbalizations of men emerge the idea that the ex-wives are not only involved in limiting contact in everyday life:

The child saw me and he attempted to greet me; but she positioned herself in the middle. (Y., man)

Today he told me that she is going to take the child from school when I was already in agreement with the child that I would take him and then we would go to play football. She is uncooperative in any event (L., man)

But, in men's narratives emerge the idea that ex-wives aim precisely to eliminate father figure in their son's life:

She also wanted that I renounce to my parental authority, I absolutely could not ... because M. (the daughter) is ... she's a part of myself and therefore, I could never and ever give up on M. (G., man)

Another father says:

She is doing everything to distance me from the child ... I asked her to see I. (the daughter) . . . I think this is my father's right. (F., man)

Men perceive this behavior as revenge:

Depriving a daughter of the relationship with her father . . . I think that this is her revenge toward me. (A., man)

In addition to the control exercised on the relationship between self and children, men report a limitation imposed also on the meeting with the grandparents and then access to the paternal lineage:

She constrains me not to take the child to the grandparents, I cannot understand it! As her mother and father are G. (the son) grandparents the same is for mine! The child has the right to see her parents same as to see mine! (E., man)

#### DISCUSSION

This work provides an original contribution to the field in order to understand the complexity and the characteristics of domestic violence associated with the separation context and to explore the specific gender differences regarding this topic.

The findings suggest that there is a common area of domestic violence perceived by both men and women and that concerns psychological aggression. In this area, consistent with World Health Organization [WHO] (2002) and Longobardi (2017) there are present some categories related to the use of violent verbal communication with the intent to harm or to exert control over another person. Men and women emphasize different forms of verbal assault and the use of intimate knowledge for degradation. This form of destructive communication could be considered specific of separation context in which violence occurs with continuity. Previously search identified emotional abuse as the most shared form of IPV (Karakurt and Silver, 2013), this area appears to be present both in men's and women's narratives also in the context of separation. They narrate to be subjected to threats, and exposed harm inflicted on victim's pets, it can be emotionally abusive, causing distress to both humans and animals (Faver and Strand, 2007).

The results of this work also show that there are two main gender differences to consider.

The main difference that emerges from the analysis of the transcriptions of men and women is related to the presence of physical violence suffered exclusively in the narratives of women. Research on IPV in women has mainly paid attention to their victimization, for very valid reasons (Straus, 2006;

Houry et al., 2008; Holtzworth-Munroe et al., 2010). In line with previous researchers women have often been considered to be the predominant victims and men the perpetrators of IPV. This could be coherent with traditional Italian family pattern in which women are devoted to child and housework and are considered weaker than men. Past findings highlight that severity, motives, and impact of IPV may be due to a gender asymmetry. Men often initiate and perpetrate more severe IPV which leads to more serious consequences or injuries (Ko Ling, 2011).

In the narratives of men no episodes of experienced physical violence appear. This confirms the literature trend that focuses on male-to-female violence, while overlooking female-to-male violence.

Specifically, regarding psychological violence experienced by men and women our results highlight a more complex scenario. Women identify a wide variety of types of domestic violence suffered, while men recognize only a few behaviors. A possible explanation is about social desirability. Men have to maintain their position in society. Face-to-face reporting of IPV behavior may induce shame, guilt, and embarrassment, which possibly lowers the likelihood of disclosure of such violence (Felson and Paré, 2005). So men may have trouble reporting certain behavior.

Women complain about the violence that affects control (money, work, movement) and aspects that undermine the identity (dehumanization and criticism).

In the context of separation more than in other condition controlling a battered person's access to work and financial resources can directly affect their possibility to separate. Men should implement also other form of violence referred by women, as a tactic to insure compliance. Minimization, denial, and blame destabilize the credibility and identity of battered/abused individuals. This appears particularly significant in relationship breakdown because people could loose their certainties.

Gender difference in this kind of context develops a reflection on the fact that also women are perpetrators of violence even if in the literature this perspective is less treated.

In our study, men reported a specific domestic violence perpetrate from women: the limiting access to meet children. This violence includes threats and/or behavior of exclusion from whole father ancestry. The narratives in this topic recall the concept of Parental Alienation (Gardner, 2002). Psychological studies focused on this specific syndrome considering it a form of psychological child abuse, that can lead to long-term traumatic psychological and physical effects in the children (Cavanna, 2013; von Boch-Galhau, 2018). Only recently a particular attention has been given to parent perspective (Balmer et al., 2018), underlying an intense psychological distress as a result of being alienated from their children. Data of our work suggest that specifically women participants use this type of violence, and that men experience significant exposure to parental alienation tactics. This finding is consistent with previous research (Bow et al., 2009).

A final reflection on the specificity of violence highlights that men use violence that affects the relationship with the outside (money, work), the possibility of autonomy (movements) and the definition of identity. On the contrary, women perpetrate violence in the area of relationships with their children. This seems in line with gender stereotypes, resurrecting or reinforcing the division between male-dominated public spaces, and the private spaces defined as women's domain (Scabini and Cigoli, 2006). Man threatens woman in the aspects on which she is weaker (e.g., women earn less, are less independent) and woman exercises a deprivation where man is more fragile. On the basis of gender differences in affect, behavior, and cognition (selfconstrual, emotional experience, selective memory), according to Gardner and Gabriel (2004) women would pay attention more on the relational aspects of interdependence in close relationship; this aspect could increase the women vulnerability to the effects of domestic violence.

# Limitations and Strengths of the Study

There were, of course, some limitations to this study, first the study does not consider the dyadic dimension as interpretative paradigm of the relationship between violence and gender as suggested by the Johnson and Ferraro (2000) studies. This should be useful to better understand the relational context surrounding IPV.

Second, data analysis in qualitative research is inherently subjective. We collect data in a Family Mediation Center, therefore, we may have missed the more severe levels of IPV.

Furthermore, the results are based on narratives consequently recall bias or unwillingness to report may influence the findings. Within the context of these limitations, however, our study suggests some possible practical implications for operators and procedures regarding this type of context to enhance programs that can empower women and men.

The strength of the present research was to analyze IPV in the context of separation, highlighting common and specific area. Within the narratives of the Italian couples, the common categories were emphasized and gender differences underlined. The narratives of the participants made it possible to highlight that the main differences are about the perception of physical violence only in women words.

As the separation can be considered a risk factor for intimate violence, our findings could be very useful because only a few studies have investigated the domestic violence in separated couples. Additionally, this study supports the need for further and more in-depth research on the gender differences in how IPV is used by men and women in different transition of family life cycle.

A greater understanding of similarities and differences in the conceptualization of domestic violence by gender can help to improve appropriate gender specific interventions and prevention efforts.

# CONCLUSION

Our findings show some types of experienced IPV characteristic of separated couples in Italian context and underline some gender specificity in men's and women's narratives about this topic. This first exploratory study raises many questions that are not sufficiently studied, but that need to be addressed about the different form of IPV carried out or suffered from

men and women in the context of relationship breakdown. If men and women differ in their conceptualizations of IPV, this suggests that interventions could be begin with the creation of a greater awareness and a common understanding about the problem, especially for non-physical violence. This common understanding is even more important and meaningful in the delicate phase of separation in which sharing is made difficult by the transition itself.

Jaffe et al. (2008) recommend that differing types and levels of IPV should be incorporated into case analyses and choices regarding mediation. The mediator must pay attention to the imbalance of power that could be generated between partners and he must also have sufficient power to intervene on couple dynamics, identifying forms of IPV. The necessary first step to ensuring the safety of mediating parties must therefore be detecting a history of IPV, and the present study suggests that we have to take into account of specificity gender matter (Ballard et al., 2011). An interesting result concerns a very particular form of violence that women exercise in case of separation and that regards the limitation of father's meetings with children that often causes parental alienation syndrome. This represents a central issue in mediation practice.

This article should contribute to the growth of the literature in Italy and provide interesting suggestion for other international context that are facing the domestic violence phenomenon. As well as data on domestic violence in the Italian context are collected exclusively on women, also in international studies the dominant portrayal of domestic violence does not cover men as victims, with rare exceptions (Costa, 2017).

This work suggests possible practical implications for researchers, clinicians, and procedures regarding this type of domestic violence, to enhance intervention programs. In fact, these findings could indicate two possible reflections and work areas: first, also men must be considered victims, and clinicians are called to promote communication and emotional expression

#### REFERENCES


about violence. Second, operators should promote empowerment paths for women to strengthen their identity, self-esteem, and self-efficacy as protection of the self from the abuses of the other.

# ETHICS STATEMENT

This research was conducted following the ethical norms stipulated by the Italian Psychology Association (AIP). Before the narrative task, written informed consent was obtained from all participants in accordance with the Declaration of Helsinki. It contained a brief explanation about the research and informed potential participants that the interview would be audio-recorded and the data processed and anonymized; it also assigned a code to each participant, in compliance with Italian Law on Privacy no. 196/2003. Research ethics committee has not yet established in the authors' institution when the research started, so an ethics approval was not required for this research as per the authors' Institutions' guidelines and national regulations.

## AUTHOR CONTRIBUTIONS

DC and FG conceived of the presented idea and supervised the findings of this work. PC developed the theory, performed the qualitative analyses, and wrote the manuscript with support from LM and FB. All authors discussed the results and contributed to the final manuscript.

# ACKNOWLEDGMENTS

We would like to acknowledge the psychologists Daria Finzi and Enrica Spadaccini who contributed to collect data and all women and men who shared their domestic violence experiences.




**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 © 2018 Cardinali, Migliorini, Giribone, Bizzi and Cavanna. 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.

# Intimate Partner Violence in the Golden Age: Systematic Review of Risk and Protective Factors

Eva Gerino, Angela M. Caldarera\*, Lorenzo Curti, Piera Brustia and Luca Rollè

Department of Psychology, University of Torino, Torino, Italy

Intimate partner violence (IPV) is identifiable as a major public health concern worldwide. The international literature highlights how this phenomenon is complex and transversal to all age groups. While the global population is becoming older, the scientific research about risk and protective factors related to IPV in the golden age is diverse, and the different findings of the various studies have not been systematized so far. Thus, in this systematic review, we aim to analyze the scientific studies that investigate the risk and the protective factors of violent dynamics between elderly couples. From the perspective of the theoretical frameworks and the methodological approaches used, we present the main conceptual themes that emerge. Following the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement, we review the articles that report the analyses of protective and risk factors of IPV perpetration. Our results indicate social support, help-seeking behavior, and the availability of community-based services addressing the issues of abuse as the main protective factors. The risk factors are related to economic conditions, belonging to an ethnic minority, cognitive or physical impairment, other conditions associated with cultural background and relational dynamics, such as intrapartner dependence and intergenerational transmission of violence and trauma, and caregiving stress. We discuss possible future directions of research to improve the understanding of IPV in the elderly population and the implications for the development of intervention policies at preventive and supportive levels.

Keywords: golden age, IPV, risk factors, protective factors, aging

# INTRODUCTION

Intimate partner violence (IPV) refers to violence between couples. The World Health Organization (WHO, 2012) defined it as "any behavior within an intimate relationship that causes physical, psychological or sexual harm to those in the relationship" (p. 1), including acts of physical and sexual violence, emotional-psychological abuse, and controlling behaviors. It is important to distinguish IPV from domestic violence (DV), a comprehensive term that includes many types of domestic abuse, such as child and elderly abuse in a household. The term "intimate partner" indicates that violence can be perpetrated by both men and women, regardless of age, marital status, or sexual orientation (Archer, 2000; Capaldi et al., 2007; Ali et al., 2016).

In their recent review, Ali et al. (2016) find different classifications of IPV in the scientific literature. Their work outlines three main perspectives used to classify IPV, according to the types of (1) abuse, (2) violence, or (3) perpetrators. Regarding the types of abuse, WHO (2002) describes physical, sexual, and psychological categories.

Edited by:

Grazia Terrone, University of Foggia, Italy

#### Reviewed by:

Roberto Baiocco, Università degli Studi di Roma La Sapienza, Italy Georgios Papanicolaou, Teesside University, United Kingdom

> \*Correspondence: Angela M. Caldarera angelamaria.caldarera@unito.it

#### Specialty section:

This article was submitted to Gender, Sex and Sexuality Studies, a section of the journal Frontiers in Psychology

Received: 05 May 2018 Accepted: 09 August 2018 Published: 04 September 2018

#### Citation:

Gerino E, Caldarera AM, Curti L, Brustia P and Rollè L (2018) Intimate Partner Violence in the Golden Age: Systematic Review of Risk and Protective Factors. Front. Psychol. 9:1595. doi: 10.3389/fpsyg.2018.01595

**105**

As for the distinction according to the type of violence, Ali et al. (2016) report two classifications. The first is proposed by Johnson and Ferraro (2000), who classify five qualitatively different types of IPV: coercive controlling violence (CCV), violent resistance, situational couple violence (SCV), mutual violent control violence, and separation-instigated violence. CCV is described as "a pattern of control and manipulation by a partner against their intimate partner" (Ali et al., 2016, p. 18), where the coercive partner may use one or a combination of behaviors, such as intimidation, coercion, control, and physical violence, to keep the partner under control. A victim shows violent resistance to violence from a coercive controlling partner. SCV is "defined as the type of violence between partners when an individual can be violent and non-controlling in a relationship with a nonviolent partner or a violent but non-controlling partner" (Ali et al., 2016, p. 18). Mutual violent control violence occurs when both partners are violent and controlling toward each other (Ali et al., 2016, p. 19). Separation-instigated violence occurs between partners who are in the process of separation.

Ali et al. (2016) refer to the second classification as the "Johnston Typology" (Johnston and Campbell, 1993). Johnston and Campbell distinguish among IPV types based on the motivations for the use of violence and outline the categories of episodic male battering, separation-engendered violence, male controlling interactive violence, and psychotic and paranoid reactions.

Regarding the classification of IPV according to the types of perpetrators, the Authors find that it encompasses different approaches. These range from gender to the perpetrator's psychopathology (Holtzworth-Munroe and Meehan, 2004) or physiological activation and emotional arousal (Jacobson and Gottman, 1998) to the type of violence understood as a behavioral response (generalized violent behavior, frustration response, and defensive behavior; Miller and Meloy, 2006).

In addition to the complexity of the many ways of categorizing the construct, IPV presents significant variations across the life span from adolescence to young adulthood (Johnson et al., 2015) and to older age (Policastro and Finn, 2017). Specifically, senior years comprise a critical stage of life, where IPV has particular implications for intervention strategies (Roberto et al., 2014).

The United Nations (2017) reports that the number of people over 60 years old more than doubled (962 million worldwide in 2017 vs. 382 million in 1980), and it is expected to become twice larger again by 2050. In the light of such increase, the study on events strongly related to physical and mental health in older age becomes crucial (Gerino et al., 2017). However, to our best knowledge, a systematic review about studies on risk and protective factors is still missing.

Many studies indicate the difficulty of obtaining clear figures about the prevalence of IPV among the general population (Devries et al., 2013). The magnitude of IPV is also underestimated (Crockett et al., 2015). Such difficulty is more evident when examining IPV in old age. For example, Policastro and Finn (2017) note that it is possible to observe IPV occurring among the elderly in two ways—either as IPV on growing old or as a new experience of violence, initiated after the partners have reached their older years. Drawing data from the National Elder Mistreatment Study (a survey of a representative sample of older adults from the US), the two researchers find that 1.7% of the participants report experiencing physical violence after the age of 60, and 3.7% report experiencing emotional coercive controlling behavior by an intimate partner. However, Policastro and Finn (2017) acknowledge the heterogeneity of the prevalence data, mentioning, among many others, the study of Acierno et al. (2009), who find that around 10% of the participant elders have experienced a form of abuse and/or neglect, and for over half of the physical mistreatment cases, the partners are reported as the perpetrators. In the sample recruited for their study, Rosay and Mulford (2017) show that 22.2% of the elderly victims reporting psychological abuse have been assaulted by an intimate partner and likewise for 27.4% reporting physical abuse.

Being involved in physical and sexual IPV, as either a victim or a perpetrator, is negatively associated with physical and mental health across the life span (Costa et al., 2015). IPV has greater health consequences for older women (Crockett et al., 2015) and a strong impact on emotional wellbeing and mental health (McGarry et al., 2016), being related to feelings of greater "worthlessness" or a loss of a sense of identity over time.

Given the aging of the global population mentioned above (United Nations, 2017), the study on risk and protective factors related to IPV in old age is one of the most important strategies for planning prevention programs in communities. However, to date, the scientific literature is varied in scope and content, presenting interesting and heterogeneous data, which need to be systematized.

# AIMS

Our study aims to present an up-to-date overview of risk and protective factors related to IPV in the golden age, focusing on the following:


# METHODS

#### Data Source and Search Strategy

We followed the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement's rationale (Moher et al., 2009). PubMed and EBSCO databases (PsycArticles, PsycInfo, eBook Collection, CINAHL Complete, Education Source, Family Studies Abstracts, Gender Studies Database, Race Relation Abstracts, Social Sciences Abstracts, Sociology Source Ultimate, Urban Studies Abstracts, and Violence and Abuse Abstracts) were searched by browsing titles, abstracts, and full texts to find eligible studies published in English, from the beginning to March 2018, with the keywords (IPV OR intimate partner violence) AND (aging OR older OR elder OR seniors OR golden age). Considering the recent development of this research domain, we did not insert time limits. The two independent reviewers' search on EBSCO yielded 986 results; 85 met the criteria and were selected. A second search on PubMed was performed to identify other papers; from the 597 results found (with a significant overlap with the previous search), only four papers met the criteria and were selected, totaling 89 papers screened by title and abstract. Subsequently, all 89 papers were screened by text, and from these, 58 papers provided specific information about IPV and aging. Eventually, the last phase entailed the selection of the papers specifically concerning risk and protective factors. Among those, 30 papers included the analysis of risk factors, while only eight included the analysis of protective factors. Considering that six papers included both risk and protective factors (Gil et al., 2015; Guedes et al., 2015; Yan, 2015; Roh et al., 2016; Souto et al., 2016; Teresi et al., 2016), the papers dealing with risk and protective factors that were included in this systematic review totaled 32.

Since we used databases containing peer-reviewed international journals, most of the studies included in the research were written in English. This implies that the research could miss hypothetical studies in other languages or those studies not published in peer-reviewed international journals. This issue is addressed in the Limitations section.

# INCLUSION AND EXCLUSION CRITERIA

The first three inclusion criteria for the papers were (a) the presence of the IPV construct, (b) an older population (average ≥ 55 years), and (c) the English language. We took the age of 55 as the cutoff because it is the lowest cutoff used in the literature (Zink et al., 2006; Poole and Rietschlin, 2012; Sood et al., 2016) to separate adulthood from the golden age; therefore, it is the most inclusive, except the study of Paranjape et al. (2009), using the age of 50. The cutoff age seems related to countries' specific demographic characteristics; for example, Adjukovic et al. (2009) use 65 as the cutoff because it is the age of retirement in their Croatian sample. We also included papers focusing on constructs connected to IPV, such as domestic violence, family violence, and elder abuse, when related to violence between partners and spouses, excluding those unrelated to intimate partner situations. Both qualitative and quantitative articles were selected in our attempt to show different approaches and methodologies regarding the subject matter. We included quantitative papers with different and cross-cultural kinds of populations. All the papers that emerged from the search with no direct link to IPV among older populations were excluded. Subsequently, we selected the papers that aimed to investigate the risk and the protective factors in this specific sample in order to systematize them in a table.

The review process is summarized in **Figure 1**, while **Figure 2** shows the growing number of published studies on the issue over time.

# RESULTS

The papers about risk and protective factors of IPV in the senior years are derived mainly from North America (9 from the US, including those dealing with ethnic minorities; 3 from Canada, and 1 from a sample of the North American indigenous population). It is remarkable that an article involves study participants of Korean descent in California and another is about Portuguese immigrants in Canada. Several papers come from Asia (1 systematic review about Asia, 4 from China, and 2 from Korea), while Europe seems less represented (1 article each from Croatia, Germany, and Albania and 2 from Portugal). IPV and aging in South America have also been studied in two samples (Colombia and Brazil). It should be noted that the Albanian, Colombian, Brazilian, and Canadian) samples have been studied and reported in the same paper (Guedes et al., 2015).

# Methodological Issues: Research Methods and Assessment Measures Used in the Retrieved Papers

The papers included in this systematic review present several methodological and theoretical differences, partially due to the coexistence of many branches and disciplines involved in these issues (e.g., nursery, criminology, psychology, social services). Furthermore, the research conducted in many areas of the world (e.g., Southeast Asia, North America, Europe, Africa) and in different social contexts (e.g., rural areas or immigrants and ethnic minorities) shows cultural differences although the selected construct (IPV) has been defined in the same way in cross-cultural papers. Among these, the most studied contexts and ethnic groups in the literature about IPV and aging are the Western (US) rural areas (Teaster et al., 2006; Brossoie and Roberto, 2015; Weeks et al., 2016; Roberto and McCann, 2018) and Asian elders (both residents in Southeast Asia and immigrants in Western countries) (Yan and Chan, 2012; Yan, 2015; Yan et al., 2015; Cheung et al., 2016; Han et al., 2017; Nam and Lincoln, 2017; Qin and Yan, 2018). The reason why so much literature has been produced about these cultures is made clearer later in this article. Considering all the papers (58 selected from the databases and noted in **Figure 1**), the most used assessment measurement adopted in quantitative research to investigate IPV among older people is the Conflict Tactics Scale (Straus, 1979), sometimes in a revised form (Sormanti et al., 2004; Sormanti and Shibusawa, 2008; Liles et al., 2012; Yan and Chan, 2012; Stöckl and Penhale, 2015; Roh et al., 2016; Nam and Lincoln, 2017). Other measurement tools, mainly used to investigate constructs similar to IPV, such as domestic violence and elder abuse, include the Multidimensional Measure of Emotional Abuse Questionnaire (MMEAQ, Hazrati et al., 2017), the Hurt, Insulted, Threatened with harm, and Screamed scale (Guedes et al., 2015; HITS, Miszkurka et al., 2016), or the Family Violence Against Older Women scale (FVOW, Paranjape et al., 2009). Researchers also utilize national and clinical services' databases to investigate the prevalence and correlates of IPV/domestic violence among elders (Salari and Maxwell, 2016; Sood et al., 2016; Policastro and Finn, 2017; Rosay and Mulford, 2017). The quantitative research papers are mainly cross-sectional; the difficulties in conducting longitudinal studies are probably due to the novelty of the issues, also caused by the victims' historical tendency to conceal their situation (McGarry

et al., 2016). In qualitative research, the most used tools are in-depth face-to-face interviews (Zink et al., 2006; Tetterton and Famsworth, 2011; Band-Winterstein, 2013, 2015; Eisikovits and Band-Winterstein, 2015; Yan, 2015; Weeks et al., 2016) and semi-structured interviews (Roberto and McCann, 2018). Band-Winterstein's papers are characterized by a phenomenological approach and discourse analysis. In qualitative studies, focus groups (Cianelli et al., 2013; Gil et al., 2015) have been used both to analyze and describe the phenomenon and to improve the participants' mental health, from a research-action and a community-ecological perspective (Brossoie and Roberto, 2015). The usefulness of including both qualitative and quantitative research is that the latter provides the prevalence and the risk/protective factors of the IPV phenomenon among the elderly, whereas the former can be helpful in explaining, somewhat clinically, the associations among the factors analyzed in quantitative research. A meta-ethnographic synthesis of qualitative evidence (McGarry et al., 2016, p. 2187) analyzes the following three fundamental themes of IPV in late life, aiming to show the variability of the phenomenon: (a) "unspoken and hidden" (b) "changing nature of IPV over time," and (c) "longevity of abuse." The first dimension refers to the hiddenness of the violence and the victims' inability to disclose and talk about IPV. The third dimension highlights the importance of the difference in the longevity of abuse; "older women may either have experienced IPV over the course of a long-term partnership or as a result of entering into new relationships later in life" (McGarry et al., 2016, p. 2188). The "changing nature of IPV over time" is also emphasized because many older women experience changes in the violence, for example, the transition from physical to psychological abuse in the relationship.

Beyond the methodological issues, it is important to mention the different theoretical and conceptual frameworks emerging from the research about IPV and aging. The first theoretical model to be produced is Sev'er's (2009) trilevel conceptual model of elder abuse (which includes IPV as a subtheme). This model aims to show the complexity of the issues involved in this phenomenon, including both personal and social characteristics and structural inhibitors and accelerators, highlighting the potentially different effects (inhibiting-accelerating) of specific circumstances. As shown by Roberto et al. (2014), two conceptual frameworks also emerge from the literature about IPV in the golden age—the feminist and the ecological models. The first

model analyzes power dynamics in late-life relationships with a feminist lens. It aims to show the gender-biased structure of latelife families, where older women are victimized and repressed by structural elements temporally prior to feminist instances and fights (e.g., Roberto and McCann, 2018 ["feminist life course perspective"]; Weeks et al., 2016). The second model is guided by the idea that it is impossible to conceive of a phenomenon such as IPV in late life without considering multiple layers, including context, cultural and societal values, family, community, and formal and informal social support. This conceptual framework is mainly oriented to intervention and research-action; in this context, it is possible to find a phenomenological approach as well (e.g., Teaster et al., 2006; Bonomi et al., 2007; Band-Winterstein, 2012, 2013, 2015; Poole and Rietschlin, 2012; Eisikovits and Band-Winterstein, 2015). More recently, Teresi et al. (2016) provide a conceptual framework to analyze elder abuse and IPV in late life. This complex ecological-cybernetical model includes stressful events (that produce symptoms); social structure and environment; presence or absence of primary and secondary prevention; psychological, social, and financial resources; and presence or absence of precipitating conditions (e.g., dementing illness and psychiatric or neurological diagnosis).

Despite the presence of these conceptual frameworks, many quantitative studies move away from a non-theoretical perspective, preferring to show empirical data without inserting them in theory-oriented research (e.g., Sood et al., 2016; Rosay and Mulford, 2017). To sum up, it is possible to identify two methodological–conceptual axes (dimensions) in the research about IPV and aging, as follows:


Although qualitative papers tend to be characterized by a theoretical framework, we prefer not to overlap the two dimensions. In fact, some quantitative articles are also theory oriented (e.g., Poole and Rietschlin, 2012; Miszkurka et al., 2016).

#### Protective Factors

The protective factors detected in IPV (see **Table 1**) during the senior years are reappraisal (1 paper), community (2), having friends (3), generally speaking, social support and networks (2), help-seeking behavior (1), protective interventions from childhood (1), self-esteem (1), coping strategies (life skills) (1), and eventually becoming an immigrant in Canada (1). Significantly, in all the literature about IPV in late life, only eight articles specifically deal with and analyze protective factors. Rather, it seems that research on this theme has proceeded in

#### TABLE 1 | Published studies on protective factors.


the direction of either identifying risk factors or extrapolating protective ones from the absence of the risk factors.

The protective factor, par excellence, emerging from the literature is social support (Zink et al., 2006; Gil et al., 2015; Roh et al., 2016), defined as comprising formal or informal social networks (community, friends, and social/protective services). Zink et al. (2006) also attach importance to the reappraisal of a victim's situation. The contribution of Zink et al. shows that all these factors can influence the effectiveness of actual coping strategies of IPV victims in late life. Social support seems to protect IPV victims from the pejorative loop. However, in another research on a particular sample (youngsters to older Korean immigrants and descendants in California) (Liles et al., 2012), social support does not emerge as a protective factor among women under the age of 40 but as a paradoxical risk factor (and it does not emerge as a protective factor for women over 55). This phenomenon is due to specific cultural values. For the authors, the traditional Korean values expressed in the social networks (mainly, patriarchal Confucianism) are not coherent with the protection of women's health from the adverse effects IPV. The relevance of the variability of cultural values to the effectiveness of the social network as a protective factor is also highlighted by Guedes et al. (2015), who find a protective effect of having friends in developed countries (where friends substitute for family ties) but not among Latin American and Eastern European participants. These findings are consistent with those of Souto et al. (2016) who identify "becoming an immigrant in Canada" (p. 12) as a protective factor. Although this could seem to be a powerful stressor that would trigger psychological and physical violence, the change of status (and state) actually allowed the battered women to be more protected by a different culture and system of formal and informal support.

Another protective factor emerging from the literature (Yan, 2015) is the help-seeking behavior of elderly victims of abuse and IPV. A recent paper (Teresi et al., 2016) analyzes elder abuse and IPV in late life in association with research on child maltreatment and abuse. Based on their collected data from studies about IPV across the life span, the authors claim that some intervention strategies could also be protective for elder people, such as legal programs; medical interventions; social services; training in violence prevention, assertiveness, and resistance; and skill enhancement and practice (even if these interventions are provided for children). This way, interventions in protection from violence since childhood can be interpreted as a protective factor for IPV in late life. Generally speaking, self-esteem and coping strategies, supported by knowledge and life skills, can be targeted to develop interventions and change models (Teresi et al., 2016).

#### Risk Factors

The risk factors emerging from the literature (see **Table 2**) are gender (7 papers), age (5), parental violence and intergenerational transmission of violence (7), low social support and isolation from the community (6), cognitive impairment, such as dementia and Alzheimer's disease (9), physical impairment (6), cultural values and factors (4), depressive symptoms (4), ethnic differences (3), immigration stress (1), unemployment and low income (3), personal factors, such as life stress (3), relational factors, such as living with an abusive partner (4), environmental factors, such as little privacy (3), verbal abuse (2), substance abuse by both perpetrator and victim (2), and caregiver stress (1). The research on IPV in the golden age provides much more literature about risk factors than protective ones. Despite this difference, most of the associations between predisposing factors and IPV in late life are still not completely explained, partly because of the inextricability of the relations between variables. For this reason, it seems useful to include papers where IPV is interpreted as a risk factor of severe symptoms, such as depression (e.g., Nam and Lincoln, 2017), to evaluate the circularity of the associations between other variables. In fact, a systematic review (Yan et al., 2015) reports depression as a risk factor of IPV in the elderly population, thus reversing the direction of the relation between the variables.

# Social-Demographic Characteristics: Gender, Age, and Socioeconomic Status

The most studied risk factor for IPV in late life is gender; the majority of the studies take for granted the association between elder abuse and IPV and their relevance to older women. We can state that for researchers, the involved population is primarily that of women. Some studies (Sev'er, 2009; Guedes et al., 2015; Policastro et al., 2015; Miszkurka et al., 2016; Santos et al., 2017) directly examine and show the prevalence of women as IPV victims although a study in the Croatian context indicates that older men can be victims of family violence as well (Adjukovic et al., 2009). The literature generally claims the necessity to adopt a gender-oriented approach (aligned with the feminist model) (e.g., Guedes et al., 2015), yet a recent review of the empirical literature about this topic (Roberto et al., 2014) highlights the absence of research about female-on-male violence.

Other risk factors associated with IPV in late life are related to demographic and non-demographic characteristics of victims. Age seems to be a relevant variable because it is negatively associated with IPV in the older population. The younger segment (aged 55–69) of the elder population is at higher risk of being involved in an IPV situation (both psychological and physical abuse). In contrast, older women (> 69) seem less at risk (Yan and Chan, 2012; Crockett et al., 2015; Beach et al., 2016; Santos et al., 2017), probably due to facts related to aging (e.g., the death of the abuser or separation/divorce) (Miszkurka et al., 2016). Race seems to be a risk factor of IPV in late life; in fact, older people (women) belonging to ethnic minorities seem more predisposed to IPV (Sormanti et al., 2004; Paranjape et al., 2009; Liles et al., 2012; Cianelli et al., 2013; Souto et al., 2016). The victims' unemployment (Paranjape et al., 2009; Yan and Chan, 2012) and low income (Guedes et al., 2015; Yan et al., 2015) are also risk factors for IPV. It is remarkable that low income and unemployment are generally more associated with females than with males. A low level of education seems to be a risk factor of IPV in late life (Han et al., 2017) although a high level of education is not necessarily a protective factor even if correlated with more victims' awareness about protective services (Stöckl and Penhale, 2015). In another research, no associations between low income and education and IPV are observed if adjusted for social support and living arrangements (Guedes et al., 2015). For these authors, the more relevant risk factor is the low level of support received from family members, as well as the isolation from the community that derives from it (Altman, 2017; Policastro and Finn, 2017).

# Mental and Physical Health

Another issue emerging from the literature is the high correlation between cognitive (e.g., Alzheimer's, neurological, and psychiatric diseases related to aging) and physical impairment and IPV in late life (Roberto et al., 2014; Yan et al., 2015; Beach et al., 2016; Miszkurka et al., 2016; Altman, 2017). This association could be interpreted both as IPV influencing mental and physical health (Qin and Yan, 2018) and as mental and physical health influencing IPV. Verbal abuse seems to be a risk factor and a predictor of physical abuse (Sood et al., 2016). Substance (particularly alcohol) abuse emerges as a risk factor of IPV among older couples (Liles et al., 2012; Miszkurka et al., 2016; Altman, 2017).

# Cultural Factors

Cultural beliefs and, generally speaking, cultural and societal values, emerge as relevant risk factors in late-life IPV (Souto et al., 2016). In fact, IPV seems to be more prevalent among ethnic minorities (yet at risk of other factors), often marked by powerful stressors (e.g., immigration) and machistic-patriarchal values (Sormanti et al., 2004; Paranjape et al., 2009; Liles et al., 2012; Cianelli et al., 2013). Late-life IPV also occurs more frequently in contexts where feminist trends have not arrived, such as in rural areas (Brossoie and Roberto, 2015; Weeks et al., 2016; Roberto and McCann, 2018) and in Confucian Asia (Yan and Chan, 2012; Yan, 2015; Yan et al., 2015; Cheung et al., 2016; Han et al., 2017; Nam and Lincoln, 2017; Qin and Yan, 2018). If aging itself brings more vulnerabilities for victims, it is probable that women remain subject to IPV and abuse due to the same social norms that impose the gender hierarchy (Crockett et al.,

#### TABLE 2 | Published studies on risk factors.


(Continued)

#### TABLE 2 | Continued


(Continued)

#### TABLE 2 | Continued


2015). Generally, racism (even when introjected) and sexism are ideological risk factors for IPV in late life (Poole and Rietschlin, 2012). Another important risk factor analyzed in the literature is the intergenerational transmission of violence and trauma. In fact, experiencing trauma in early life seems to be a predictor of acting with (or receiving) violence in late life (Guedes et al., 2015; Yan, 2015; Miszkurka et al., 2016; Policastro and Finn, 2017; Rosay and Mulford, 2017). The literature shows the importance of paying attention to every variable at stake; personal, relational (e.g., dependence on the partner by a victim with an impairment), and environmental factors can play a determining role in the phenomenon (Poole and Rietschlin, 2012). In fact, the research claims that caregiver stress and burden (in this case, the caregiver is the partner or the spouse) can also be risk factors for IPV (Gil et al., 2015). To our best knowledge, there is no published research yet about the differences (and different risk factors) between "IPV grown old" and new experiences of IPV in later life (Cheung et al., 2016).

#### DISCUSSION

#### Summary of Main Findings Protective Factors

This systematic review involving older romantic couples has made it possible to highlight how research has not yet clearly identified the protective factors for victims or couples in situations where violent dynamics are or could be manifested. Few articles (8 considered eligible for our systematic review) deal with this specific aspect.

The main protective factor that seems to be investigated by the limited research on this issue is social support. However, scientific publications do not seem to agree on this element, and the results seem non-homogeneous and univocal among the different cultures. Other possible protective factors could be help-seeking behavior and local/national services that deal with both assistance and information on the dynamics of abuse. This last aspect seems particularly significant for IPV prevention and the development of skills among the population during the entire life cycle. Regarding the risk factors that have emerged, the number of publications is greater. Similar to the literature on violence between couples, even studies on the elderly population show that older women are exposed to greater risk, while female-to-male violence is less explored. About this last result, we ask ourselves how much the "hidden number" affects the scarce findings about male victims of violence inflicted by their female partners. In fact, this factor implies an underestimation of the actual prevalence of the phenomenon (WHO, 2014, 2016) because of a men's certain reluctance to declare or denounce (and thus the tendency to minimize their involvement in violent dynamics between couples) IPV situations. For females as victims of violence between couples, an additional risk factor seems linked to economic conditions (low income and unemployment). The low level of education seems to be a risk factor although a high level of education is not a protective factor in itself.

#### Social-Demographic Variables

However, Guedes et al. (2015) question the results about the socio-demographic (economic and educational) status in their study of the role played by social support in the double role of risk and protective factors (Gerino et al., 2017). In fact, Guedes et al. (2015) point out that a low level of family support, loneliness, and isolation from the community increase the risk of being a victim of IPV; on the contrary, high levels of social support protect against the risk of suffering violence.

Other demographic variables that expose older people to a greater risk of IPV seem to be age (with "younger elders" aged 55–69 as the more exposed population) and membership in ethnic minorities, as well as cognitive and physical impairment. However, in the case of impairment, the need to study the causal relationship between the two factors (psycho-physical state and development of violent dynamics between couples) is particularly evident. Additionally, the analyzed studies show how substance addiction (particularly alcoholism) increases the risk of IPV. Depression also appears to be a risk factor, as well as a consequence of IPV. Even cultural beliefs, social values of reference (specifically machistic-patriarchal values), as well as racism and sexism, would have significant impacts on the manifestation of the IPV phenomenon.

#### Relational Dynamics

Finally, the relational dynamics between couples, with reference to intrapartner dependence, the family, the partners' development history (specifically the intergenerational transmission of violence and trauma), and caregiving stress (in a manner often consistent with older couples, where one partner is affected by physical or mental illness), are identified as risk factors. For the development of the phenomenon of physical abuse, previous experience of verbal abuse would constitute a specific risk condition.

#### Methodological and Application Issues

All these findings, in connection with the outcomes and the enrichments that further research may bring, could help (1) to target specific public awareness and information policies, and (2) to offer helping professionals (such as psychologists, social workers, etc.) recommendations on how to best address situations of particular risk or vulnerability. Among the outcomes of the present review, the results showing how different disciplines are involved in the analysis of the phenomenon and the increasing number of published studies are noteworthy. Together with the use of the tools for analysis and detection that are already present in the literature, it would be advisable to prepare validated guidelines for screening and managing these complex situations.

Regarding the methodological issues of the analyzed studies, it has become evident that these lack a clear definition and specifications of the IPV construct as critical elements. In particular, the literature seems to be missing an important differentiation of the studies through an analysis of IPV in the light of the phenomenon's complexity and evolution. The construct (as defined in the research design), especially in quantitative studies on risk and protective factors, seems affected by the lack of distinction by type (as explained by the WHO), involving motivations at the origin of the phenomenon and the kinds of victims and perpetrators (see the models reported in the introduction section), in the direction of theoryoriented procedural optics (a clear explication of the theoretical framework).

On an even more general level, in line with the observations of McHugh et al. (2005) and Bell and Naugle (2008), studies concerning domestic violence between couples are affected by some critical issues of both theoretical and methodological relevance. In fact, current theories—both sociocultural (feminist and power theories) and individual (social learning theory, background/situational model, personality/typology theories) fail to fully grasp the complexity of the factors involved in the phenomenon (McHugh et al., 2005; Bell and Naugle, 2008) and to be effective in terms of prevention and treatability (Bell and Naugle, 2008). For example, the feminist theory does not adequately explain women's violence toward their male partners, the presence of IPV between lesbian couples, and the lack of a significant relationship between sociocultural changes in attitudes toward the female gender (from more to less traditional) and IPV rates (McHugh et al., 2005; Bell and Naugle, 2008). This limitation implies the presence of bias in both the explanation of the violent interactions between couples and in the design of the variables to be included in the research on explanatory models, particularly in detecting antecedent or precipitating factors from a procedural perspective (Wilkinson and Hamerschlag, 2005; Bell and Naugle, 2008).

Greater attention is paid to contextual and cultural differences, but in our opinion, the need for cross-cultural comparative studies is increasingly evident, with emphasis on the issue of cultural minorities (e.g., in Western contexts). Nonetheless, always from a methodological perspective, it is noteworthy that no longitudinal studies deepen the knowledge on both risk and protection factors and how these vary over time. Again, it would be important to investigate which factors differentiate conflicting couples from those with IPV, as this would allow focusing on both the precipitating factors and the protection elements. In particular, in line with the findings of Roberto et al. (2014), it could be useful to deepen the differences between the situations where the abusive dynamic between couples is long-lasting (occurring in younger age and continuing up to old age), and those where IPV has its onset in the advanced phases of the life span. This would help both to increase the knowledge about the phenomenon and to design specific interventions. For this purpose and in general, it could be interesting to plan future research with mixed models (qualitative-quantitative studies) and with further attention to the peculiarities of the senior phase of the life span.

Taking as a reference Bell and Naugle's (2008) proposal, a useful explanatory model should include "multiple contextual units" (1101) that in turn involve relevant proximal variables. In fact, the authors' proposed model allows an even contextual analysis of the dynamics involved in violent episodes by using a micro- and a macro-analytic perspective and enables integrating the dynamic combination of multiple factors. This perspective, although still in progress, could be useful for a better understanding of the IPV construct involving the elderly, keeping in mind the necessity to elevate the complexity of the current

interpretation (Bell and Naugle, 2008), as well as following what is indicated by McHugh et al. (2005) on the postmodern approach.

#### Limitations of the Study

This review is limited by the availability of rigorous scientific publications on IPV in old age, particularly on both protective and risk factors involved in the development of this relational phenomenon in the life cycle. The study has several limitations. First, as mentioned in the methods section, we have exclusively considered the studies included in databases containing peerreviewed international journals and published in English; this means that we have not considered possible studies in other languages or published in other types of journals (e.g., not peer reviewed). Thus, it is important to interpret the data concerning cross-cultural issues with caution.

Second, this paper is a systematic review (not a meta-analysis) related to protective and risk factors for IPV, without considering studies on intervention strategies and their outcomes. We have included only the scientific contributions that specify the constructs that are the objects of the studies, or we have analyzed in detail the elements relevant to the focus of this work, aware that this does not exhaust the comprehension of the phenomenon in its complexity.

Finally, this review is limited to the older population and does not consider studies involving participants belonging to other age groups. This choice of field is both a restriction on the application (generalizability) of the observations presented in this contribution and an analysis in line with the need to study more deeply into the peculiarities of the conditions of the elderly. A comparative study of similarities and differences in IPV manifestation across different age groups would be informative. This could be a desirable perspective and a compelling challenge.

#### Implications for Future Research and Prevention Projects

Considering the aspects highlighted so far, an increasing interdisciplinary approach in the study of IPV among the elderly is recommended. This implies the need for a greater complexity of explanatory models, adequately structured for the elderly,

#### REFERENCES


which take into account the complexity and particularity of this phase of the life cycle, including the elements of resilience and the fragility of this growing population. Such studies could provide important leads for policy and action to prevent IPV, starting from the precursors of violence that could be changed by preventive intervention, taking into account the research evidence that violent dynamics result from the interactions among contextual, individual, relational, and situational factors.

We consider it crucial to further explore risk and protective factors. Studies should differentiate between those associated with the "elder" onset of the phenomenon and those associated with IPV relapse/recidivism, monitoring the situation of the person entering the cycle of violence and the trajectory of the development of the abuse condition over time, including the factors of possible remission. Longitudinal and cross-national studies on the senior years and IPV should be conducted as well. Both these issues will offer the opportunity to create a protective network for people in difficulty and perhaps to create ad hoc services addressing this problem. To achieve these goals, we consider of primary importance the increasingly structured and constant cooperation among the professionals involved in planning scientific research and the practitioners of the community-based services. This interaction would allow mutual enrichment and critical attention to both the theoretical framework and the results of clinical work in specialized services.

#### AUTHOR CONTRIBUTIONS

EG has been the responsibility of the design of the paper. EG and AC have been responsible for analyzing the systematic design and writing paper and in the creation of tables and images. LC searched for papers on databases and worked in their systematization. He collaborated in the process of paper writing. PB has been responsible for supervising the results and for the systematization of the papers. LR has the overall responsibility of the project and as supervisor of the paper, collaborated during the process of paper writing and in the finalization of it. All the author have been involved in the preparation of the discussions and conclusions.


**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 © 2018 Gerino, Caldarera, Curti, Brustia and Rollè. 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 Importance of Culture in Addressing Domestic Violence for First Nation's Women

#### Donna M. Klingspohn\*

Mental Health Recovery and Social Inclusion, School of Nursing and So Work, University of Hertfordshire, Hatfield, United Kingdom

Indigenous women in Canada face a range of health and social issues including domestic violence. Indigenous women (First Nations, Inuit and Métis) are six times more likely to be killed than non-Aboriginal women (Homicide in Canada, 2014; Miladinovic and Mulligan, 2015). Aboriginal women are 2.5 times more likely to be victims of violence than non-Aboriginal women (Robertson, 2010). These and other statistics highlight a significant difference in the level of violence experienced by Indigenous women to that experienced by women in the mainstream population in Canada. The historical impacts of colonization and forced assimilation are viewed as the main social determinant of health for aboriginal people in Canada, as they led to intergenerational trauma, with communities struggling today against discrimination, stigma, poverty and social exclusion. Most disturbing and damaging are the outcomes of domestic violence, mental health and addiction issues (Prussing, 2014). First Nation's women who want to leave a violent situation have limited access to helping services, as most are located in large cities and towns, far from remote reserves where many of the women live. Services were originally designed by and for the mainstream population. First Nation's women who manage to access these programs often find staff with limited cultural competence and program supports that have little cultural safety or relevance for them. Indigenous culture is defined in various levels of legislation as having a set of specific rights based on their historical ties to a particular region, with cultural or historical distinctiveness from the mainstream and other populations (Indigenous Peoples at the UN, 2014). In Canada, indigenous cultural beliefs are closely tied to belief in a creator, ancestors and the natural world, influencing their spirituality and their political perspectives (Waldram et al., 2006). Cultural safety, a concept that emerged in the 1980's in New Zealand, is viewed as an environment that is spiritually, socially, emotionally and physically safe for people; where cultural identity is recognized and valued through shared respect, meaning, knowledge and the experience of learning together. This paper will explore current evidence-based literature to determine if there is empirical evidence to support program policies and practices that reflect culturally safe, competent and relevant domestic violence services to address the cultural needs of Indigenous women in Canada.

Keywords: domestic violence, cultural safety, impacts of colonization and forced assimilation, intergenerational trauma, culturally specific vs. mainstream approaches to healing

#### Edited by:

Luca Rollè, Università degli Studi di Torino, Italy

#### Reviewed by:

Anna Lisa Amodeo, Università degli Studi di Napoli Federico II, Italy Susana Baron, Complutense University of Madrid, Spain

> \*Correspondence: Donna M. Klingspohn dkrgconsult@gmail.com

#### Specialty section:

This article was submitted to Gender, Sex and Sexuality Studies, a section of the journal Frontiers in Psychology

> Received: 06 April 2018 Accepted: 14 May 2018 Published: 04 June 2018

#### Citation:

Klingspohn DM (2018) The Importance of Culture in Addressing Domestic Violence for First Nation's Women. Front. Psychol. 9:872. doi: 10.3389/fpsyg.2018.00872

# BACKGROUND

European colonial ideology, reflective of racism and sexism, was used to repress and control indigenous peoples in Canada, and worldwide (New Zealand, Australia, South America and Africa), resulting in significant impacts to their health and social wellbeing (Bourassa et al., 2004).

Originally a matriarchal society, First Nation's women in Canada were respected and honored for their spiritual and mental strength; wealth, power and inheritance were passed down through mothers.

European colonists enacted legislation reflecting their patriarchal perspective, where women were not viewed as persons; recognizing only indigenous men as leaders of their communities. The respected and honored role of community leader was lost to aboriginal women for centuries (Cornet, 2001).

The passing of legislation in Canada such as the Indian Act of 1876, coupled with forced assimilation policies such as the residential school program, caused immeasurable harm, particularly for women. It has been claimed that South Africa's policy of apartheid was actually based on Canada's Indian Act (Saul, 2010).

The Indian Act denied women the right to possess land and marital property, unless they were a widow. However, even a widow could not inherit her husband's property upon his death as everything, including the house, went to his children. The Act changed slightly in 1884, with an amendment that allowed men to will their estate to their wives, but a wife could only receive the estate if the government's Indian Agent determined she was of "good moral character."

"Once the Indian Act was passed, the responsibilities of our men and women changed drastically. As a result of being confined to a reserve, our traditional men and women lost their responsibilities in using their strengths, either physically or mentally. Women were thought of as property by our O: gwe ho:we men who became acculturated into believing that they had to think like white men. The entitlement to status under the Indian Act itself enabled that to happen, wherein the male would gain status and his wife, and his children would gain his status."

Beverley Jacobs "International Law/The Great Law of Peace" (Jacobs, 2000, p. 108).

Residential Schools were the prime example of forced assimilation policies of the colonial government, where children were forcibly removed from their homes and communities to attend church administered residential schools. The real goal of these schools was to erase the traditional family and culture from the children and assimilate them into European colonial culture (Bombay et al., 2014). Instead of receiving a supportive educational experience, the children were abused—physically, sexually and psychologically—leaving lasting scars that have impacted generations of indigenous people. The Manitoba Justice Institute (1999) stated that "Residential schools laid the foundation for the epidemic today of domestic abuse and violence against Aboriginal women and children." Though assimilation policies were officially renounced with a formal apology by the Canadian Federal Government in 2008, new legislation and policy has been very slow to develop (Prime Minister Stephen Harper's Statement of Apology, 2008).

With the changes in economies over the years, men in Indigenous communities lost traditional jobs such as fur trading. With few economic roles available for replacement in remote communities, many men found their gender roles under attack, with more women becoming the economic providers. The resulting sense of social, cultural and economic insecurity has become a powerful factor in domestic violence (Douglas, 2013).

These historical impacts left First Nation's women living in poverty and socially excluded, facing multiple stigmas and experiencing domestic violence, compounded by intergenerational trauma. The residential school experience left many women with mental health issues including complex post-traumatic stress disorder, depression and substance misuse, as well as a suicide rate that is five times that of non-aboriginal people in Canada (Belanger, 2014).

# LITERATURE REVEALS

Critical evaluation and synthesis of the research literature is important to improve evidence-based decision making for policy and practice, by identifying valid evidence, bias, knowledge gaps and helping to separate fact from lore.

The objective of this review was to examine the evidencebased literature regarding effectiveness of mainstream, women serving agencies in delivering culturally safe and competent recovery services for aboriginal women experiencing domestic violence and also, a mental illness.

The main characteristics of the literature review, following Cooper's Taxonomy, helped systematically assess the quality of the literature and provided a guiding framework for subsequent research (Sipe and Stallings, 1996).

The main focus of the search was on historical impacts, correlated to significant health and social issues for First Nation's women and how they are currently addressed in policy and practice in service agencies.

The goal of the literature review was to reveal any particular cultural interventions that promote recovery and social inclusion, inform policy and practice to improve cultural safety and competency in recovery services and most importantly, improve the quality of life for indigenous women.

The innovative approach to this review was to find participatory research studies that allowed us to "hear the voices" of Indigenous women, to gain their perspectives on how they view the issue of domestic violence and identify strategies they would find effective while supporting their cultural beliefs. The returns of the review were subject to a thematic content analysis and separated into the identified emerging themes (see Appendix I).

#### SUMMARY OF FINDINGS

Campbell (2002) conducted a review of the research on mental and physical health sequelae of domestic violence, in a paper published in the Lancet, titled "Health consequences of intimate partner violence," concluding that the significant relationship of domestic abuse and mental health outcomes should be of concern and interest to clinicians as well as researchers.

Ramon (2015) echoed similar concerns in her paper "Intersectionalities: Intimate Partner Domestic Violence and Mental Health Within the European Context" demonstrating that the issue of domestic violence, concurrent with mental health problems, is a global issue that would benefit from the use of a recovery approach, changing the perspective of "victim" to "survivor. Jaffer and Mobina (1992) chaired a task force on Family violence in British Columbia. Their report, "Is Anyone Listening?" revealed nearly 30% of women in Canada experienced violence of a physical or sexual nature at least once in their relationship.

Sinha and Mair (2013) stated that British Columbia had the highest reported rates of violence against women of any province in Canada, indicating that not much has changed since the earlier report in 1992. One has to wonder what contributes to these statistics in BC.

In the article "Understanding the elevated risk of partner violence against Aboriginal women," Brownridge (2008) looked at data from two national surveys comparing violence against aboriginal women to violence experienced by non-aboriginal women. The comparison revealed that aboriginal women are 4 times more likely to be victimized and posit this statistic to intergenerational impacts of colonization and loss of culture.

Data from the Homicide Survey indicated that Aboriginal women were disproportionally represented as homicide victims, and similarly, victimization data indicate that Aboriginal women have higher rates of self-reported spousal and non-spousal violence (Homicide in Canada, 2014).

The Four Worlds Centre for Development Learning (2003) examined Aboriginal domestic violence in Canada, to map the nature and extent of the problem and uncover the facilitating factors of family, community, cultural, professional and governmental systems, to develop an intervention framework and strategies to reduce violence. Bopp et al. (2003) despite some progress, there are still gaps between the incidence of domestic violence in Indigenous communities and the capacity of those communities and agencies to systematically and effectively address the problem.

Given the increasing statistics of domestic and other abuse in First Nation's communities, coupled with significant mental health issues, it is evident that the current service provision is not addressing the contributing factors.

It is posited that a strong prevention framework, based in cultural strategies and recovery principles would help to decrease both the incidence of domestic abuse and the development of concurrent mental health issues. However, current services focus on crises intervention and stabilization, lacking dedicated funding for prevention, leaving women in a cycle of violence and despair, with little hope for recovery.

Cultural safety, a concept that emerged in the 1980's in New Zealand, is viewed as an environment that is spiritually, socially, emotionally and physically safe for people; where cultural identity is recognized and valued through shared respect, meaning, knowledge and the experience of learning together (Williams, 1999). Many Aboriginal people don't utilize mainstream health care services, not only due to their remote location but also due to a lack of trust. They experience stereotyping and racism, consequently view Western health care and other services as alienating and intimidating. In their paper examining First Nation's women's experience with mainstream health care, Browne and Fiske (2001) found Indigenous women were marginalized and disadvantaged by encounters of racism, discrimination and structural inequalities.

In 2010, Canada endorsed the United Nations (UN) Declaration on The Rights of Indigenous Peoples, with Article 22 of the Declaration stating that:

"States shall take measures, in conjunction with indigenous peoples, to ensure that indigenous women and children enjoy the full protection and guarantees against all forms of violence and discrimination." (page 9)

The Canadian government's Status of Women Committee released a parliamentary report in 2011, examining the issue of violence against Aboriginal women. They made recommendation that culturally appropriate services are vitally important, as there is a significant difference in the way domestic violence is viewed in the mainstream compared to aboriginal communities (Robertson, 2010).

# CURRENT APPROACHES/RECOMMENDATIONS

Mainstream services often reflect a feminist approach; women who are victims of violence are supported to leave their relationship, develop self-sufficiency and learn to take care of themselves and their children. First Nation's women believe they are married for life, therefore, do not envision an outcome where they leave their husbands. Their goal is to unite the family and put to rest the intergenerational trauma that has ruled their lives for centuries (Robertson, 2010).

In evaluating projects of the Aboriginal Healing Foundation, cultural safety was highlighted as critical to healing, and that relationships based on acceptance, trust and safety are the first steps in the healing process (Castellano and Archibald, 2007).

In the 2011 Community Guide to end Violence Against Aboriginal Women, there were five principles identified as best practice for cultural safety:


In the recovery framework "Honoring Our Strengths", culture is understood as the "outward expression of spirit," and revitalization of the spirit is a vital best practice to ensuring health and well-being among Indigenous people. Recognition of the importance of ceremony, language and traditions help to focus on strengths and reconnection with oneself, their history, family, community and land (Exhibition et al., 2010).

"Cultural safety extends beyond cultural awareness and sensitivity within services and includes reflecting on cultural, historical and structural differences and power relationships within the care that is provided. It involves a process of ongoing self-reflection and organizational growth for service providers, and the system as a whole, to respond effectively to First Nations"

#### (Honouring our Strengths, 2011, p. 8).

(The Indigenous Physicians Association of Canada the Royal College of Physicians Surgeons of Canada, 2009) believe that if mainstream health care is to be truly effective in improving the health of First Nations, Inuit, and Métis clients, it must provide culturally safe care. Any definition of cultural safety should include a strategic and pragmatic plan to change the way healthcare is delivered to Aboriginal people.

The traditional teaching of the Medicine Wheel guides a holistic healing process, examining the intersectionalities of domestic violence in relation to physical, mental, emotional and spiritual domains (Dapice, 2006). Many healers and indigenous services use the wheel's domains as the underpinning of their approach to achieving balance in well-being.

Culture is all about history and society; complex and dynamic, as opposed to a given set of beliefs or practices. Service providers must understand that the ongoing impacts of colonization have had a range of negative effects on First Nations peoples' health, rather than just having an understanding of specific cultural practices (Browne and Varcoe, 2006).

# CULTURE AND TRAUMA INFORMED PRACTICE

The (First Nations Health Authority, 2016) in their Policy Statement on Cultural Safety and Humility, recommends that all health services:

"Increase opportunities to educate health care professionals, those training to become health professionals, and others working in the health system on the history of First Nations health, as well as the concepts of cultural safety, and cultural humility and the relevance to First Nations health. Training to include: Recognizing the role of history and society, their impacts, and their relationship to culture in shaping health and health experiences of First Nations. This includes recognizing the role of trauma and offering trauma-informed care" (p. 15).

Trauma informed care and practice embraces a recovery focused, strengths-based approach, with an understanding and response to the impacts of trauma, where psychological, physical and emotional safety are paramount (for providers and service users) and provides opportunities for control, empowerment and recovery.

Given the history of intergenerational trauma experienced by Indigenous women, all service providers delivering services to address domestic violence must have a clear understanding of the traumatic effects of colonization, its impacts on indigenous women and their culture and develop competency in the types of cultural approaches that will be effective.

# ROLE OF PREVENTION

A review of the literature related to cultural interventions for domestic violence reveals the need for primary, secondary and tertiary prevention approaches (Shea et al., 2010).

Kiyoshk (2003) promotes primary prevention to reduce risk factors for family violence by integrating spiritual practices and ceremonies into family group counseling, utilizing cultural healing methods such as smudging, the talking circle, and the sweat lodge with Aboriginal men.

In a study on secondary prevention to stop risk factors becoming actual violence, Norton and Manson (1997) reported on the effectiveness of home visits to conduct weekly family violence groups that incorporated sharing of food and talking circles. Building a positive relationship with the counselor improved the outcomes for women taking part in the family violence program.

Willmon-Haque and Bigfoot (2008) reviewed literature on historical trauma and poverty, calling for early intervention in family violence to prevent further trauma to children. Services such as advocacy, promoting cultural awareness and culturally relevant services as well as involving communities in research are seen as vital to supporting mental health.

A recent development in prevention focuses on the responsibilities of the men. A program called (The Moose Hide Campaign, 2016) begun as a grassroots organization in BC to bring awareness to violence against women and children, has now expanded across Canada. Men are invited to join and stand against violence with a commitment to "honor, respect, and protect the women and children in their lives and to work together to end violence against women and children." Though the campaign began in the Indigenous community, non-indigenous men are welcomed, with the movement now having well over one million members (moosehide campaign.ca).

Battered Women's Support Services in Vancouver (BWSS, 2018) have successfully utilized youth targeted social media, to spread key prevention messages to stop violence against women and children (https://www.bwss.org/20-ways-youthcan-prevent-violence)<sup>1</sup> .

The Federal Government has a Family Violence Prevention Program designed to improve the safety of indigenous women, children and families. Funding is provided for prevention projects such as "awareness campaigns, conferences, workshops, stress and anger management seminars, support groups, and community needs assessments on and off reserve" (Government

<sup>1</sup>Battered Women's Support Services, 20 Ways Youth Can Prevent Violence Against Girls and Women. https://www.bwss.org/20-ways-youth-can-prevent-violence/

of Canada, Family Violence Prevention Program, 2017); (https:// www.aadnc-aandc.gc.ca/eng/1100100035253/1100100035254).

Annualized funds also support a network of shelters across Canada with services for women and children living on reserve.

Tertiary prevention looks at reducing of the worst effects of domestic violence. Heilbron and Guttman (2000) examined the outcomes for women who participated in a Healing Circle, providing a spiritual framework for group therapy. The women reported feeling empowered, safe, and comfortable. The healing circle sharing of Indigenous traditions and teachings provided "personal meaning" for the women, helping them develop a stronger connection to Aboriginal healing methods and ultimately their own cultural roots.

# CONCLUSIONS

Reflecting on the literature related to domestic violence and Indigenous women, there are indications of a clear need for the recognition and inclusion of culture in any helping services. Aboriginal women face multiple forms of stigma; being an aboriginal female, having a mental health issue, and being a victim of domestic violence (Hurtado, 1997). Statistics related to domestic violence show a disturbing difference in the incidence of violence and death comparing statistics of indigenous women to women in the mainstream. The key difference between the two groups is based in historical impacts. Colonial legislation and forced assimilation policies, such as the residential school program, led to significant traumatic impacts to the health and wellbeing of Indigenous people, and in particular, for women for multiple generations.

Moore (2001)states any recovery strategies developed for First Nation's women must be self-determined, working holistically toward reducing the negative impacts of all contributing factors, such as social exclusion, hopelessness and poverty, on the wellbeing of these women. This collaborative recovery approach would help to change the perspective of the women from victim to survivor.

Domestic violence services developed for the mainstream embrace a feminist model that does not fully consider the unique cultural needs of Indigenous women, nor are indigenous cultural principles clearly reflected in the services currently provided. Staff working in such programs must be educated in the history and culture of Indigenous women and be able to offer trauma informed practice with cultural safety and competency (Douglas, 2013). It has been proposed that a restorative justice model, embraced by First Nation's communities, is culturally a better fit, however there are concerns that it may lack the accountability that is needed to address the seriousness of the abuse (Ptacek, 2009).

Upon consideration of all recommendations found in the current literature, there is overwhelming evidence that effective domestic violence services delivered to Indigenous women must actually involve the women in the planning and delivery of recovery services that embrace their traditional beliefs and cultural principles. Planning of policies and program services should also involve key persons in the community, including Elders and tribal band council members. A range of Indigenous healing strategies should be offered to women and their families, such as smudging, talking/healing circles and sweat lodges to help reconnect to their ancient culture and address the issue of intergenerational trauma. Prevention programs, targeted to men and youth, may be effective in changing attitudes and behaviors toward indigenous women and girls.

Effective programs follow a cultural path, help one regain balance and share in what is viewed as the "circle of life." The use of healing traditions, such as those mentioned above, are designed to address the domains of the Medicine Wheel in planning holistic programs to reflect balance in the spiritual, emotional, mental and physical realms (Hunter et al., 2006). While we see some progress in developing culturally appropriate policy and programs, there is still a long way to go.

As a signatory in 2010 to the United Nations (UN) Declaration on The Rights of Indigenous Peoples, Canada must meet those declared obligations by developing effective services to end discrimination and violence toward Indigenous women and children (United Nations General Assembly, 2007).

A Message from the Elders:

"Women, like Mother Earth are life givers and nurturers of our children, families, communities and nations. By gathering our Indigenous women, we are stepping into our traditional values, ceremonies, teachings and cultures; embodying personal healing and connections to our ancestors, and to future generations to come. With the guidance of our elders, personal healing helps us to hold our connection to Mother Earth, the Creator and all our relations."

> Bev Gillard, Cree Elder Chairperson of the Elders Advisory Council, Circle of Indigenous Nations Society in the West Kootenay and Boundary region (Province of British Columbia, 2016).

# SEARCH ENGINES

A wide range of were utilized to review databases such as CINAHL; MEDLINE; Wiley InterScience; ATSIHealth; ProQuest; EBSCO; Google Scholar; Voyager; EMBASE; PubMed. As well, academic databases searched were: Women's Studies International; iPortal for Indigenous Studies (University of Saskatchewan); OISE's Aboriginal Peoples Curricula Database (University of Toronto); Indigenous Research Resources (Okanagan College); Indigenous Foundations (University of British Columbia).

# AUTHOR'S NOTE

DK has worked in the field of Psychiatric Nursing in Canada for over 35 years. Her master's dissertation in mental health recovery and social inclusion focused on the need to hear the voices of Indigenous women in conducting research and planning for services to address mental health and domestic violence issues. The issues of cultural safety and competency have guided her work in this area.

# AUTHOR CONTRIBUTIONS

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

#### REFERENCES


# SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyg. 2018.00872/full#supplementary-material


**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 © 2018 Klingspohn. 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 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.

# Informal Caregiving Relationships in Psychosis: Reviewing the Impact of Patient Violence on Caregivers

Juliana Onwumere1,2 \*, Zheng Zhou<sup>1</sup> and Elizabeth Kuipers <sup>1</sup>

<sup>1</sup> Department of Psychology, King's College London, Institute of Psychiatry, Psychology & Neuroscience, London, United Kingdom, <sup>2</sup> Bethlem Royal Hospital, South London and Maudsley NHS Foundation Trust, Beckenham, United Kingdom

A modest association can be found between people with a schizophrenia spectrum diagnosis (psychosis) and perpetrating acts of violence. When a person with psychosis does engage in violence, it is their informal carers, when compared to those from the general population, who are more likely to be the targets, and violence will often occur within the family home. Despite the importance of carer support for improving patient outcomes, our understanding of how carers are impacted by patient initiated violence in psychosis remains limited. This paper reviews literature documenting the effects of patient-initiated violence in psychosis on carer functioning. The review comprised searches of Medline, PsychInfo, Embase, and Web of Science databases and the hand searches of reference lists from relevant published papers. The review was limited to English language publications from inception to 11th September 2017, and where carer experiences following reports of violence from patients with psychosis were specifically recorded. Data from 20 papers using mixed methodologies were reviewed. Patient violence in psychosis was linked to poorer carer outcomes, including carer reports of burden, trauma, fear, and helplessness. There is, however, a significant need for further studies to systematically quantify the impact and correlates of patient initiated violence on psychosis caregivers, and improve prevention.

Keywords: psychosis, violence, aggression, carers, families

His mother didn't want him to be reported to the police and I was sympathetic towards that. I decided we'd do it her way, and that was a mistake, it was mistake that she paid for.

(Ferriter & Huband, 2003, p555)

# INTRODUCTION

Schizophrenia spectrum disorders (psychosis) affect 7 per 1000 of the adult population with over 20 million people worldwide living with a diagnosis of schizophrenia alone (McGrath et al., 2008; World Health Organization, 2018). The disorders, with their first onset commonly occurring in young adulthood, are often long-term and highly burdensome (Whiteford et al., 2013). They are associated with a significantly reduced life expectancy (Hayes et al., 2017), stigma (Dickerson et al., 2002), and small social networks (Sündermann et al., 2014; Palumbo et al., 2015). Despite these challenges, many individuals with psychosis remain in close contact with informal carers from whom they receive valued support. Informal carers are a diverse group, but are mainly close

#### Edited by:

Luca Rollè, Università degli Studi di Torino, Italy

#### Reviewed by:

Abin Varghese, Bhopal Memorial Hospital & Research Centre, India Martina Smorti, Università degli Studi di Pisa, Italy

\*Correspondence:

Juliana Onwumere juliana.1.onwumere@kcl.ac.uk

#### Specialty section:

This article was submitted to Gender, Sex and Sexuality Studies, a section of the journal Frontiers in Psychology

Received: 25 April 2018 Accepted: 02 August 2018 Published: 03 September 2018

#### Citation:

Onwumere J, Zhou Z and Kuipers E (2018) Informal Caregiving Relationships in Psychosis: Reviewing the Impact of Patient Violence on Caregivers. Front. Psychol. 9:1530. doi: 10.3389/fpsyg.2018.01530

**126**

family relatives (e.g., parents, partners, siblings) of patients and predominately female. The evidence base confirms that patients with carer support can achieve superior outcomes compared to peers without. The outcomes are varied but include significantly lower rates of relapse and overall number and length of psychiatric admissions (Norman et al., 2005), and improved rates of mortality (Revier et al., 2015; Ran et al., 2016) and service engagement (Stowkowy et al., 2012).

Though it has proved beneficial for improving patient outcomes, the caregiving role can impact negatively on carer health and wellbeing (Perlick et al., 2005; Flyckt et al., 2013; Gupta et al., 2015). Common mental disorders, including depression and anxiety, are significantly elevated in psychosis carers compared to the general population (Hayes et al., 2015). Carers also report experiencing exhaustion, grief reactions, and sleep disturbance (Patterson et al., 2005; Onwumere et al., 2017; Smith et al., 2018). As part of their role, many carers have also been exposed to episodes of anti-social behavior from the relatives they care for, these include episodes of both verbal and physical aggression (Belli et al., 2010; Onwumere et al., 2014).

# PSYCHOSIS AND VIOLENCE

Societal concerns about mental health and violence will often peak in the aftermath of a reported random act of violence (e.g., homicide) committed by an individual with mental health problems. This tends to ensure that violence, and its risk assessment, remains an important issue for mental health professionals (Shopp, 1996). Contrary to common media stereotypes, people with psychosis are more likely to have a history of victimization experiences (Bebbington et al., 2004; Honings et al., 2017), including violent victimization (Dean et al., 2007a; Short et al., 2013; ten Have et al., 2014). They also have an elevated risk for self-harm (De Hert et al., 2001). However, people with psychosis are also more likely, than the general population, to perpetrate acts of violence, including homicide (Fazel et al., 2009; Short et al., 2013). The statistical association between violence and psychosis is frequently reported as modest (Coid et al., 2006; Taylor, 2008; Douglas et al., 2009; Fazel et al., 2009), and is particularly evident during the first psychosis episode (Nielssen et al., 2007; Spidel et al., 2010; Large and Nielssen, 2011), or in some studies, during the first year of problems (Meehan et al., 2006).

Data taken from a large scale study of first episode psychosis cases in England indicated that nearly 40% of patients were aggressive at first service contact and more than half were reported as being physically violent (Dean et al., 2007b). A systematic review and meta-regression analysis of first episode psychosis highlighted that 28% of patients were aggressive prior to service contact and 31% following contact with mental health services (Winsper et al., 2013). Similarly, as part of a smaller scale study of 34 adults attending a service for people with ultra-risk psychosis mental states, 38% were reported to have had a history of violent behavior (Hutton et al., 2012). Further, approximately one fifth of adolescents meeting diagnostic criteria for psychosis and attending a community based children and young person's psychiatric service in England were recorded as having a history of physical aggression (Khalid et al., 2012).

Parallel to investigating rates of violence in psychosis, much of the research attention in this area has also focused on identifying the purported risk factors and clinical correlates of patient violence (e.g., Swanson et al., 2006; Bo et al., 2011). For example, we know that patient violence in psychosis has been positively linked to several factors. These include: younger patient age (Dean et al., 2007b; Large and Nielssen, 2011; Coid et al., 2013); substance abuse (Coid et al., 2006; Fazel et al., 2010; Spidel et al., 2010); lower educational attainment (Large and Nielssen, 2011); poor vocational activity (Swanson et al., 2006), being from an ethnic minority group (Dean et al., 2007b; Coid et al., 2013), female gender (Swanson et al., 2006); male gender (Dean et al., 2007b); social difficulties (Amore et al., 2013); history of victimization (Swanson et al., 2006; Spidel et al., 2010), and patients with a forensic history (Large and Nielssen, 2011). Violence risk has also been linked to specific symptom clusters including mania (Dean et al., 2007b; Large and Nielssen, 2011); hallucinations (Swanson et al., 2006); delusional beliefs, particularly those related to persecution, being spied upon, and conspiracy (Joyal et al., 2011; Coid et al., 2013; Onwumere et al., 2016), or where the patient perceives personal threat and/or experiences thoughts that over-ride their sense of control (Chan, 2008; Nederlof et al., 2011). Disposition to anger (Nederlof et al., 2011), particularly where anger relates to delusional beliefs, is also linked to an increased risk of violence perpetration (Coid et al., 2013; Ullrich et al., 2013). Higher rates of violence are reported in individuals with untreated psychosis (Keers et al., 2014); before commencement of pharmacological treatments (Large and Nielssen, 2008; Nielssen and Large, 2010); in patients who are non-adherent with treatments (Witt et al., 2013), and in those with a history of involuntary and/or a greater number of psychiatric inpatient admissions (Large and Nielssen, 2011; Dack et al., 2013).

When we look at other factors, we know that positive links have also been observed between patient violence and coresidence with family members (Estroff et al., 1998; Swanson et al., 2006; Kageyama et al., 2015); poorer family relationships including patient reports of not feeling listened to by their family (Swanson et al., 2006), and attempts made by carers to establish behavioral limits with patients (Straznickas et al., 1993). A review of 4,168 patients with a diagnosis of schizophrenia suggested that patients, when compared to control groups, were more likely to come to the attention of police authorities through their involvement in family violence (Short et al., 2013). Further, in approximately one third of recorded adult domestic homicides in England and Wales, over an 11 year period, perpetrators were experiencing psychotic symptoms (Oram et al., 2013).

While only a relatively small proportion of the total global population will be diagnosed with a psychotic illness, and acts of violence vary in terms of context and severity, the importance of understanding the impact of violence on a victim has remained largely ignored in the psychosis caregiving literature (Solomon et al., 2005). This is despite the fact that this is another form of domestic violence or abuse. There is a paucity of data on reports of violence from patients with psychosis who have informal carers (Thompson, 2007), and scarce consideration of the implications for its assessment and management. The limited evidence base is surprising given the contribution to and importance of the caregiving role to patient outcomes in psychosis, and the large numbers of patients who live with or continue to remain in close contact with informal carers (The Schizophrenia Commission, 2012). Lifetime rates of carer exposure to patient violence varies but has been estimated in some studies to fall within the 50–60% range (e.g., Onwumere et al., 2014; Kageyama et al., 2015), with approximately one third of carers reporting incidents of violence in the preceding year (Kageyama et al., 2015). In addition, the risk factors associated with violence perpetration within the general population (e.g., substance misuse) are also elevated in psychosis populations (Hartz et al., 2014). Further, and perhaps most compelling, is the observation that carers, particularly those who are female and living with the patient (e.g., typically mothers), are more likely to be the identified target of violent acts compared to other family members and the general population (Nordström and Kullgren, 2003a,b; Nielssen et al., 2007; Belli et al., 2010; Ural et al., 2012). Carers are also more likely to sustain greater injuries (Nordström and Kullgren, 2003a). Whilst ∼8.4% of carers have issued legal orders (e.g., restraining orders) against the relatives they provide care for, following violence related issues (Solomon et al., 1995), the evidence suggests less likelihood of patient initiated violence directed toward carers ever being reported to law enforcement agencies (Nordström and Kullgren, 2003a).

# STUDY AIMS

Patient initiated violence in psychosis is an important problem for many stakeholders including family members, who are the common victims. Interpersonal violence is a public health issue that exacts a significant impact on individual wellbeing (World Health Organization, 2002). Efforts to prevent violence in family settings must commence with a more comprehensive and informed understanding of the subjective experience and impact on victims. The current paper therefore aims to review the literature on the reported effects on carers who have been exposed to violence from patients with psychosis. It specifically seeks to address the research question: What are the reported effects of patient initiated violence on carers' physical and psychological wellbeing? The research and clinical implications will be discussed.

# METHOD AND TERMINOLOGY

#### Design

A systematic review of the relevant literature with a qualitative synthesis of the findings.

#### Search Criteria

In accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (Moher et al., 2009), a search was undertaken of four electronic databases (Medline, PsychInfo, Embase and Web of Science) from inception to 11th September 2017. The search also included a hand search of the reference list of relevant papers to check for further applicable studies that had not been identified from the initial database review. The review was limited to: (i) studies where explicit links were reported between reports of patient initiated violence in psychosis and carer functioning; (ii) peer reviewed papers that were published in English language journals. Studies were excluded if they were review studies and those reporting data soley from non-psychosis patient populations (e.g., organic disorders like dementias), and/or psychosis conditions secondary to a primary disorder. Studies employing mixed diagnostic groups were excluded if psychotic disorders constituted <30% of the sample.

Studies ineligible for inclusion were review studies, and those reporting data soley from non-psychosis patient populations (e.g., organic disorders like dementias) and/or psychosis conditions secondary to a primary disorder. Studies that employed mixed diagnostic groups were eligible for inclusion if psychotic disorders constituted at least 30% of the sample. Given the interchangeability in the use of terms to reflect violence (e.g., aggression) (O'Callaghan and Richman, 2010), we purposively included a broad definition of violence to account for any acts of aggression toward an individual or property, designed to threaten or inflict harm, irrespective of reported severity. A detailed list of keywords and Medical Subject Headings (MeSH) were used (with applicable search truncations and wild cards) to maximize the search capabilities and relevant paper selection. The selected terms and headings varied according to the specific database. Search terms related first to "psychosis" (serious mental illness OR severe mental disorder OR schizophren<sup>∗</sup> OR schizo-affective, psychosis OR psychotic), "violence" (aggression OR violence OR abuse), and carers (caregiver<sup>∗</sup> OR carer<sup>∗</sup> OR famil<sup>∗</sup> OR relative<sup>∗</sup> OR parent<sup>∗</sup> OR partner<sup>∗</sup> OR spouse<sup>∗</sup> OR sibling<sup>∗</sup> ). The Boolean operator "AND" was used to combine the three primary search term categories.

#### Article Selection

The titles and abstracts of identified articles from the initial search were screened, independently, by the first two authors against eligibility criteria and to remove duplicates. Selected papers were read against inclusion criteria. Disagreement between the reviewers about a decision to include or exclude were resolved through discussion. The study selection process is outlined in **Figure 1**. Data from the selected studies were tabulated and presented in terms of author, publication year, study design, sample, and summary of the key findings.

#### RESULTS

The initial search strategy identified 4,498 articles, which were reduced to 2,513, following exclusion of duplicates. Twenty papers met full criteria for selection. See **Table 1** for summary of studies.

#### Study Origin and Design

The cited studies were international including six from Asia (Gopinath and Chaturvedi, 1992; Hanzawa et al., 2013; Hsu

and Tu, 2013; Chaturvedi et al., 2014; Kageyama et al., 2016; Varghese et al., 2016); four from the USA (Friedrich et al., 1999; O'Brien et al., 2006; Thompson, 2007; Copeland and Heilemann, 2008); three from the United Kingdom (Gibbons et al., 1984; Ferriter and Huband, 2003; Onwumere et al., 2014) and Australia (Vaddadi et al., 1997, 2002; Loughland et al., 2009), and two from Sweden (Kjellin and Ostman, 2005; Nordström et al., 2006). The remaining studies had their origin in Canada (Chan, 2008) and Switzerland (Lauber et al., 2003). All studies, with exception of one, were cross sectional in design. The O'Brien et al. (2006) study was a 3 month longitudinal study.

# Carer Demography

The total number of carer participants identified was 1,875 and most were female. In the 16 studies reporting this data, the composition of female participants in individual studies ranged from 39% (Gopinath and Chaturvedi, 1992) to 100% (Copeland and Heilemann, 2008). Only 65% (n = 13) of studies offered details on the mean age of carer participants. Where details were offered, carer participants were aged mainly in their early to mid-50s (Vaddadi et al., 2002; Thompson, 2007; Chan, 2008; Loughland et al., 2009; Hanzawa et al., 2013; Onwumere et al., 2014) or 60s (Ferriter and Huband, 2003; Lauber et al., 2003; Kageyama et al., 2016). There were, however, carer participants who were notably younger with their mean ages falling in the mid-30s (Friedrich et al., 1999; Hsu and Tu, 2013) and 40s (Vaddadi et al., 1997; Varghese et al., 2016).

Carer participants were heterogeneous in how they related to patients, although most were reported as being the parents. The composition of parents in individual study samples ranged from 27% (Kjellin and Ostman, 2005) to 100% (Ferriter and Huband, 2003; Nordström et al., 2006; Copeland and Heilemann, 2008; Hsu and Tu, 2013; Kageyama et al., 2016).

Carer participants also included siblings, who were sampled in 10 studies (Gibbons et al., 1984; Friedrich et al., 1999; Kjellin and Ostman, 2005; O'Brien et al., 2006; Chan, 2008; Loughland et al., 2009; Hanzawa et al., 2013; Chaturvedi et al., 2014; Onwumere et al., 2014; Varghese et al., 2016), and partners, who were reported in 9 studies (Gibbons et al., 1984; Gopinath and Chaturvedi, 1992; Vaddadi et al., 1997, 2002; Kjellin and Ostman, 2005; Chan, 2008; Chaturvedi et al., 2014; Onwumere et al., 2014; Varghese et al., 2016). Carer participants who were the children of patients were included in six studies (Vaddadi et al., 2002; Kjellin and Ostman, 2005; Loughland et al., 2009; Hanzawa et al., 2013; Onwumere et al., 2014; Varghese et al., 2016).

# Patient Demography

Homogeneous schizophrenia spectrum patient samples were employed in the majority of studies (75%, n = 15), with a further study using a psychosis prodrome sample (O'Brien et al., 2006). Only four studies employed mixed diagnostic samples where schizophrenia spectrum diagnoses ranged from one third (e.g., 31%—Kjellin and Ostman, 2005) to ∼80% of the overall sample (e.g., 78% Vaddadi et al., 2002). The additional diagnostic groups sampled were personality disorders, affective psychosis, bipolar affective disorder, and mood disorders (e.g., Vaddadi et al., 1997, 2002; Kjellin and Ostman, 2005; Thompson, 2007; Varghese et al., 2016).

# Clinical Setting

Just over half of the studies sampled carers of patients who were living in community settings (Gibbons et al., 1984; Gopinath and Chaturvedi, 1992; Vaddadi et al., 2002; Lauber et al., 2003; O'Brien et al., 2006; Thompson, 2007; Chan, 2008; Copeland and Heilemann, 2008; Loughland et al., 2009; Chaturvedi et al., 2014; Kageyama et al., 2016). Inpatient only samples were used in five studies (Vaddadi et al., 1997; Ferriter and Huband, 2003; Kjellin and Ostman, 2005; Nordström et al., 2006; Hsu and Tu, 2013) and three studies used mixed inpatient and community dwelling groups (Hanzawa et al., 2013; Onwumere et al., 2014; Varghese et al., 2016). The study from Friedrich et al. (1999) did not offer any information on clinical setting.

# Assessment of Violence

The operationalization of patient violence varied across studies. For example, in the Chan (2008) and Kageyama et al. (2016) studies, acts of physical, and psychological aggression were assessed. Three studies looked at acts of physical aggression only (Kjellin and Ostman, 2005; Nordström et al., 2006; Onwumere et al., 2014), with all remaining studies, but one, investigating verbal and physical aggression. It was unclear in the Copeland and Heilemann (2008)study whether verbal and/or psychological aggression was also included, in addition to physical aggression. The methods used to record data about patient violence varied considerably across studies and included patient medical records (e.g., Kjellin and Ostman, 2005; Thompson, 2007) and symptom rating scales (e.g., O'Brien et al., 2006); carer selfreport questionnaires (e.g., Nordström et al., 2006; Chan, 2008; Loughland et al., 2009; Hanzawa et al., 2013; Chaturvedi et al., 2014); carer semi-structured interviews (e.g., Vaddadi et al., 2002; Lauber et al., 2003; Copeland and Heilemann, 2008; Hsu and Tu, 2013; Onwumere et al., 2014), and combinations of the different assessment methods (e.g., Gibbons et al., 1984; Gopinath and Chaturvedi, 1992; Ferriter and Huband, 2003).

For the majority of studies, the assessment period focused on any episode of patient violence that had occurred since the initial illness onset (Gopinath and Chaturvedi, 1992; Ferriter and Huband, 2003; Lauber et al., 2003; Kjellin and Ostman, 2005; Hanzawa et al., 2013; Hsu and Tu, 2013). In three studies, the assessment review period was more limited and thus focused only on reports of violence that had taken place during the preceding 12 months (Chan, 2008; Loughland et al., 2009; Kageyama et al., 2016). Two studies focused on reports during the preceding 6 months (Varghese et al., 2016) or one month (Gibbons et al., 1984), while another prospectively measured violence over 3 months (O'Brien et al., 2006). A life time prevalence of patient violence was the focus in four studies (Vaddadi et al., 1997, 2002; Copeland and Heilemann, 2008; Onwumere et al., 2014). In addition to lifetime prevalence, Vaddadi et al. (2002) also recorded reports of violence in the last 12 months. Four studies did not share any information on the assessment period under study (Friedrich et al., 1999; Nordström et al., 2006; Thompson, 2007; Chaturvedi et al., 2014).

# Impact of Patient Violence on Carer Functioning

Fourteen studiesreported a positive link between patient violence and reports of carer burden (Vaddadi et al., 1997, 2002; Friedrich et al., 1999; Ferriter and Huband, 2003; Lauber et al., 2003; Chan, 2008; Hanzawa et al., 2013; Chaturvedi et al., 2014) including financial burden (Thompson, 2007), and emotional distress (Gibbons et al., 1984; Vaddadi et al., 1997, 2002; Hsu and Tu, 2013; Kageyama et al., 2016). For example, Vaddadi et al. (2002), in a sample of 101 carers, identified a positive relationship between carer reports of emotional distress on the General Health Questionnaire (Goldberg and Williams, 1988) and patient aggression. Positive links between patient violence and trauma symptoms in carers were described in two studies (Loughland et al., 2009; Hanzawa et al., 2013). Hanzawa et al. (2013) observed that carers of patients who had been violent also reported significantly higher levels of intrusion, avoidance and hyperarousal symptoms on a self-report trauma measure (i.e., Impact of Event Scale–Revised, Weiss and Marmar, 1997; Asukai et al., 2002). In contrast, the findings from Gopinath and Chaturvedi (1992) suggested that it was patient difficulties with self-care, inactivity and depressed mood that carers reported as being most distressing and not patient aggression.

In five studies, carers reported experiencing fear (Nordström et al., 2006; Copeland and Heilemann, 2008; Hsu and Tu, 2013), which included beliefs that their life was in danger (Loughland et al., 2009) and a fear of violence recurrence in the future (Friedrich et al., 1999). Data from Hsu and Tu (2013) qualitative investigation suggested that patient violence led to carer reports of feeling powerless and frustrated over their perceived inability to control patient behavior and effect positive change. Carers described making a deliberate choice to remain quiet, out of fear that their relative might retaliate (Hsu and Tu, 2013).

Three studies observed positive links between patient violence and expressed emotion, which included carer reports of patient focused criticism (O'Brien et al., 2006; Chan, 2008); hostility (Onwumere et al., 2014), and emotional over involvement (intrusiveness) (Chan, 2008). In the Vaddadi et al. (2002), patient violence was associated with carers reporting a poorer relationship between the patient and themselves.

# DISCUSSION

Despite the fact that carers can be integral to securing optimal outcomes for those with psychosis, violence, of any type, is likely to impact negatively on any family relationship. The importance of identifying and responding to carers' individual needs has now been recognized within several clinical treatment guidelines (NICE, 2014; Galletly et al., 2016; Norman et al., 2017). The current review suggests there have been few investigations that have purposively sought to directly and systematically record the outcomes for carers who have been exposed to patient violence in psychosis. Where outcomes have been identified, patient violence is seemingly linked to a wide range of negative carer outcomes that can include burden, emotional distress, fear, and high expressed emotion (EE). These findings are offered against a body of literature which attests that poorer carer functioning and negative caregiving relationships; for example, high EE, are linked to patient poorer outcomes that include higher rates of relapse and hospitalization (Bebbington and Kuipers, 1994; Cechnicki et al., 2013; Hesse et al., 2015).

The current findings, which are based on a heterogenous group of studies, provide a useful template from which to explore, in greater detail, the carer experience of patient violence in psychosis. Though most carer participants surveyed were female, this picture is consistent with the profile of carers typically reported in psychosis research (e.g., Smallwood et al., 2017; Smith et al., 2018). In addition, the studies were diverse in country of origin and continent sampled, and their respective systems of health care provision (e.g., National Health Service; Health insurance).

It is noteworthy that the majority of studies reviewed had sampled carers at a single time point. The importance of this data should not be underestimated since it provides a much needed starting base to address pertinent questions on violence in caregiving relationships in psychosis. However, given the often repeat nature of violence, there is need to identify the potential longer-term implications of violence exposure for carer health, family outcomes, and service provision. For example, trauma presentations in carer groups are gradually receiving more research attention (Kingston et al., 2016). Exploring pathways between violence exposure, carer trauma reactions and coping styles (e.g., Loughland et al., 2009) would be beneficial and supported by multi time point studies.

Reports of patient violence toward caregivers are likely to be an underestimate, particularly when data are based upon selfreport, which can be influenced by issues of social desirability (Swanson et al., 2006). Stigma and efforts to avoid adversely affecting the care and public image of their relative are likely to impact carer willingness to disclose abuse in their relationship (Kageyama et al., 2015; Onwumere et al., 2016, 2018). Further research should aim to incorporate additional sources of data; for example, accident and emergency data. Likewise, future studies exploring patient violence in caregiving relationships should seek to assess its broader impact on carer functioning and relationships, and highlight potential pathways through which patient violence may disrupt these.

# LIMITATIONS

The review was designed to offer a platform and direction for further studies but had key limitations. The selected studies tended to lack detailed information on carer participants such as weekly hours spent with patient, and the exact nature of their caregiving responsibilities; factors that could have provided more context to the findings. With exception of one, all studies were cross-sectional, thus precluding conclusions about causality. Few studies had solely set out to report on patient violence and its impact on carer outcomes, which limited the amount of data interrogation one could undertake. In accordance with the Cochrane Collaboration Tool for risk of bias (Higgins and


(Continued)


September 2018 | Volume 9 | Article 1530



(Continued)


NR

**135**

Altman, 2008), the studies were deemed at high risk for reporting bias due to homogeneity in methods used to record outcomes on carer impact (i.e., self-report). Finally, it is unclear to what extent the current findings are specific to psychosis caregivers or are observed in other severe mental health caregiver groups such as bipolar affective disorder. Future reviews may wish to extend the population group under study to include other diagnostic groups, which can help in the process of determining the scale of the problem and assessing the need for psychosis specific or trans-diagnostic responses.

#### IMPLICATIONS

The cognitive model of caregiving responses in psychosis highlights the importance of carer appraisals about the patient and the illness on overall outcomes (Kuipers et al., 2010). Though much has been written about issues of domestic violence and mental health (Howard, 2012), there has been a noticeable neglect of these issues when they are reported by informal carers of people with psychosis. For policy makers, greater awareness of the different family settings that interpersonal violence can and does occur, and the additional unique and complex needs faced by informal carers is required. A consideration of nuanced and targeted informant campaigns, specifically designed for those in caregiving roles and with an understanding of the broader issues should be given. More research is required to improve our understanding of the impact of patient violence on carer outcomes and the implications for their caregiving relationship. The data should help to facilitate the development of tailored interventions for carers and patients to help prevent such problems, minimize the risk of patient violence and the potential negative psychological and physical sequelae for carers. Recent findings from Bowman et al. (2014) suggest that the quality of life in the siblings of early psychosis patient groups can be negatively affected by a patient's history of violence. Thus, exploring the needs of other family members who may not be in primary caregiver roles but are nevertheless affected by patient violence would seem a helpful way forward. In services amongst clinicians,

#### REFERENCES


more efforts are required to routinely and systematically record patient violence in caregiving relationships and its impact. Developing carer focused support interventions, which could be incorporated into community treatment models, are likely to benefit both carer and patient outcomes (McCann et al., 2011), and are consistent with recommended treatment approaches (NICE, 2014, 2015). Though most adults with psychotic disorders do not engage in violence, domestic violence in psychosis should be an issue of public health and concern. Focusing on building a better understanding of the patient sub groups who engage in acts of violence toward their caregivers might support the development of preventative and targeted interventions, which would have the potential to improve outcomes for all.

#### CONCLUSION

In psychosis, our findings indicate that patient violence in caregiving relationships can impact carer wellbeing and outcomes. Historically, however, carer needs and their issues have tended to be overlooked and marginalized. The current findings underscore the importance of focusing clinical and research efforts on carers and caregiving relationships affected by patient violence.

#### AUTHOR CONTRIBUTIONS

JO led on the conceptualization and design of the project. JO and ZZ led on the database searches and data synthesis. All authors contributed to preparing the manuscript.

#### FUNDING

For JO, this paper represents independent research partly funded by the National Institute for Health research (NIHR) Biomedical Research Centre at South London and the Maudsley NHS Foundation Trust and King's College London. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health.


amongst individuals with early onset psychosis. Int. J. Law Psychiatry 33, 171–176. doi: 10.1016/j.ijlp.2010.03.007


**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 © 2018 Onwumere, Zhou and Kuipers. 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.

# Psychological Impact of Stalking on Male and Female Health Care Professional Victims of Stalking and Domestic Violence

#### Daniela Acquadro Maran<sup>1</sup> \* and Antonella Varetto<sup>2</sup>

<sup>1</sup> Department of Psychology, University of Torino, Turin, Italy, <sup>2</sup> A.O.U. Città della Salute e della Scienza di Torino, Turin, Italy

The aim of this work was to investigate stalking experiences in a sample of Health Care Professionals, or HCPs, who experienced domestic violence in their previous relationships with an intimate romantic or non-romantic who had become their stalkers. A comparison between males and females was made to highlight the differences among the genders. The findings showed that, for the most part, the victims experienced stalking by a stalker that was not of the same gender. Moreover, the nature of the relationship was romantic, for the most part, for both female and male subjects, suggesting that the principal motivation of stalking is the disruption of an intimate relationship. Regarding domestic violence, females described the phenomenon from a different perspective, indicating verbal, physical, and sexual abuse, while males indicated only verbal abuse. Females tended to amplify, more than the males, depression, and state and trait anxiety. Even if all symptoms were expressed in both females and males, the males exhibited a lack of confidence in their bodies, and the emotional literacy made the expression of distress more difficult. At the same time, the expression of anxiety presented in the women permitted them to become progressively less victimized over time; depression and anxiety allow the recognition of these symptoms as signs of distress and to intervene to reduce them.

Edited by:

Shulamit Ramon, University of Hertfordshire, United Kingdom

#### Reviewed by:

Carlene M. Turner, Norfolk State University, United States Nicole Anderson, Brigham Young University, United States

\*Correspondence:

Daniela Acquadro Maran daniela.acquadro@unito.it

#### Specialty section:

This article was submitted to Gender, Sex and Sexuality Studies, a section of the journal Frontiers in Psychology

> Received: 20 December 2017 Accepted: 26 February 2018 Published: 13 March 2018

#### Citation:

Acquadro Maran D and Varetto A (2018) Psychological Impact of Stalking on Male and Female Health Care Professional Victims of Stalking and Domestic Violence. Front. Psychol. 9:321. doi: 10.3389/fpsyg.2018.00321 Keywords: anxiety, depression, distress, prevention, hospitals, gender differences, relationships

# INTRODUCTION

Stalking has been defined by Westrup (1998) as a set of repetitive behaviors directed toward a target that perceives those behaviors as unwelcome and intrusive. As a consequence, victims of stalking experience fear for their safety (Sheridan et al., 2003) and/or for closure (Dennison, 2007). Since the 1990s, national surveys have been made in several countries (see for example Tjaden and Thoennes, 2000) using different criteria both to define the phenomenon (e.g., the duration of the stalking campaign), the method used (e.g., interview by phone or online survey), and the gender of the subject involved (female and/or male). This limitation did not allow the stalking campaign to be defined in a unique way, influencing (and being influenced – De Fazio, 2009) the boundaries of the phenomenon. Despite these limitations, Spitzberg and Cupach (2003, 2007) identified the prevalence rate of victimization among females and males in the general population. While the prevalence rate in females ranged from 8 to 32%, the range was 2–13% in males. Thus, the risk of victimization among females is higher, particularly in young individuals (18–29 years) and may

extend to when the stalker tries to establish or reestablish a relationship. On average, the analysis of the literature by Spitzberg and Cupach (2007) highlighted a duration of 22 months, with a frequency of the contact that Sheridan and Davies (2001) found in their investigation to be 'every day' for more than one third of the victims.

Further research tried to understand the nature of the victim– stalker relationship and previous domestic violence (defined as "any violence between current or former partners in an intimate relationship, wherever and whenever the violence occurs," Walby and Allen, 2004, p. 4), the behavior that characterized the stalking campaign, the physical and emotional consequences, and the coping strategies adopted. Regarding this relationship, Sheridan et al. (2003) described the stalker as an ex-intimate, an acquaintance (e.g., friend, relative, colleague, or patient) or a stranger. From their work, it appears that prior domestic violence increases the chances of becoming a victim of stalking.

Moreover, as underlined by Senkans et al. (2017), if the intimate relationship was characterized by violence, this will be perpetrated also during the stalking campaign; the end of the relationship is not the end of the abusive behavior (Palarea et al., 1999; Douglas and Dutton, 2001; Roberts, 2005). The behaviors that characterize the stalking campaign have been described as unwanted communication (any contact with the stalkers, such as telephone calls, e-mails, letters or cards, text messages), unwanted approaches (the stalker followed, damaged property, visited or waited outside home and/or workplace) (McEwan et al., 2012), harassment and intimidation (threats, asking for information, spreading lies) (Spitzberg and Cupach, 2014). In their investigation, in which 107 victims of stalking were involved, Ferreira and Matos (2013) found that the majority of participants reported they had experienced partner abuse before the beginning of the stalking campaign (caused by the end of the relationship). Among these, the findings revealed that stalking victims suffer from a very diverse spectrum of behaviors, more frequently unwanted communication and contact, while harassment and intimidation are the least frequent. The stalking campaign leads to physical and emotive consequences for the victims, described by Littel (1999) as 'soul destroying.' Among the physical consequences, Spitzberg and Cupach (2007, 2014) indicated e.g., forms of addictions, appetite, or sleep disturbances, while among the emotive consequences they suggested the examples of anger, confusion, and fear. Mullen et al. (2006) highlighted the importance of investigating depressive and anxious reactions because these symptoms are linked to the experience of stalking victimization (Dressing et al., 2005), in particular in victims of ex-partners (Davis and Frieze, 2000). To confirm this, findings from the investigation by Sheridan and Lyndon (2012), which involved 1,214 self-defined stalking victims, showed that victims who had a prior relationship with their stalkers were more likely to experience a greater number of emotive, physical, and social consequences than other types of victim– stalker relationships. To cope with the phenomenon, victims use several strategies. Spitzberg and Cupach (2014) categorized these coping strategies (see also Amar and Alexy, 2010) as the following:


Cupach and Spitzberg (2004) highlighted that victims engage in multiple strategies to cope with stalking campaigns. Ferreira and Matos (2013) highlighted that the most used coping strategies by victims of stalking, previously victims of domestic violence, were to search for help from friends or family (moving outward, following the categorization of the coping strategies suggested by Spitzberg and Cupach), to negotiate with and to confront the stalker (moving toward), and to avoid the stalker (moving away).

An analysis of the literature revealed that one of the samples most investigated was HCPs (see McIvor et al., 2008). Findings from investigations involving psychiatrists, physicians, nurses, psychologists, and among others have shown that this population is at higher risk of victimization than the general population (Ashmore et al., 2006; Abrams and Robinson, 2011; Whyte et al., 2011; Paraschakis and Konstantinidou, 2012; Mastronardi et al., 2013). In their literature analysis, Spitzberg and Cupach (2003) argued that the average incidence was 13.9% for samples from ordinary populations, while the prevalence rate of victimization in HCPs ranged from 12 to 50% (Acquadro Maran et al., 2014a). In this population, the risk of victimization has been linked to both to the nature of their work and to the expectation about the engagement in the relationship (Galeazzi et al., 2005). HCPs have close contact with people suffering from physical and/or mental disease. Their professional behavior, involving being devoted to caring for the patient, could be misunderstood as a desire to begin a relationship (intimate romantic or non-romantic). The motion to change the nature of the relationship from the HCP could cause disillusioned beliefs, feelings of frustration, desires of revenge, and so on (Galeazzi et al., 2005). Thus, a stalking campaign begins with the aim of establishing a relationship. On the other hand, people (partners, friends, and relatives) have expectations about the quality of the care that HCPs provide, also outside the workplace. These expectations are linked to the attention to the relationship and, as has emerged in previous research (Acquadro Maran et al., 2017), the failure to fulfill it could trigger emotive reactions such as anger and jealousy. In this case, the stalking campaign could begin with the desire of revenge.

A gap in the literature is related to the experience of domestic violence in this population and the experience of stalking victimization. An analysis of the literature showed that HCPs are asked to prevent and/or to intervene in domestic violence (see e.g., Krug et al., 2002). The possibility that they could themselves be victims of domestic violence was not considered. However, previous investigations showed that HCPs are victims by acquaintances and ex-partners (Ashmore et al., 2006; Whyte et al., 2011).

## Current Study

The aim of this work was to investigate stalking experiences in a sample of Italian female and male HCPs who experienced domestic violence and stalking victimization. The Italian context has been characterized since 2009 by the introduction of the anti-stalking law (Penal Code, article 612 bis, 2009). This law states that: "Provided the act is not recognized as a more serious crime, it is a criminal offence, punishable with imprisonment ranging from 6 months up to 4 years, to continuously threaten or harass another person to such an extent as to cause a serious, continual state of anxiety or fear, or to instill in the victim(s) a motivated fear for his/her own safety or for the safety of relatives or other persons linked to the victim(s) by virtue of kinship or emotional relationship or to force the victim(s) to change his/her living habits.". Previous research on Italian HCP victims of stalking, showed that among 107 nurses, 28 were victims of an ex-partner, while 44 were victims of acquaintances (Acquadro Maran et al., 2014a). In an investigation among 256 HCPs, it emerged that stalkers were ex-partners of 88 victims, while for 95 they were acquaintances (Acquadro Maran et al., 2017).

The general goal of our study was to evaluate the stalking experience in a sample of HCPs that experienced domestic violence in their previous relationship with an intimate romantic or non-romantic that became their stalkers. A comparison between males and females was made to highlight the differences between the sexes. The variables investigated were those described by the literature (see above): nature of the relationship, stalking behaviors, the frequency and the duration of the stalking campaign (one item each), the consequences, and the coping strategies used. Given the lack of literature on this topic, we did not have specific hypotheses about the gender difference in the stalking experience of females and males who experienced domestic violence.

The data were gathered from a survey on stalking victimization involving more than 4000 HCP who worked in 6 Italian hospitals (public sector), and 1901 questionnaires were filled out (47.2%). The HCPs self-declared victims of stalker were 272 (14.3%). A selection was made among the cases of self-declared victimization. The criteria of inclusion were the nature of the relationship [intimate romantic (such as partner and ex-partner) and intimate non-romantic (relatives, close friends – see Spitzberg, 2002)] and the presence of a form of domestic violence (verbal abuse, physical harm, sexual abuse – Bennett Cattaneo et al., 2011). Victims of other types of stalkers (acquaintance, unknown) and those who did not suffer from domestic violence were not included.

# MATERIALS AND METHODS

# Participants

In accordance with the above-mentioned criteria, and among more than 270 cases of self-declared victimization (205 female, 67 male), 147 (54%) were selected. The participants were aged 19–60 years (mean age 36 years, SD = 11.24), and most of them 96 (65.3%) were female. Overall, most of them were nurses (59, 40.1%), psychologists, (37, 25.2%), physicians (22, 15%), health technicians (14, 9.5%), and health care operators (9, 6.1%). Six (4.1%) HCPs did not answer this question. About a quarter were single (36, 24.5%), 35 participants were engaged (23.8%), 35 were married (23.8%), 18 were cohabiting (12.2%), 18 were divorced (12.2%), and one was widowed (0.7%). The remainder of the sample (4, 2.7%) did not answer this question. The stalker was in most cases a male (101, 68.7%), and he/she was aged 17–80 (M = 35.8, SD = 12.47). In 13 (8.8%) cases, the stalker was unemployed. All victims experienced verbal abuse before the beginning of the stalking campaign, 15 (10.2%) experienced physical harms, and 4 (2.7%) experienced sexual abuse. All respondents took part on a voluntary basis.

# Materials

Participants were asked to anonymously complete several sections of a self-administered questionnaire. The first section described the purpose of the questionnaire and contained the instructions for replying, as well as the anonymity and privacy statements. The modified Italian version of the questionnaire constructed by the Network for Surviving Stalking (NSS) with Dr. Lorraine Sheridan (Forensic Psychologist, University of Leicester), a questionnaire on depression, and two scales on anxiety were used to describe the experience of victimization. The Italian version of the stalking questionnaire covered issues such as demographic details of the participants and the stalkers and the duration and frequency of stalking. These were followed by yes/no type questions about the following:


The coping strategies were subsequently categorized as suggested by Spitzberg (2002) and Spitzberg and Cupach (2007) as moving toward, moving away, moving against, moving inward, or moving outward (e.g., told friend, relative, or neighbor).

The Beck Depression Inventory (BDI, Beck et al., 1961; Italian version by Scilligo, 1988) and the State-Trait Inventory (STAI,

Spielberger, 1983; Italian version by Pedrabissi and Santinello, 1989) were used to investigate the psychological consequences of the victimization in each stalking campaign. The BDI is a 21 question survey designed to determine the presence of depression symptoms. Scoring permits the classification of depression as minimal (scores 0–13), mild (14–19), moderate (29–28), and severe depression (> 29) (in this study Cronbach's α = 0.96). The STAI consists of two forms to measure the state (Y1 form, how the victim of stalking feels 'right now,' at this moment) and trait (Y2 form, how the stalking victims feel most of the time) anxiety. Each scale includes 20 items. The total scores can range between 20 and 80, where 40 is the threshold value considered predictive of anxiety symptoms. A rating scale defines the level of severity, with 40–50 indicating mild, 50–60 indicating moderate, and >60 indicating severe anxiety. Cronbach's α was 0.87 and 0.86, respectively. All the questionnaires were selfadministered.

#### Procedure

A letter with the invitation to take part in the investigation on HCPs victims of stalking was sent out to six hospitals. In the letter, we explained the purpose of the research, the voluntary nature of participation, the anonymity and privacy statement in accordance with Italian Law and with the Declaration of Helsinki, the scales that would be used and the procedure for completing and collecting the questionnaires. Hospital administrations and local guarantee committees evaluated, endorsed, and authorized the research, allowing researchers to use the data for scientific purposes. Upon approval, Department Chiefs from each unit/service were asked for authorization to administer the questionnaire to the HCPs and to define the method of delivery of the questionnaires.

Each participant was given a printout of the questionnaire, the information letter, and the informed consent form in accordance with the Declaration of Helsinki. The first page of the questionnaire contained the aim of the research, the instructions for completing and returning the questionnaires and the contact details of the researchers (the authors of this paper) for any doubts or problems. The stalking phenomenon was described on the first page. The definition by Galeazzi and Curci (2001), similar to that set forth in article 612 bis, 2009 of the Italian Penal Code, was used: "a repetitive pattern of behavior, intrusive surveillance and control, unwanted communication or contact with a victim, which causes a state of fear and/or anxiety and/or annoyance (for the victim him- or herself and/or for his or her loved ones)" (Acquadro Maran et al., 2017, p. 2608).

All HCPs were asked to complete the first part of the questionnaire (socio-personal data). After this section, one question discriminated victims and non-victims: "referring to the previous description of the phenomenon, have you been a victim of stalking during the lifetime?." Consequently, for those subjects that self-declared as a victim of stalking, the questionnaire continued onto the next sections. For those who answered 'no,' the questionnaire ended. For all, the request was to place the questionnaire in a sealed box situated in the locker room. The scheduled date for collection was after a 3-week period (after 1 week there was a reminder placed on the sealed box).

## Data Analysis

The data were processed using SPSS version 24 to produce mainly descriptive and inferential statistics. Descriptive measures (means ± SD) were calculated for all test variables for the two groups of victims (male, female). The test of Chi-square (χ 2 ) was used to measure the differences between groups in terms of the categorial variables (sex male/female, yes/no answer to the stalking questionnaire) and the different cut-offs that indicated the level of depression and state and trait anxiety symptoms (minimal, mild, moderate, or severe). Differences were considered statistically significant if p < 0.05. Correlations were calculated to examine the relation between the number of physical and emotional symptoms reported by female and male victims of stalking and depression and anxiety symptoms, and between the number of methods of harassment and the coping strategies used by victims (female, male).

# RESULTS

#### Female HCPs Victims of Stalking

Female HCPs who self-declared as victims of domestic violence and stalking were 96 (65.3–46.8% of victims among the female HCPs victims of stalking), aged 19–60 years (M = 35, SD = 11.40). Most of them were nurses (38, 39.6%), psychologists, (27, 28.1%), physicians (15, 15.6%), health technicians (8, 8.3%), and health care operators (7, 7.3%). About one third were married (27, 28.1%), 27 participants were engaged (28.1%), 19 were single (19.8%), 12 were cohabiting (12.5%), 10 were divorced (10.4%), and one did not give an answer.

The stalker was in most cases a male (88, 91.7%). He/she was aged 17–80 years (M = 35.2, SD = 12.54). The stalker was an employee in most cases (83, 86.4%), and 25 (26%) were HCPs. The nature of the relationship with the victims was intimate romantic in 67 cases (69.8%) and intimate non-romantic in 31 (32.3%) cases. The domestic violence was described by the victims as verbal abuse (all respondents), physical harm (15, 15.6%), and sexual abuse (3.1%). The behaviors that characterized the stalking victimization are in **Table 1**. Among 'others' behaviors, one female described 'knocking at the window.' On average, females experienced five different behaviors, and in most cases, the victims (59, 61.5%) affirmed that the frequency of the behaviors was 'every day.' A total of 10 victims (10.4%) declared that they were still victims, 13 (13.5%) did not know, and the rest (73, 76%) said no. The duration of the stalking campaign was on average, more than 1 year (range 2–120 months, M = 16.57, SD = 26.11).

The stalking campaign left the victims with both physical and emotive consequences (see **Table 2**). Female victims of stalking suffered from 1 to 4 different physical consequences (M = 1.23, SD = 0.60) and from 1 to 3 different emotive consequences (M = 1.21, SD = 0.49). Regarding depression and anxiety, the results showed that, for the most part, for females the level of depression was minimal (**Table 3**), even if the females experienced symptoms more often than males. Regarding trait anxiety, the findings showed that females were more prone than males to reach the cut-off for moderate anxiety (STAI-Y2). The number of emotive consequence was significantly related to

the increase in depression symptoms (**Table 4**). To cope with the stalking campaign, the victims adopted different strategies (**Table 5**). All victims adopted at least one strategy of the moving away type. Inferential statistics showed that when the number of stalkers' behaviors increased, the use of the following coping strategies decreased: moving away, moving against, moving inward, and moving outward (**Table 6**).

#### Male HCPs Victims of Stalking

Male HCPs who self-declared as victims of stalking were (51, 34.7–76.1% of victims among the male HCPs victims of stalking), aged 21–60 years (M = 38, SD = 11.49). Most of them were nurses (21, 41.2%), psychologists (10, 19.6%), physicians (7, 13.7%), health technicians (7, 13.7%), and one (2%) health care operator. Three HCPs did not give any information about their work. More than one third were single (17, 33.3%), eight participants were engaged (15.7%), eight were married, eight were divorced, six (11.8%) were cohabiting (12.5%), and one was a widower. Three participants did not give an answer to this question.

The stalker was in most cases a female (34, 66.7%). He/she was aged 20–65 years (M = 37.07, SD = 12.38). The stalker was an employee in most of the cases (37, 73.5%), and 16 (31.3%) were HCPs. The nature of the relationship with the victims was intimate romantic in 37 (72.5%) cases, while in 14 cases (27.4%) it was intimate non-romantic. The domestic violence was described by all victims as verbal abuse. None of the male respondents indicated physical harm or sexual abuse. The stalking campaign was characterized by different behaviors (see **Table 1**). Among 'others' behaviors, the participants described the 'the stalker threatened self-injury.' On average, males experienced three different behaviors. Half of the victims (26, 51%) affirmed that the frequency of the behaviors was 'every day.' A total of 6 victims (11.8%) declared that they were still victims, 10 (19.6%) did not know, and the rest (35, 68.6%) answered no. The durations of the

TABLE 2 | Physical and emotional symptoms characterizing the male and female HCPs experience of stalking and domestic violence victimization (N = 147).


stalking campaigns were similar to those of the females, but it was a little longer, on average being more than 1 year, with a range of 2–132 months (M = 16.65, SD = 22.69). Male victims suffered, more than females, from the unwanted written communication

Percentage values are given in parentheses. χ <sup>2</sup> = chi-square; p = p-values; n.s. = not statistically significant.

TABLE 1 | Behaviors characterizing the male and female HCPs experience of stalking victimization (N = 147).


Percentage values are given in parentheses. χ <sup>2</sup> = chi-square; p = p-values; n.s. = not statistically significant.

TABLE 3 | Level of depressive and anxiety symptoms indicated by the male and female HCPs experiencing stalking and domestic violence victimization (N = 147).


Percentage values are given in parentheses. χ <sup>2</sup> = chi-square; p = p-values; n.s. = not statistically significant.

method of harassment ("sending e-mails, letters, or cards"). The physical and emotive consequences that characterized the stalking campaigns are in **Table 2**. Male victims suffered from 1 to 3 different physical consequences (M = 1.07, SD = 0.34) and at least one emotive consequence (M = 1.05, SD = 0.35). The males suffered, more than the females, from 'anger.' The results showed that for the most part, for the males, the levels of depression, state trait, and anxiety were minimal (**Table 3**) and that they were less prone than females to experience these symptoms. The physical consequence was significantly related to both state and trait anxiety symptoms (**Table 4**). To cope with the stalking campaign, victims adopted different strategies (**Table 5**). Similar to the sample of females, all the males adopted at least one strategy of the moving against type. The stalkers' use of different behaviors was significantly related with the decrease in the use of the moving away and moving against coping strategies (**Table 6**).

TABLE 4 | Correlation between the number of physical and emotional symptoms reported by male and female HCP victims of stalking and domestic violence and depressive and anxiety symptoms (N = 147).


<sup>∗</sup> = p < 0.005; ∗∗ = p < 0.010.

TABLE 5 | Typology of coping strategies used by male and female HCP victims of stalking and domestic violence (N = 147).


Percentage values are given in parentheses. χ <sup>2</sup> = chi-square; p = p-values; n.s. = not statistically significant.



<sup>∗</sup> = p < 0.005; ∗∗ = p < 0.010.

#### DISCUSSION

The aim of this work was to compare female and male HCP victims of domestic violence and stalking. The findings showed that, for the most part, the victims experienced stalking by a stalker who was not of the same gender, confirming that the phenomenon is most frequently inter-gender, particularly when the victim was a female (Sheridan et al., 2016). Moreover, the nature of the relationship was romantic for the most part for victims, both female and male, suggesting that the principal motivation of stalking was the disruption of an intimate relationship (Tassy and Winstead, 2014). Regarding the domestic violence, females described the phenomenon from a different perspective, indicating the verbal, physical, and sexual abuse, while males indicated only the verbal abuse. These findings did not support those from investigations of male victims of domestic violence by Drijber et al. (2013); the men were physically as well as the psychologically abused females, and were often an (ex)-partners. Interestingly, in our sample, male victims of stalking were more prone to experience unwanted written communication than females. This confirmed that female stalkers tend to adopt more behaviors that permit them to be connected with their victims (Purcell et al., 2001). The duration of the stalking campaigns was similar in both females and males, with a little longer duration in males; as suggested by Meloy and Boyd (2003), female stalkers are more patient and tough. Male victims are also more prone than females to express anger with their stalkers, though they did not reach threshold values from the psychopathological point of view. This finding confirmed that the expression of this feeling (especially behaviorally) is culturally associated with men (Simon and Nath, 2004).

From the screening for depression, BDI emerged that the discouragement did not prevail for the most part in either male or female victims. However, females tend to amplify, more than males, depression, state, and trait anxiety (in particular the moderate level of trait anxiety). The expression of anxiety symptoms was also seen through the body; indeed, females experienced somatic and cognitive (such as confusion) symptoms. Moreover, in females, there was a higher influence of the victimization in some cognitive aspects that could have had an impact on work efficiency, on the ability to apply social and organizational rules (Acquadro Maran et al., 2014b), medical procedures, and to care the patients (Collins and Long, 2003). The anxiety and the somatization were evident, for example, in the higher percentage of sleep disturbances in female. Even if all symptoms were expressed in both females and males, in males a lack of confidence in their body and of their emotional literacy makes the expression of distress (in each channel, such as emotive and cognitive) more difficult. At the same time, the expression of anxiety presented in women is permitted to become progressively less victimized over time; depression and anxiety permitted the recognition of these symptoms as signs of distress and to intervene to reduce them (Spence-Diehl, 2004).

An interesting finding was in regard to the coping strategies. Victims, both female and male, involved in this investigation confirmed that the coping strategy of moving away was the most used in this population (Acquadro Maran et al., 2017), alone

or in association with another. However, our findings suggested that when harassment behavior increased, the number of coping strategies adopted by the victims decreased. An explanation could be in the fatigue resulting from coping with repetitive and intrusive behavior that distress the victim (Davis et al., 2002), leading to exultation. This result was not in accordance with Spitzberg et al. (1998, p. 43) who argued that "the more a person is obsessively pursued, the more this person attempts to cope, and the increased coping is merely a barometer of the stalking and its disruptiveness, rather than a method of effectively diminishing the negative effects of the stalking" (p. 43). According to Davis and Frieze (2000), the link between coping strategies and the stalking campaign needs attention from scholars; the adoption of an appropriate coping strategy (e.g., sought help from colleagues) could determine the stop of the stalking campaign (Kaplan, 2006). In particular, in HCP victims of stalking, the urgency to intervene is linked to the need to limit the consequence of the stalking campaign, in order to be efficient and effective at work.

There were, of course, limitations to this study. First, since the sample was non-randomly selected, the results should be taken with caution and should not be generalized. Moreover, the sexual orientation of the stalker and victim was not investigated; thus, comparisons between heterosexual and non-heterosexual individuals were not made. Studies that had directly assessed the stalking campaign based on sexual orientation found that men were more likely to engage in a stalking campaign at the end of a relationship than women were (Derlega et al., 2011). Furthermore, in this current study, the data on the contradiction between being a HCP victim of stalking/domestic violence and caring for victims of stalking/domestic violence were not collected. Future research should investigate the psychological impact in HCPs who are victims of stalking and domestic violence and caring for victims of stalking and domestic violence. At least one other limitation was in reference to the domestic violence experience. Our work was based on the more wellknown categorization of domestic violence [physical, sexual, emotive – see World Health Organization [WHO], 2001 and Garcia-Moreno et al., 2006], but the questionnaire was not tailored to provide detailed information about the experience (for example the economic violence or violence associated with ethnic/religious motives were not investigated). We suggest that future studies examine the experience of domestic violence and its link with sexual orientation and stalking victimization in a more comprehensive way. As argued by Sheridan et al. (2014), the study of those variables could allow a better understanding of the dynamics of the stalking phenomenon, its consequences and for the exploration of the efficacy of coping strategies adopted by victims and their social context.

Despite these limitations, we hope this study offers interesting insights and suggests implications for HCPs and the organization in which they are working. First, attention is generally focused on female victims of domestic violence and stalking. This study highlights, one more time, the importance of considering men as potential victims of domestic violence and stalking (Tarzia et al., 2017). The indication is that HCPs, and the entire health care system, need to improve their ability to recognize the signs of victimization in men, to provide more suitable intervention for individuals (e.g., counseling, Zaccagnino et al., 2017) and the social context (e.g., to protect them and their families). Moreover, the auspice is to consider HCPs not only as providers of care in victims of domestic violence and stalking but also as potential victims themselves. For HCP victims of domestic violence and stalking, due to the nature of their work, it could be more difficult to admit the victimization, particularly when the nature of experiencing violence is intimate. At the same time, the perceived contradiction of being victims and providers of care in victimization cases (Guldimann et al., 2015) could result in a minimization or a denial of the problem (Acquadro Maran et al., 2018). Clearly, such an attitude can be harmful both for the patient/victims and for the HCP victims. In HCP victims, the experience could result in a reluctance to seek support, with a consequently prolonged exposure to the stalking campaign and its effect on well-being. Finally, health care organizations (e.g., hospitals) should contribute to prevent of the phenomenon and should intervene in domestic violence and stalking phenomena. Prevention programs include, for example, information courses on the phenomena (e.g., underlying the prevalence of victimization among HCPs), the risk of victimization (in the general population and in HCP population), and defense strategies (also those offered by the Italian anti-stalking law). Health care organizations should also offer individual measures, such as intervention programs, counseling, and psychological help, to reflect on victimization experiences. Future research should look to replicate – with a larger sample – the current analyses to test the psychological impact of the different forms of domestic violence in HCPs victims of stalking.

# ETHICS STATEMENT

This study was carried out in accordance with the recommendations of Italian Law statement about privacy, the Code of Italian Psychologist, and the Ethical Committee of the Univesrità degli Studi di Torino with written informed consent from all subjects. All subjects gave written informed consent in accordance with the Declaration of Helsinki. The protocol was approved by the Hospital administrations and local guarantee committees evaluated, endorsed, and authorized the research, allowing researchers to use the data for scientific purposes.

# AUTHOR CONTRIBUTIONS

DAM and AV substantially contributed to the conception or design of the work or to the acquisition, analysis, or interpretation of the data for the work. DAM and AV drafted the work or revised it critically for important intellectual content. DAM and AV made the final approval of the version to be published. DAM and AV prepared the agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

# REFERENCES



**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 © 2018 Acquadro Maran and Varetto. 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 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.

# Domestic Violence and Education: Examining the Impact of Domestic Violence on Young Children, Children, and Young People and the Potential Role of Schools

#### Michele Lloyd\*

The School of Education, University of Hertfordshire, Hatfield, United Kingdom

This article examines how domestic violence impacts the lives and education of young children, children, and young people and how they can be supported within the education system. Schools are often the service in closest and longest contact with a child living with domestic violence; teachers can play a vital role in helping families access welfare services. In the wake of high profile cases of child abuse and neglect, concerns have been raised about the effectiveness of multi-agency responses to children living with abuse. In the United Kingdom, the case of 4-year-old Daniel Pelka who died in 2012 following abuse and starvation by his mother, who experienced domestic violence, and her partner, led to a serious case review. It found recording systems in Daniel's school were not used consistently, and details held by different agencies were not collated to enable the formation of a coherent assessment. The lack of integrated working cited in the report echoes findings from previous serious case reviews. A strong correlation exists between domestic abuse and child abuse, with approximately half of all domestic violence situations involving direct child abuse. Children can also be affected indirectly by violence occurring in their home by seeing or hearing it taking place. This article examines the impact of domestic violence on the mental health of children, and the impact on their education. Violence in children's lives often causes disruption to their schooling and harms the quality of their educational experiences and outcomes. The abuse children experience can result in emotional trauma, physical and psychological barriers to learning, and disruptive behavior in school, while the underlying causes of these problems remain hidden. Knowing when and how to seek advice from multi-agency professionals is an essential part of effective practice among school staff. Despite their vital role in identifying signs of abuse and signposting referral pathways, research indicates teachers often lack confidence and knowledge for such work. The article examines how the professional learning and professional confidence of teachers can be developed, and how recent policy and practice developments in the United Kingdom have the potential to influence work in this area.

Keywords: domestic violence, education, early childhood, children, young people, schools, teachers, multiagency working

#### Edited by:

Piera Brustia, Università degli Studi di Torino, Italy

#### Reviewed by:

Eunice Lumsden, University of Northampton, United Kingdom Connie J. Beck, The University of Arizona, United States

> \*Correspondence: Michele Lloyd m.lloyd1@herts.ac.uk

#### Specialty section:

This article was submitted to Gender, Sex and Sexuality Studies, a section of the journal Frontiers in Psychology

> Received: 13 May 2018 Accepted: 11 October 2018 Published: 13 November 2018

#### Citation:

Lloyd M (2018) Domestic Violence and Education: Examining the Impact of Domestic Violence on Young Children, Children, and Young People and the Potential Role of Schools. Front. Psychol. 9:2094. doi: 10.3389/fpsyg.2018.02094

# INTRODUCTION

fpsyg-09-02094 November 10, 2018 Time: 13:42 # 2

Every school is likely to have children affected by domestic violence. The aim of this article is to examine how domestic violence impacts the lives of young children, children, and young people, and the potential role that schools can play in helping to address their needs. Wellbeing and healthy relationships are the foundations of learning. The immediate and long-term costs of domestic violence can thus be high, affecting children's education as well as having long-term developmental consequences.

Many types of abuse occur within the domestic sphere. In the United Kingdom the government definition of domestic violence and abuse is: 'any incident or pattern of incidents of controlling, coercive, threatening behavior, violence or abuse between those aged 16 or over who are, or have been, intimate partners or family members regardless of gender or sexuality. The abuse can encompass, but is not limited to: psychological, physical, sexual, financial, emotional' (Home Office, 2018). Although this definition applies to those aged 16 or over, children also experience the harmful effects of domestic violence, as will be examined further below. There is growing awareness of the emotional harm of domestic violence, evidenced in the United Kingdom in the offense of controlling or coercive behavior in intimate or familial relationships which has a maximum custodial sentence of 5 years, a fine or both (Home Office, 2015). As well as being affected by physical abuse, children can be affected by nonphysical domestic abuse based on coercive control, such as isolation, continual monitoring, financial abuse, and verbal and psychological abuse (Katz, 2016). Domestic violence is part of the landscape of child protection. Government documentation on child protection titled 'Working Together to Safeguard Children' (HM Government, 2013) details safeguarding responsibilities of professionals and organizations, and promotes a child-centered approach based on the needs and views of children (Holt, 2014).

While the term domestic violence is used in this article, a number of terms are present in literature such as intimate partner violence and inter-parental violence, and sometimes terms are used interchangeably. Since there is a governmental definition of domestic violence, and also for consistency, I will adopt the term domestic violence predominantly here. The terms intimate partner violence and inter-parental violence will be used when referring to studies that specifically adopt these terms.

Domestic violence and child protection is a complex, multifaceted area. It is common for domestic violence and, as specified in this case, intimate partner violence (IPV) to co-occur with other problems: 'children's experiences of and responses to IPV exposure cannot be viewed in isolation from other adversities and inequalities' (Etherington and Baker, 2018, p. 70). The co-occurrence of stressful problems in early life is often referred to as adverse childhood experiences (ACEs). ACEs is a construct emerging from a long line of studies into traumatic events occurring in childhood such as domestic violence, sexual, physical and emotional abuse, household dysfunction, and neglect (Felitti et al., 1998; Dube et al., 2001). Research studies find that having these ACEs has long-lasting effects into adulthood. ACEs can be a source of long-term psychological distress as well as having longitudinal effects on physical health, substance misuse, interpersonal violence and self-harm (Hughes et al., 2017). The 'toxic trio' of domestic violence, substance misuse, and parental mental health problems can render children at risk of harm and complex trauma. Poverty all too frequently intersects with ACEs. Although poverty is viewed as a social marker regarding the distribution of domestic violence risk, the association is not causal (Ray, 2011). Domestic violence cuts across all socio-economic groups and all backgrounds. Victims of all backgrounds, predominantly women, face common difficulties when leaving an abusive partner. Research demonstrates that it is at the point of leaving, or after she has left, that a woman is in most danger (Calder and Regan, 2008). It is not uncommon for victims of domestic violence to remain living with perpetrators, even risking their own safety, rather than risking themselves, and their children, becoming homeless.

Teachers are well placed to play a pivotal role in identifying and responding to domestic violence since they have contact with children more than any other service. As emphasized by Sterne and Poole (2010, p. 17), 'although staff in schools may not be able to stop the violence at home, they are in a position to make a considerable difference to children's lives.' Statistics from the Department for Education [DFE] (2017a) show that of the 646,120 children referred to children's social care in England 2016–2017, the highest number of referrals, 27.5%, came from the police. The second highest percentage of referrals came from schools at 17.7% followed by health services at 14.4%. School referrals combined with education services referrals of 2.6% means that education accounted for 20.3% of referrals overall. Once referred and assessed, statistics show the percentage of children in need according to identified factors. In 2016–2017 (Department for Education [DFE], 2017a) the most common factor was domestic violence which applied to 49.9% of children in need; this incorporated violence directed at children or adults in the household. The second most common factor was mental health at 39.7% which likewise encompassed mental health of the child or adults in the household. The prevalence of domestic violence is not only high among children in need but also among the wider population with as many as one in six young people in the United Kingdom reporting experiencing it during their childhood (Radford et al., 2011).

Exposure to domestic violence generates a multitude of responses and needs and it is important for children and young people not to be regarded as a homogeneous group or lacking the capacity for posttraumatic growth and recovery: 'it is wrong to stereotype all children as inevitably and permanently damaged by living with domestic violence' (Mullender et al., 2002, p. 121). Although some children experiencing domestic violence will exhibit difficulties in their schoolwork, the education of others will not be adversely affected: 'some children living with domestic abuse achieve highly in school; throwing themselves into school life and work can provide an escape' (Sterne and Poole, 2010, p. 23). Similarly, while some students affected by domestic violence will experience

educational settings as a source of continuity and security, others will experience them as challenging. It is therefore essential to take into account the range of responses to domestic violence among children. This article firstly looks at how domestic violence is conceptualized with regard to children, and how it affects them across different ages. This is followed by an examination of multi-agency working between schools and other organizations. The importance of recognizing individual and family contexts is considered before analyzing prevention education in schools. Finally, recent developments in policy and practice in the United Kingdom are examined in terms of both the challenges and opportunities they pose.

# CONCEPTUALIZING DOMESTIC VIOLENCE IN RELATION TO CHILDREN

Domestic violence is manifested in various ways and has been conceptualized by some as taking direct and indirect forms. Indirect abuse can result from inter-parental violence where children are not the subject of direct abuse. However, children witnessing inter-parental violence, and hearing it without necessarily seeing it, can still feel its effects: 'While often characterized as witnesses to inter-parental violence, which implies a passive role, children actively interpret, attempt to predict and assess their roles in causing the violence' (Baker and Cunningham, 2009, p. 199, emphasis in original). Indeed, the terms direct and indirect abuse have been interpreted as potentially misleading and perhaps simplistic. Callaghan et al. (2018, p. 1566) argue it is too restrictive to view domestic violence as abuse between partners in an intimate dyad whereby children are perceived as 'affected by' the abuse: 'Far from passive witnesses, they are not "exposed" to violence and abuse; rather they live with it and experience it directly, just as adults do.' Regarding children as 'affected by' domestic violence diminishes its impact on them. Instead, Callaghan et al. (2018) call for children to be recognized as direct victims of violence and abuse which in turn could improve professional responses to their needs.

# IMPACT OF DOMESTIC VIOLENCE ON YOUNG CHILDREN, CHILDREN, AND YOUNG PEOPLE

Domestic violence occurs at all ages. Sterne and Poole (2010) point out that the duration of children's encounters with domestic abuse has a greater bearing on their stress levels than the severity of the abuse. Harm caused by domestic violence can be physical, emotional, behavioral, cognitive, and social, and effects are usually overlapping and interrelated. Although harm can be present across all age phases, I will differentiate by three age groups, namely young children aged 1–4, children aged 5–10, and young people aged 11–16 since challenges and issues arising from domestic violence are different across these ages. It should be noted, however, these age groups are approximate and children's experiences and responses will be influenced by individual needs and context.

# IMPACT ON YOUNG CHILDREN

The effects of domestic violence can be felt in early childhood. Research shows that psychosocial development is more problematical among toddlers exposed to IPV who additionally experience physical abuse (Harper et al., 2018). In some cases domestic violence during early childhood leads to emotional problems. Among pre-school children it can cause separation anxiety from the non-abusing parent, commonly their mother. Pre-school children's restricted ability for coping due to their young age means that behavioral and psychological disengagement is one way they react to inter-parental violence (Baker and Cunningham, 2009). Pre-schoolers sensitized to the noise of family violence may cope by tuning out noise, consequently posing difficulties for those wishing to interact with them in the school setting. According to Baker and Cunningham (2009), pre-school children will react to interparental conflict in a variety of ways including becoming withdrawn, anxious, engaging in repetitive play, regressive behavior, having inhibited independence, sleep problems, tantrums or impaired understanding. The signs and symptoms of domestic violence and inter-parental violence are not always easily detectable. Moreover, it is difficult for staff in pre-school to know whether children's conduct is associated with experience of domestic violence or regular behavior expected of this age group. If staff suspect abuse, and/or notice changes in pre-school children, background checks into the home environment will help inform their professional judgment. Staff can check if the child has a previous history of abuse and if a parent has a history of violence including toward adults or animals since they are likely to be violent toward children as well (Beckett, 2007). It is important for pre-school staff to exchange information with other healthcare professionals such as health visitors who work with children from birth to five. Guidelines in the United Kingdom recommend health visitors undertake routine screening for domestic violence and share information with pre-schools and schools as appropriate. The quality of the parent–child relationship also needs to be considered by pre-school staff, for example is the child reluctant to go home or fearful in the presence of a parent. Early years teachers and support staff can develop strategies for supporting pre-school children displaying symptoms through giving positive feedback, focusing on desirable rather than undesirable behavior, validating the child's feelings, and preparing for transitions during the day (Baker and Cunningham, 2009).

#### IMPACT ON CHILDREN

Separation anxiety due to domestic violence is not limited to pre-schoolers and young school-aged children experiencing such

anxiety could be clingy, and feign illness or be disruptive at school in the hope of being sent home. In relation to the physical impact of domestic violence Calder and Regan (2008) state effects include, but are not limited to, injury, eating problems, and stress-related conditions such as asthma and bronchitis. Emotional effects, they note, are manifested in disruption to schooling including non-attendance, attention and concentration difficulties, sleep disturbance, withdrawal, insecurity, guilt, depression and low self-esteem. Behaviorally, the impact might be changes in conduct, unpredictable behavior, aggression, anger, and hyperactivity. Being the perpetrator or victim of bullying can also ensue (Children's Commissioner, 2018). Some children facing trauma at home display hypervigilance and hyperarousal at school, constantly watchful and fearful of danger (Sterne and Poole, 2010). Domestic violence can negatively affect cognitive skills, language development and educational attainment.

#### IMPACT ON YOUNG PEOPLE

In older children potential indicators of domestic violence include self-blame, depression, self-harm, suicidal ideation, substance abuse, risk-taking behavior, criminal behavior, poor social networks, disaffection with education, and eating disorders (Children's Commissioner, 2018). Research indicates that experiencing domestic violence has a differential impact along gender lines. Girls are more likely to internalize symptoms in the form of withdrawal, anxiety and depression, whereas boys, though still susceptible to anxiety and depression, are more prone to externalizing symptoms through violence against peers or antisocial behavior (Baldry, 2007). Research with young people found that being listened to, taken seriously, and jointly involved in finding solutions were key means of helping them cope; in cases where no one listened, young people felt 'doubly disadvantaged' (Mullender et al., 2002, p. 121). The effects of domestic violence clearly have implications for student wellbeing and learning examined in more detail in the following section.

# RESEARCH WITH SCHOOL TEACHERS

Research has demonstrated how domestic violence impacts on students' engagement in learning when living within, as well as leaving, abusive homes. Those leaving domestically violent homes face the additional threat of temporary homelessness or overcrowded accommodation. Research with school teachers in England has shown that the sequential issues of domestic violence and homelessness can lead to unstable accommodation, with children being re-housed frequently, obliged to live with relatives or friends, or living long distances from school due to lack of local housing (Digby and Fu, 2017). Non-permanent accommodation has an impact in the classroom through children's lack of ability to participate socially and academically. Primary and secondary school teachers in Digby and Fu's (2017) sample spoke of the effects of homelessness on children they worked with such as lack of space at home to study, limited access to a computer for homework, increased anxiety and stress, and living in noisy, overcrowded accommodation which affected their sleep. The teacher participants also noted that while children in younger age groups became withdrawn, the tendency was for older pupils to exhibit anger and aggression. The study revealed the adverse effects on teachers themselves who described feeling emotionally exhausted as well as frustrated at not always being able to help their students. Children living in a refuge are additionally vulnerable to being teased and bullied at school due to the stigma associated with refuge accommodation (Sterne and Poole, 2010). Given the multiple effects of domestic violence, teachers and support staff in schools need to be equipped with knowledge, understanding and skills to identify and respond to internalized and externalized symptoms discussed next.

## DOMESTIC VIOLENCE, SCHOOLS AND MULTI-AGENCY WORKING

United Kingdom government guidelines underline the importance of multi-agency working in child protection (HM Government, 2015). In order to strengthen education as part of multi-professional team working the government recently made a commitment to giving schools a greater role in forthcoming statutory guidance for safeguarding children (HM Government, 2018). Despite this emphasis 'Surprisingly little attention has been paid to the inter-organizational information exchange in the educational context' (Baginsky et al., 2015, p. 355) which this article seeks to examine. The United Kingdom government's statutory guidance for schools and colleges titled 'Keeping children safe in education' (Department for Education [DFE], 2016) emphasizes that safeguarding children is everyone's responsibility. Rather than being the exclusive concern of the Designated Safeguarding Lead in school 'any staff member can make a referral to children's social care' (Department for Education [DFE], 2016, p. 7). However, it is evident from research and Serious Case Reviews (SCRs) into child abuse and child deaths in the United Kingdom that school staff are sometimes unclear about their role in the child protection process, and that effective training is needed to enable school staff to better support children and their parents.

Serious case reviews have repeatedly cited failure to respond to early signs of abuse, poor record keeping, and sharing information too slowly as contributing to ineffective practice (Department for Education [DFE], 2016). The SCR into the death of 4-year-old Daniel Pelka in 2012 found that recording systems in his school were not used consistently, different social work and health organizations held partial information which was not collated to enable the formation of a coherent assessment, and insufficient training of school staff resulted in their not being clear of their role in child protection, nor whom to contact with concerns (Wonnacott and Watts, 2014). Daniel was frequently hungry when he went to school where he searched for food, including in bins (Lock, 2013). Although his mother said he had health problems, he had further unexplained injuries which did not prompt a referral. His mother's experiences of the 'toxic trio'

of domestic violence, substance misuse, and mental ill-health complicated matters further still. Lack of professional confidence among child protection workers can be a barrier to multi-agency working as in the case of Daniel Pelka 'where uncertainty and apprehension lead to inaction' (Baginsky et al., 2015, p. 355). Effective child protection requires understanding of collaborative roles: 'children are best protected when professionals are clear about what is required of them individually and how they need to work together' (Holt, 2014, p. 56).

Collaboration across agencies is similarly examined in a report entitled 'The multi-agency response to children living with domestic abuse' (Office for Standards in Education, Children's Services and Skills (Ofsted et al., 2017) which calls for health practitioners, social workers and the police to share child protection information more readily with schools. Evidence from inspections in six local authorities in England demonstrated aspects of good practice within schools for addressing domestic abuse including: schools having awareness-raising assemblies; disseminating posters and information booklets; hosting visits from charities and the police; counselors, play therapists and learning mentors working with child victims; and providing parents with support service information. The latter took the form in one school of giving out pens with a telephone number disguised as a bar code. Having support resources available in school is an important way of informing young people's friends of how to respond to disclosure since young people experiencing violence sometimes confide in their friends (Refuge, 2008). Impediments identified by teachers in the inspection report by Ofsted et al. (2017) included limited resources for working with children affected by domestic abuse; and psychological harm being taken less seriously than physical harm. The report calls on schools to prioritize education about healthy relationships which was not always in evidence from the inspections. Schools responding to domestic violence also entails working with parents, especially mothers who tend to be the non-offending parent. Working with parents requires a contextsensitive approach which forms the focus of the following section.

#### RECOGNIZING THE CHILD AND FAMILY CONTEXT

Parental non-disclosure of domestic violence coupled with wariness toward social services have a deep-rooted history, due in part to feelings of guilt, shame and fear of children being taken into care. Young people themselves have expressed fears of being removed from home (Ellis et al., 2015). Research shows that making a disclosure to professionals or other adults can be traumatic for children, with instances of family members becoming angry and upset and holding the child responsible for consequences (Children's Commissioner, 2018). Cultural taboos can render disclosure of domestic violence, including 'honorbased' violence, even more difficult for members of certain communities. Interventions by social workers are sometimes perceived, if not directly experienced, as punitive rather than supportive. In order to facilitate identification and disclosure of domestic violence victims must be treated in a non-judgmental way and their complex needs recognized. Professionals, including teachers who are the Designated Safeguarding Lead in their school, require knowledge, training and strategies for inquiring about abuse, and how to manage both disclosures and nondisclosures.

The range of needs among those living with domestic violence requires a professional response informed by victim context. Welfare services need to adopt an intersectional approach to domestic violence and its attendant issues (Ramon, 2015) whereby disability, race and ethnicity, gender, age, socioeconomic status, immigration status, and sexual orientation of children and parents alike are taken into account. Immigration status, for instance, can be a factor in non-disclosure. In recognition of the importance of intersectionality Etherington and Baker (2018) advocate service providers engage in reflexivity by examining whether their provision ignores or attends to children's multiple social locations. The needs and access to resources of a middle-class child, for example, will differ to those of a child living in persistent or recurring poverty. An intersectional, child-centered approach is promoted by Etherington and Baker (2018), one which takes into account the specificity of children's individual experiences, and is sensitive to the characteristics shaping their experiences. Where interconnected factors such as domestic violence and mental health problems affect a family's context, they need to be understood and documented in conjunction with each other rather than in isolation (Lloyd et al., 2017).

#### SCHOOL ENGAGEMENT WITH DOMESTIC VIOLENCE PREVENTION AND EDUCATION WORK

In addition to making referrals to social care 'Schools also have an essential role in educating children about domestic abuse' (Ofsted et al., 2017, p. 28). Yet research has revealed a lack of work in school on domestic violence. A survey commissioned by the domestic violence charity Refuge (2008) involving 513 young women aged 18–21 revealed that just 13% had learned about domestic violence while at school and nearly 70% responded they would have welcomed such lessons.

Engaging in prevention education and awareness raising in school can increase domestic violence disclosure from young people, though research shows mixed outcomes, with increased disclosure following some educational programs but not others (Ellis et al., 2015). Moreover, participating in a school-based program results in some young people more likely to disclose to a family member than to professionals (Ellis et al., 2015). Trust in professionals plays a key role in domestic violence disclosure. Experiences of abuse can lead to young people having diminished trust in adults and in their ability to support and protect them, sometimes a consequence of teachers in school not acting upon student disclosure of abuse in the home (Swanston et al., 2014). Teachers building trust with young people is therefore of vital importance.

Prevention programs in school are more effective when promoted through whole-school policies and practices than through single-component programs or individual teachers (Harne and Radford, 2008). Program evaluations also show that while one-off education initiatives have some value in raising awareness of domestic violence, attitudinal change is better sustained when learning is revisited and reinforced in subsequent years (Harne and Radford, 2008). Adopting a gendered approach is another preferred format for changing attitudes as it underlines that domestic violence is rooted in unequal power relations between men and women; although men can be victims too, the majority are women and they are subject to domestic violence in more severe and repeated forms (Women's Aid, 2009). Furthermore, where prevention programs in schools include a male facilitator, there is a higher likelihood of boys changing their attitudes (Ellis et al., 2006).

A more recent evaluation of a United Kingdom schoolbased domestic violence prevention program was undertaken by Fox et al. (2016). They evaluated a 6-week education program (1 h each week) delivered by domestic abuse practitioners in seven secondary (high) schools and compared participant questionnaire responses with those of participants in six schools not receiving the intervention program. The study had a total of 1,203 participants. When pre- and post-test responses were statistically analyzed, findings showed that boys and girls alike who had participated in the intervention program became less accepting of domestic violence and were more likely to seek help for abuse in comparison to those in the control group. Comparable degrees of attitude change occurred across those who had experienced abuse and those who had not experienced it. Although those in the intervention group indicated a higher likelihood of engaging in help-seeking behavior from pre- to post-test, this trend was not maintained at the 3-month follow-up data collection stage leading Fox et al. (2016) to argue that young people require more than a one-off program to persuade them of the benefits of seeking help for abuse. Congruent with previous evaluation research, Fox et al. (2016) emphasize that in order to help ensure the sustainability and effectiveness of prevention education teachers need to be trained and supported to integrate such education into the school curriculum.

For prevention and support work in school to be effective, teachers themselves evidently need to feel supported by school processes and management (Sterne and Poole, 2010). When addressing the needs of children living with domestic violence school staff should be prepared with information about services, signposting to external agencies, ensuring student safety, and knowing what to do next following disclosure. Without this information, students could be put in a worse situation than before (Howarth et al., 2016). Just as teachers need to have a clear understanding of their role in safeguarding children, so too they need to know the boundaries of their role. Research warns of the dangers of teachers acting beyond their professional scope such as asking a child to talk about their experiences without being suitably qualified which can have a traumatic effect on the child (Swanston et al., 2014). A sensitive approach is needed to help both students and their parents already living with domestic violence. Yet teachers' responses to research reveal they often lack the professional confidence and expertise to provide domestic violence prevention education and intervention support, highlighting the need for effective staff training at both initial and continuing professional teacher education, and to include school nurses (Refuge, 2008).

The content, manner and personnel delivering domestic violence education in schools clearly require careful consideration to enhance student engagement and handle student vulnerability (Fox et al., 2014). Different models of educational program delivery have been employed in school. Some entail teachers delivering school-based initiatives themselves, others favor delivery from external specialists, while some opt for collaborative implementation. Although external facilitators have specialist knowledge, expertise and experience of discussing sensitive topics with young people, teachers have more in-depth knowledge of students and their individual circumstances (Fox et al., 2014). Working in partnership with external facilitators provides a way for teachers to develop their professional learning and confidence. The respective strengths of external specialists and teachers can be complemented through collaboration:

. . .preventive interventions when co-delivered with specialist organizations might offer the possibility for school staff to increase their skills in dealing with disclosures and subsequently help improve the health and well-being of children and young people. (Ellis et al., 2015, p. 60)

The need for effective professional learning and training of school staff applies to the issue of interpersonal violence too. Cross-national European research in secondary (high) schools found that teachers frequently had limited confidence and knowledge to address the problem of interpersonal violence and abuse (Barter et al., 2015). Findings from the study echoed domestic violence program evaluations in that rather than interpersonal violence and abuse being left to the efforts of an individual teacher championing the cause, the issue should be addressed at institutional level as a whole-school concern. Schools have an essential role to play, then, in tackling domestic violence and the following section examines how recent policy and practice have the potential to influence work in this area.

# DEVELOPMENTS IN POLICY AND PRACTICE

Following long-running campaign calls for the introduction of mandatory relationships and sex education (RSE) in schools, the United Kingdom government announced in March 2017 it will introduce compulsory lessons in all schools in England (in Wales, Scotland and Northern Ireland RSE is expected but not compulsory). Current guidance on sex education in England was introduced in 2000 but content has not kept pace with social change, especially in respect to social media, online pornography and 'sexting.' While parents will still have the right to withdraw their child from sex education, draft government proposals have suggested parental withdrawal should only be possible until three terms before the child is aged 16, after which the child should be

fpsyg-09-02094 November 10, 2018 Time: 13:42 # 6

able to decide to attend (Department for Education [DFE], 2018). Subsequent to a consultation period schools will be required to teach the new RSE content from September 2020 (Department of Health, and Social Care and Department for Education, 2018). Statutory curriculum content in schools promoting healthy relationships, and raising awareness of unhealthy relationships and the unacceptability of violence in relationships, is a positive step toward equipping young people for modern-day life.

Another move in the right direction is the United Kingdom government's consultation document which proposes the implementation of a designated senior lead for mental health in every school and college (Department of Health, and Department for Education [DOH and DFE], 2017). Educational settings present a valuable opportunity to promote mental wellbeing and prevent mental ill-health since half of all mental health conditions start by the age of 14 (World Health Organization, 2013). Although 61% of schools currently offer counseling (Department of Health, and Department for Education [DOH and DFE], 2017), concerns have been raised about schools' ability to resource such support coupled with the length of time students sometimes need to wait. There have also been long-standing concerns around high referral thresholds for external support services such as Child and Adolescent Mental Health Services (CAMHS) which some children affected by domestic violence have not been accepted for due to referral criteria (Swanston et al., 2014). Long waiting times to access such services have been a further source of distress; the average waiting time is 12 weeks but the longest is up to 100 weeks (Department of Health, and Department for Education [DOH and DFE], 2017) during which time problems can escalate requiring more intensive and more costly support. The government's consultation document proposes a 4-week waiting time for National Health Service mental health services for children and young people, and recommends content on mental wellbeing be part of the Personal, Social, Health and Economic education syllabus in schools. The government's aim to implement training for the designated senior lead for mental health to all areas by 2025 has been criticized for being too slow. Concerns have been raised about the added pressure these proposals would place on an already overstretched teaching workforce facing recruitment and retention difficulties, and about the level of funding needed to ensure teachers have well-developed training for the vital role of designated senior mental health lead (Education and Health and Social Care Committees, 2018). While training programs can improve teachers' confidence and skills to deal with children's emotional needs (Place2Be, 2015), budgetary and workload pressures mean training opportunities are unlikely to be available in all schools, discussed further below. With the number of young people with a diagnosable mental health condition standing at one in ten (Department of Health, and Department for Education [DOH and DFE], 2017), strategies aimed at addressing the causes and symptoms of mental health problems must be adequately resourced if they are to be effective.

Another area for consideration when tackling domestic violence in children's lives is the role of school nursing. The number of school nurses has been reduced in recent years. They are not always represented at child in need meetings, nor is relevant information always shared with them (Ofsted et al., 2017). Furthermore, term time working arrangements for school nurses mean they are not available during school holidays. Their restricted availability was referred to in the SCR of Daniel Pelka. School nurses attending relevant meetings and being employed during school holidays could facilitate greater consistency of care, better informed assessments, and improved multi-agency working.

Coming to school hungry is not conducive to learning and some schools provide breakfast and breakfast clubs. Since those living in poverty are at increased risk of domestic violence, having breakfast at school at no cost, or reduced cost, can be a valuable means to aid learning. For those impacted by domestic violence breakfast clubs can be an opportunity for quality time for parents and young children attending together (Sterne and Poole, 2010). An evaluation of breakfast clubs set up in high deprivation areas in the United Kingdom found reduced hunger in students, enhanced concentration and behavior, and improved social skills (Graham et al., 2017). Many schools offer homework clubs too. Those in temporary accommodation as a result of domestic violence may lack space or computer access to do homework and homework clubs at school can be a facilitator of learning. Extra-curricular activities and afterschool clubs can also provide positive experiences. While the cost of extra-curricular activities is sometimes prohibitive, schools can provide confidential financial assistance, although some parents may be reluctant to seek financial help. Domestic abuse based on coercive control is another possible impediment to participation. Because coercive control can result in the abused parent, predominantly mothers, and their children becoming isolated and lacking opportunities for relationships with those beyond their immediate family, after-school clubs might be denied to these children but where participation is permitted they can be a means for children to develop social skills and confidence (Katz, 2016).

Paramount to effectively supporting students is the adoption of a holistic, child-centered approach. If teachers are aware of issues in students' home lives, they will be better informed to provide tailored support to meet the individual needs of students regarding their learning, and social and emotional development. School staff need to be able to confidently ask students if anything is wrong at home and take appropriate action (Mullender et al., 2002). Research with young people affected by domestic violence found they valued teachers, tutors, learning mentors and school counselors in helping to identify abuse and access support (Howarth et al., 2016). In terms of educational attainment, additional learning support, perhaps in a one-toone or small group context, could help improve the educational outcomes of students. This would require a sensitive approach, however, particularly as students get older and may not wish to be singled out from peers.

Early intervention strategies to help children and young people experiencing domestic violence can be strengthened through organizations engaging in joined-up thinking and working. An illustrative example is Operation Encompass, an early intervention initiative being piloted in selected areas of the United Kingdom which entails police notifying a school by 9 am if a child has witnessed or experienced a domestic abuse incident the previous evening<sup>1</sup> . A key adult at school (the Designated Safeguarding Lead or Deputy) is informed of the case and cascades information to teaching staff to allow immediate and ongoing support to be given to the child. As a traumainformed charity Operation Encompass takes into account the child's past trauma, where applicable, and the child's responses and coping strategies. Operation Encompass can explain to the school why a child is absent or has been dropped off at school by someone else. The initiative is enabling police and schools to work in partnership to mitigate the impact of abuse and has the potential to be an exemplar of collaboration.

The Freedom Programme<sup>2</sup> is another initiative being run, including in schools, to teach and empower victims of domestic violence to recognize signs of abuse and make positive changes in their lives. Organizations interested in the program need to make a commitment in terms of ensuring their staff are trained in the program and have time for implementing it. The Freedom Programme can also work with children due to start school and has proved effective in bringing about positive change in women's and children's lives.

Schools play a role in providing help when dealing with the fallout of domestic violence in others ways too, such as setting up practical arrangements to minimize the risk of child abduction by the offending parent following parental separation (National Children's Home, 1994). Volunteering at their children's school, where appropriate, can serve as a way of non-abusive parents spending more time with their children and helping to protect them (Hamby, 2014).

Perhaps unsurprisingly, some teachers feel overwhelmed dealing with issues facing both children and their parents. Pressure on teachers to address problems among children and parents, such as increasing mental health issues, have left some feeling they are becoming like social workers. Mullender et al. (2002, p. 219) emphasize the importance of teachers listening to children vulnerable to domestic violence and offering emotional support:

This is not the same as becoming social workers, which teachers understandably fear in an already over-stretched working life and without the necessary training. Rather, it means being an effective channel for children to gain access to welfare services outside of school, by opening up an early opportunity for them to confide that something is wrong.

Despite good practices taking place in schools and with partner organizations, funding cuts in the United Kingdom have meant some support services for victims of domestic violence are no longer available (Lloyd and Ramon, 2017; Ofsted et al., 2017). Survey findings from domestic violence support services in England show that 60 per cent of respondents cited funding cuts, and the associated uncertainty, as their most significant challenge (Women's Aid, 2018). Reduced funding has led to services being unable to offer support to all women and children referred to them, loss of welfare service staff, and lower capacity to deal with increasing referrals of women with complex needs. Children and parents, predominantly mothers, living with domestic violence have been impacted by cuts to services resulting in schools taking on a greater role in supporting them. The role played by schools in supporting vulnerable children has implications for how teachers work with other agencies and in the next section I will look at how there can be tension between increased school autonomy and agencies working together.

# THE PARADOX OF GREATER SCHOOL AUTONOMY AND WORKING TOGETHER

Difficulties documented in research and governmental reports concerning inter-organizational working may be exacerbated by government policy devolving greater power to individual schools. Previously, state schools were funded by government and run by the local authority. Academy schools, initiated under the Labor government, and free schools under the Coalition government and subsequent Conservative government, are still state-funded but are not overseen by the local authority; they receive funding directly from central government affording them increased budgets. With budgetary independence and increased autonomy for their own governance these schools are able to set the pay and conditions for staff rather than abiding by national teacher pay and conditions required of local authorityrun schools. Academies and free schools are attended by over two-thirds of secondary school students and a quarter of primary school students (Department for Education [DFE], 2017b). Despite academy and free schools still being expected to liaise closely with local authorities on matters such as child protection and safeguarding, they have greater self-determination in shaping the relationship they have with local authorities (Baginsky et al., 2015).

Academy and free school status also has a bearing on school staff training and continuing professional development (CPD) opportunities. Although local authorities are still a provider of CPD, increasingly schools are buying-in, frequently expensive, CPD and training from private providers (National Association of Schoolmasters Union of Women Teachers [NASUWT], 2018). Schools are thus operating in a market system, especially pertinent now the majority of secondary schools are academies with budgetary autonomy. Schools can choose between 'marketleading' training providers who offer consultancy services. Funding for school staff training comes, in part, from Pupil Premium grants given to schools in England to support the education of the most disadvantaged students. Based on rates for 2017–2018, for each student eligible for free school meals, their school will receive a payment of £1,320 (primary) and £935 (secondary). Current school practice is for the cost of staff training to be paid for by Pupil Premium and from a school's own budget. Their budgetary independence means academy and free schools will have greater freedom to determine the nature and extent of staff training by external private providers. There have, however, been cuts in real terms in United Kingdom funding for education since 2010 (Belfield et al., 2018) posing implications for school budgets and accordingly for staff pay

fpsyg-09-02094 November 10, 2018 Time: 13:42 # 8

<sup>1</sup>http://www.operationencompass.org/

<sup>2</sup>http://www.freedomprogramme.co.uk/

and training. A survey of 1,615 teachers found budgetary and workload barriers impeded their access to training: 'Teachers report that their school does not have enough money to fund training/CPD and that external training/CPD is often very expensive' (National Association of Schoolmasters Union of Women Teachers [NASUWT], 2018, p. 14).

Greater school autonomy has additional implications for Local Safeguarding Children Boards (LSCBs) whose role is to coordinate local work to safeguard children. As a multi-agency body LSCBs are attended by representatives from the local authority and relevant organizations such as health services and the police. However, research by Baginsky and Holmes (2015) indicates that increasing fragmentation of educational services has seen academy schools (including free schools), and private fee-paying (non-state) schools being represented on less than half of LSCBs. While over 80 percent of boards were represented at senior level by local authority schools, the same was true of only 20 percent of boards attended by academy schools (Baginsky and Holmes, 2015).

# DOMESTIC VIOLENCE AND STUDENT WELLBEING WITHIN AN ATTAINMENT-DRIVEN EDUCATION SYSTEM

Given the spectrum of behavioral responses to domestic violence teachers need to be attuned to changes in children, some becoming withdrawn, others disruptive. Confrontational responses can, however, be difficult to account for: 'If underlying contributory factors are not obvious or understood, those children are likely to be labeled as problematic' (Ofsted et al., 2017, p. 14). This can lead to school staff misinterpreting students' behavior and disciplinary action might ensue. Indeed, data show a growing number of students excluded from school have mental health needs (Education and Health and Social Care Committees, 2018), and children impacted by domestic violence (Ofsted et al., 2017) and/or living in poverty (House of Commons Education Committee, 2018) are more likely to be excluded from school in comparison to their peers. This is worrying in the context of schools focusing on examination results and league tables. Teachers and educationalists lament the marketization of education whereby examination results have become a key measure by which schools define themselves and are defined by others, and schools are set in competition with each other in the form of league tables (Berry, 2016; Berry, 2017; Scott and Scott, 2018). Research with teachers shows such changes are negatively impacting teacher-student relationships and student wellbeing, with teachers reporting having less time to attend to the needs of individual students, and reporting that their own stress levels sometimes adversely affect their interaction with students (Hutchings, 2015). Baginsky et al. (2015, p. 358) discern tension between the prioritization of examination results and children's wellbeing:

Potentially there may be an inherent conflict between, on the one hand, pressure on institutions to demonstrate high levels of academic attainment and discipline by pupils in a competitive educational "market" and, on the other, the role of schools in recognizing and meeting the pastoral needs of children who are vulnerable or disadvantaged.

Where teacher performativity and student outcome measures in the form of examination results are at variance with the more holistic nurturing of students, efforts to support those impacted by domestic violence could be hampered and diminished. Concerns about schools becoming 'exam factories' have led to calls for a rebalancing of the education system whereby schools do not give precedence to academic outcomes at the expense of student wellbeing and personal development.

# CONCLUSION

What happens in childhood and adolescence has profound implications for wellbeing in adult life. The prevalence of domestic violence as the most common factor cited in cases of children in need in England in 2016–2017 (Department for Education [DFE], 2017a) emphasizes the need for addressing this enduring problem through prevention, early intervention and education. So too is wider attitudinal and social change needed whereby domestic violence is no longer trivialized as 'just another domestic' or portrayed as the fault of, predominantly women, victims, as evidenced in our earlier research into media representations (Lloyd and Ramon, 2017). Domestic violence must be addressed as a public health concern and not only as a privatized, individualized problem. The ways in which gender violence is based on and reinforced through women's wider structural inequality and lack of power in relation to men needs to be recognized if violence within the domestic sphere is to be tackled effectively.

Encouragingly there is some evidence of domestic violence research in the context of education, though it remains relatively under-investigated. The continuing fragmentation of the United Kingdom school system and plurality of school types highlight the need for increased research to evaluate schools' engagement in multi-agency working and to gain insight into effective practice. Some teachers and school support staff are themselves victims of domestically violent relationships and workplace support would be beneficial both for individuals and the school setting as a whole. Future research could usefully ask teachers and support staff their views on their professional learning and training needs in this important area of work.

Too frequently blame, shame and guilt cast a shadow over lives affected by domestic violence. Multi-agency working and in-school education and support can help prevent abuse and optimize outcomes for children, young people and their families living with the consequences of domestic violence.

# AUTHOR CONTRIBUTIONS

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

#### REFERENCES

fpsyg-09-02094 November 10, 2018 Time: 13:42 # 10


https://england.shelter.org.uk/\_\_data/assets/pdf\_file/0011/1474652/2017\_12\_ 20\_Homelessness\_and\_School\_Children.pdf doi: 10.1001/jama.286.24.3089


**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 © 2018 Lloyd. 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-09-02094 November 10, 2018 Time: 13:42 # 11

# The Willingness to Intervene in Cases of Intimate Partner Violence Against Women (WI-IPVAW) Scale: Development and Validation of the Long and Short Versions

Enrique Gracia\*, Manuel Martín-Fernández, Miriam Marco, Faraj A. Santirso, Viviana Vargas and Marisol Lila

Department of Social Psychology, University of Valencia, Valencia, Spain

#### Edited by:

Luca Rollè, Università degli Studi di Torino, Italy

#### Reviewed by:

Cristina Sechi, University of Cagliari, Italy Jaime Barrientos, Universidad de Santiago de Chile, Chile

> \*Correspondence: Enrique Gracia enrique.gracia@uv.es

#### Specialty section:

This article was submitted to Gender, Sex and Sexuality Studies, a section of the journal Frontiers in Psychology

> Received: 08 April 2018 Accepted: 14 June 2018 Published: 17 July 2018

#### Citation:

Gracia E, Martín-Fernández M, Marco M, Santirso FA, Vargas V and Lila M (2018) The Willingness to Intervene in Cases of Intimate Partner Violence Against Women (WI-IPVAW) Scale: Development and Validation of the Long and Short Versions. Front. Psychol. 9:1146. doi: 10.3389/fpsyg.2018.01146 Willingness to intervene when one becomes aware of a case of intimate partner violence against women (IPVAW) reflects the level of tolerance and acceptance of this type of violence in society. Increasing the likelihood of intervention to help victims of IPVAW is also a target for prevention strategies aiming to increase informal social control of IPVAW. In this study, we present the development and validation of the Willingness to Intervene in Cases of Intimate Partner Violence (WI-IPVAW) scale. We report data for both the long and short versions of the scale. We analyzed the latent structure, the reliability and validity of the WI-IPVAW across four samples (N = 1648). Factor analyses supported a bifactor model with a general non-specific factor expressing willingness to intervene in cases of IPVAW, and three specific factors reflecting different intervention preferences: a preference for setting the law enforcement process in motion ("calling the cops" factor), a preference for personal intervention ("personal involvement" factor), and a preference for non-intervention ("not my business" factor). Configural, metric, and partial scalar invariance across genders were supported. Two short versions of the scale, with nine and six items, respectively, were constructed on the base of quantitative and qualitative criteria. The long and short versions of the WI-IPVAW demonstrated both high reliability and construct validity, as they were strongly related to the acceptability of IPVAW, victim-blaming attitudes, perceived severity of IPVAW, and hostile sexism. These results confirm that both the long and short versions of the WI-IPVAW scale are psychometrically sound instruments to analyze willingness to intervene in cases of IPVAW in different settings and with different research needs (e.g., long versions for clinical and research settings, and short versions for large population surveys). The WI-IPVAW is also useful for assessing prevention policies and public education campaigns design to promote a more responsive social environment in cases of IPVAW, thus contributing to deter and reduce this major social and public health problem.

Keywords: intimate partner violence, violence against women, willingness to intervene, public attitudes, intervention preferences, help-giving, bystander intervention, measurement

# INTRODUCTION

fpsyg-09-01146 July 13, 2018 Time: 16:9 # 2

The World Health Organization defines intimate partner violence against women (IPVAW) as a "global public health problem of epidemic proportions" (World Health Organization [WHO], 2013, p. 7). IPVAW has profound consequences not only for the physical and psychological health of victims, but also for the well-being of their children, and for society in general (e.g., Campbell, 2002; Ellsberg et al., 2008; Devries et al., 2011; World Health Organization [WHO], 2013; Guedes et al., 2016). IPVAW is considered the most common form of violence suffered by women (Garcia-Moreno et al., 2006; Devries et al., 2013; Stöckl et al., 2013). In high-income countries, the estimated prevalence of IPVAW is 23.2%, and the percentage of IPVAW homicides, 41.2% (World Health Organization [WHO], 2013). In Europe, a survey among the 28 European Union (EU) Member States estimated that an average of 22% of European women had been victims of physical and/or sexual violence by their partners since the age of 15, with a lifetime prevalence across countries ranging from 13 to 32% (European Union Agency for Fundamental Rights, 2014). In Spain, where this study was conducted, various sources estimate IPVAW lifetime prevalence at around 13%, among the lowest in the EU (Vives-Cases et al., 2011; European Union Agency for Fundamental Rights, 2014; Ministerio de Sanidad, Servicios Sociales e Igualdad, 2015; Gracia and Merlo, 2016).

An ecological model recognizes that beyond individual and relational explanatory levels, larger contextual and societal factors are central to understand IPVAW (Heise, 1998, 2011; World Health Organization [WHO], 2002; Gracia et al., 2015a). As Gracia and Lila (2015, p. 16) pointed out, 'violence against women is a complex phenomenon that needs to be understood within the wider social context and within the social and cultural norms that permeate it.' Public attitudes toward IPVAW shape the social context in which IPVAW takes place and play an important role in perpetuating the levels of this type of violence in our societies (Carlson and Worden, 2005; Flood and Pease, 2009; Waltermaurer, 2012; Gracia and Lila, 2015; Copp et al., 2016; Powell and Webster, 2018). Public willingness to intervene when one becomes aware of a case of IPVAW reflects the level of tolerance and acceptance of this type of violence and can contribute either to deter or facilitate it (Browning, 2002; Gracia and Herrero, 2006; Emery et al., 2011; Wright and Benson, 2011; World Health Organization [WHO], 2013; Jewkes et al., 2015). In the current study, we set out to develop a scale measuring public willingness to intervene in cases of IPVAW.

One reason for studying willingness to act in cases of IPVAW is that, despite still being a largely unreported offense, at the same time IPVAW is widely known in the victims' social environment (Gracia, 2004; Taylor and Sorenson, 2005; Taylor et al., 2016). For example, in a survey across the 28 European Union member states, nearly 23% of respondents reported knowing a woman among their family members or friends who had been victim of IPVAW, 17% reported knowing women in their immediate neighborhood, and 9% knew a woman where they worked or studied (European Commission, 2016). Those who are aware of IPVAW incidents are in a position to do something to help the victims and stop the violence (e.g., offering help, taking personal action, or setting the law in motion), but they can also choose not to get involved, to ignore the situation, and do nothing (Banyard and Moynihan, 2011; Taylor et al., 2016). Therefore, whether or not those who are aware of this violence are willing to intervene is a not a trivial matter.

Attitudes of non-intervention in the victim's social circle may facilitate or reinforce the perpetrator's behavior, but may also inhibit victims' disclosure, making it more difficult for them to seek help and escape the violence. On the other hand, prointervention attitudes (e.g., reporting to the authorities or direct intervention) among those aware of this violence can have a protective effect for victims, and may inhibit or deter IPVAW by increasing the social and legal costs for perpetrators (Koepsell et al., 2006; McDonnell et al., 2011; Gracia, 2014; Voith, 2017). Willingness to intervene among those who are aware of IPVAW incidents is also relevant because victims tend to seek help among informal sources of help (friends, family, neighbors, coworkers, etc.) rather than formal sources such as the police (Liang et al., 2005; Ansara and Hindin, 2010; McCart et al., 2010; McDonnell et al., 2011; Wee et al., 2016). Moreover, pro-intervention attitudes among these potential informal sources of help, when shared collectively, can contribute to shape local social norms that help deter this type of violence (Wee et al., 2016; Voith, 2017; Powell and Webster, 2018). As Voith (2017, p. 4) noted in her review, "the protective effects of pro-IPV-intervention norms in a community are twofold, in that community members will directly intervene if they witness IPV and perpetrators are less likely to continue the use of violence against their partners as a result of social pressure".

Another reason to study and accurately measure public willingness to act in cases of IPVAW is that evidence suggests non-intervention attitudes are still quite prevalent, as shown in one report on attitudes toward violence against women in the EU (Gracia and Lila, 2015). For example, data from surveys carried out in different countries indicate that a sizable number of respondents preferred not to get involved even if they were aware of a case of violence against women ("not my business," or "is a private matter" were among the reasons given for not intervening). In addition, across the EU (European Commission, 2016), the most common reason given by those who knew victims of domestic violence but did not speak about it to anyone was that it was "none of their business" (26%). "Lack of proof " (18%), "not wanting to create trouble" (16%), "concerned about negative consequences or retaliation" (11%), "did not know who to speak to" (8%), and "it was not serious enough" (6%), were some other reasons. In Spain, where the present study was conducted, most of the officially reported cases of IPVAW are made by the victims themselves, and only around 4% of such reports come from family members or other third parties (Consejo General del Poder Judicial, 2016). Increasing the likelihood that people will intervene to help victims of IPVAW is therefore a target for prevention strategies aiming to translate public awareness of this social problem into a greater sense of personal responsibility and involvement, thus contributing to the informal social control of IPVAW (Gracia et al., 2009).

# Present Study

fpsyg-09-01146 July 13, 2018 Time: 16:9 # 3

Drawing from the above, there is an evident need to advance our knowledge about public willingness to intervene in cases of IPVAW and related key issues such as the prevalence of pro- or non-intervention attitudes, intervention preferences, its correlates or determinants, or assessing the effectiveness of interventions targeting these attitudes. The availability of reliable and valid instruments measuring public willingness to intervene in cases of IPVAW is central to this type of research. Although some measurement instruments have been developed to examine willingness to help in cases of violence, most of this research has been conducted in the context of bystander intervention behavior in cases of dating violence, and sexual harassment or rape situations (Stein, 2007; Banyard, 2008; Banyard and Moynihan, 2011; Branch et al., 2013; Banyard et al., 2014; McMahon et al., 2014). Other studies assessing willingness to intervene have limited generalizability as they use small non-community samples (e.g., college students), and others instruments report low reliabilities (Baldry and Pagliaro, 2014; Baldry et al., 2015; Cinquegrana et al., 2018). In addition, data from large population surveys on public attitudes toward intervention in cases of IPVAW are not usually based on measurement instruments with adequate reliability and validity, or rely on single items (Gracia and Lila, 2015). Clearly, there is still a need for psychometrically sound instruments measuring willingness to intervene in cases of IPVAW, appropriate for use with community samples, and suitable for large-scale surveys.

In this study, we present the development and validation of the Willingness to Intervene in Cases of Intimate Partner Violence (WI-IPVAW) scale. We aim also to develop reduced versions of the full WI-IPVAW scale, as large population surveys or studies with limited space or time require the use of short forms while retaining adequate psychometric properties (Smith et al., 2000; Goetz et al., 2013). By reporting data for both the long and short versions of the scale, we aim to provide tools to analyze willingness to intervene in cases of IPVAW in different settings and with different research needs (e.g., long versions for clinical and research settings, and short versions for large population surveys). By using advanced statistical analyses, we will address important issues such as social desirability and measurement invariance and ensure that the shortened versions of the WI-IPVAW scale retain high quality psychometric properties.

For validity purposes, we will explore the relationship between the long and short versions of the WI-IPVAW scale and other relevant constructs regarding attitudes toward IPVAW such as IPVAW acceptability, victim-blaming attitudes, perceived severity of IPVAW, and hostile sexism (Taylor and Sorenson, 2005; Gracia and Herrero, 2006; Flood and Pease, 2009; Lila et al., 2013; Gracia, 2014; Herrero et al., 2017; Martín-Fernández et al., 2018b). Gender, age and education differences in willingness to intervene in cases of IPVAW will be also explored (Carlson and Worden, 2005; Fincham et al., 2008; Flood and Pease, 2009; Gracia et al., 2009; Gracia et al., 2015b). Attitudes of acceptability of IPVAW have been considered a key issue to understand IPVAW prevalence in society (Flood and Pease, 2009; Gracia et al., 2015b; Copp et al., 2016; Martín-Fernández et al., 2018b). These attitudes have been linked to public, professionals, and victims' perceptions and responses to IPVAW (Taylor and Sorenson, 2005; Gracia and Herrero, 2006; Rizo and Macy, 2011; Gracia et al., 2014). We hypothesize that the lower the IPVAW acceptability, the greater the willingness to intervene in cases of IPVAW. Victim-blaming attitudes are also among those factors often used to explain and justify IPVAW. These attitudes can influence public responses toward known cases of IPVAW (Liang et al., 2005; Ansara and Hindin, 2010; Gracia, 2014; Gracia and Tomás, 2014). We expect that lower scores of victimblaming attitudes will be associated with greater willingness to intervene in cases of IPVAW. The perceived severity of IPVAW incidents may also influence responses to IPVAW (Gracia et al., 2009, 2014). According to Latané and Darley's (1970) model of bystander intervention, perceived severity is a precondition to the decision to intervene. According to this model, if some incidents of IPVAW are perceived as not serious enough, bystanders will be less willing to intervene (Gracia et al., 2009). We anticipate that the greater the perceived severity of IPVAW, the greater the willingness to intervene in cases of IPVAW. Hostile sexism is a gender prejudice manifestation that conveys negative images and beliefs about women (Glick and Fiske, 1996), and has been related to attitudes toward intervention in cases of IPVAW (Lila et al., 2013; Herrero et al., 2017). We hypothesize that the lower the hostile sexism, the greater the willingness to intervene in cases of IPVAW. Finally, gender, age and education differences in willingness to intervene in cases of IPVAW will be also explored (Carlson and Worden, 2005; Fincham et al., 2008; Flood and Pease, 2009; Gracia et al., 2009, 2015b).

## MATERIALS AND METHODS

#### Participants

Four samples were recruited for the current study. The first one was an incidental sample used to conduct a pilot study, composed of 148 Valencia University undergraduates who participated for course credits (31 males and 117 females), aged 19–32 years old (M = 21.29; SD = 2.60). The second, third, and fourth samples were recruited through online sampling. Online sampling is an effective and cost-efficient sampling method (Thornton et al., 2016; Topolovec-Vranic and Natarajan, 2016). A total pool of 2,698 responses was collected. We equilibrated these samples by gender and removed those participants who were younger than 18 years old, omitted socio-demographic information, or were duplicated responses. Participants from samples 2, 3, and 4 were randomly drawn from the remaining pool of responses. The socio-demographic characteristics of the samples are shown in **Table 1**.

The second sample consisted of 500 participants (231 males and 269 females), aged 18–80 (M = 33.83; SD = 14.77), and was used to study the psychometric properties of the scale. The third sample consisted of 1000 participants (490 males and 510 females), aged 18–82 (M = 35.40; SD = 13.46). This sample was used to test different levels of measurement invariance and to conduct the criterion-related validity analyses. The fourth sample consisted of 200 participants (94 males and 106 females), aged

TABLE 1 | Socio-demographics.


18–71 (M = 29.39; SD = 11.82), and was used to assemble two short versions of the scale.

#### Measures Willingness to Intervene in Cases of IPVAW (WI-IPVAW)

The development of the WI-IPVAW was based on an initial pool of 96 items. These items were developed from a review of European surveys addressing attitudes toward intervention in cases of violence against women (Gracia and Lila, 2015), and other previous research addressing public attitudes and response preferences in cases of IPVAW (Gracia and Herrero, 2006; Gracia et al., 2009). The item development and selection process was also informed by literature identifying scenarios where IPVAW also takes place, other than behind closed doors at home, and is witnessed by third parties (Banyard and Moynihan, 2011; Hamby et al., 2015; Taylor et al., 2016). This initial pool of items presented hypothetical scenarios describing IPVAW situations, occurring in different places, and that could be witnessed by the respondent, or disclosed to him/her by the victim (e.g., next door apartment, staircase or communal areas in buildings, street, shops, bars, etc.). These scenarios included various expressions of IPVAW behaviors (e.g., physical aggression, insults, threats, violent arguments, fights, etc.), and different types of potential responses or involvement (i.e., calling the police, scolding or reprehending the aggressor, protecting the woman victim, ignoring the situation, doing nothing, etc.). The initial pool of items was then reviewed by a panel of six experts on IPVAW to establish construct representativeness and clarity (Beck and Gable, 2001; Delgado-Rico et al., 2012). The experts were asked to rate the representativeness (i.e., whether the item is suitable to measure willingness to intervene in cases of IPVAW), and the clarity (i.e., how concise the item is) of the items on a 7-point Likert-type scale (1 = "Very unrepresentative/unclear"; 7 = "Very representative/clear"). An item was considered representative and/or clear if the average score in the expert ratings was above 5 on the 7-point scale (i.e., the "somewhat representative/clear" category). After this review, 31 items were selected. Respondents were asked to rate their perceived likelihood of intervening in the hypothetical scenario described in each item on a 6-point Likert-type scale (1 = "Not at all likely," 6 = "Extremely likely"). The final version of the WI-IPVAW scale is shown in Appendix 1 (see Supplementary Material).

#### Acceptability of IPVAW (A-IPVAW; Martín-Fernández et al., 2018b)

The short form of the A-IPVAW scale was used in this study. This instrument is composed of eight items tapping attitudes of acceptability of IPVAW (e.g., It is acceptable for a man "to shout his partner if she is continuously arguing and nagging him"). Respondents rated the acceptability of a range of men's behaviors against their female partners on a 3-point Likerttype scale (0 = "Not acceptable," 1 = "Somewhat acceptable," 2 = "Acceptable"). The A-IPVAW scale was cross-validated in the general Spanish population, and also with IPVAW male offenders. This scale has showed adequate internal and external validity, as it has been related to perceived severity of IPVAW and ambivalent sexism (Martín-Fernández et al., 2018b). Our results showed reasonable internal consistency across Samples 2, 3, and 4 (Cronbach's α = 0.75, 0.72, 0.68, respectively).

#### Victim-Blaming Attitudes Toward IPVAW (VB-IPVAW; Martín-Fernández et al., 2018a)

This instrument is composed of five items assessing the tendency to blame victims of IPVAW (e.g., "A man will change his behavior toward his partner if she becomes more obedient"). Respondents rated their level of agreement with each statement on a 4-point Likert-type scale (1 = "Strongly disagree," 4 = "Strongly agree"). Evidence of the instrument's validity has been demonstrated based on its relationships with other variables such as the acceptability and perceived severity of IPVAW, and ambivalent sexism (Martín-Fernández et al., 2018b). It also presented high internal consistency in Samples 2, 3, and 4 (Cronbach's α = 0.81, 0.84, 0.83, respectively).

#### Perceived Severity of IPVAW (PS-IPVAW; Gracia et al., 2009, 2011)

This scale presents eight IPVAW scenarios (e.g., "During an argument, a man hits his partner and then asks her to forgive him"), the severity of which respondents assessed on a 10 point Likert-type scale (ranging from 1, "Not severe at all," to 10, "Extremely severe"). The PS-IPVAW scale has previously been validated in the general Spanish population, and also with police officers and male IPVAW offenders, presenting adequate psychometric properties. It has also been related to sexism, empathy, personal responsibility, and IPVAW victim-blaming attitudes (Gracia et al., 2009; Lila et al., 2013; Gracia and Tomás, 2014; Vargas et al., 2015). The scale showed good internal consistency in Samples 2, 3, and 4 (Cronbach's α = 0.83, 0.85, 0.87, respectively).

#### Ambivalent Sexism Inventory Short Version (ASI; Glick and Fiske, 1996; Rollero et al., 2014)

The reduced hostile sexism subscale was used for the current study, composed of six items assessing attitudes of prejudice and discrimination against women based on the assumption of women's inferiority and their differences from men (e.g., "Women seek to gain power by getting control over men"). The Spanish version of the items was used (Expósito et al., 1998). The complete ambivalent sexism inventory has been validated in more than twenty countries (Glick et al., 2000, 2002), and the hostile sexism subscale has demonstrated strong relationships with attitudes toward intervention in IPVAW cases among police officers, IPVAW responsibility attribution, and acceptability of IPVAW (Lila et al., 2013, 2014; Martín-Fernández et al., 2018b). It presented good internal consistency in Samples 2, 3, and 4 (Cronbach's α = 0.89, 0.88, 0.87, respectively).

#### Balanced Inventory of Desirable Responding Short Form (BIDR-16; Hart et al., 2015)

The Impression Management subscale was used for the pilot study. This subscale is composed of eight items evaluating the tendency of participants to provide overestimated selfdescriptions to create a socially desirable image (e.g., "I never cover up mistakes"), and presented moderate reliability in the first sample (Cronbach's α = 0.68).

#### Procedure

Two online forms were designed to collect the data. The first form included the WI-IPVAW, the BIDR items of the Impression Management subscale, and a set of socio-demographical questions (i.e., gender, age, nationality, and education level). This form was used only for Sample 1. Participants were informed about the objectives of the study and gave their informed consent, agreeing to participate in the study if they press the "continue" button. The second form included the WI-IPVAW, the PS-IPVAW, the short forms of the A-IPVAW, VB-IPVAW, Hostile Sexism, and the same socio-demographical questions. After the participants had given their informed consent and agree to participate in the study, they completed the online form. Participants received no payment. The data were collected from October 2017 to December 2017.

#### Data Analysis

A pilot study was conducted first using the sample of college students (Sample 1) in order to explore the psychometric properties of the WI-IPVAW and the effect of social desirability on the items. One of the major threats to the content validity of any scale assessing personality traits or attitudinal components is the social desirability bias. This bias is a major concern when the assessment involves socially sensitive issues, as IPVAW (Grimm, 2010). Therefore, the aim of this preliminary evaluation was to refine the instrument before administering it to a larger sample. To this end, the descriptive statistics and the item-test corrected correlations were computed, and the internal consistency of the scale was evaluated by means of Cronbach's α. The latent structure of the scale was also assessed through an exploratory factor analysis (EFA). Before conducting the EFA, the suitability of the data matrix was tested, computing Bartlett's sphericity test and the Kaiser-Meyer-Olkin (KMO) statistic. To determine the number of factors to extract, a parallel analysis based on minimum rank factor analysis was conducted (Timmerman and Lorenzo-Seva, 2011). An EFA was then performed using the polychoric correlation matrix and the weighted least-squares means and variances adjusted estimation method (WLSMV), as this procedure is especially recommended for categorical data (Muthén and Kaplan, 1985, 1992; Asparouhov and Muthén, 2010). The fit of the model was assessed using the CFI, TLI, SRMR, and RMSEA fit indices. CFI and TLI values ≥ 0.95 are indicative of very good fit, and values between 0.90 and 0.95 indicate minimally acceptable model fit (Bentler, 1995; Hu and Bentler, 1999). RMSEA values ≤ 0.06, and ≤0.08, indicate very good and acceptable fit, respectively, and SRMR values ≤ 0.08 are considered to reflect well-fitting models (MacCallum et al., 1996). Once the latent structure of the scale had been established, the social desirability of each item was evaluated. To do so, a confirmatory factor analysis (CFA) was conducted with the addition of a social desirability factor to the EFA model. All the items of the BIDR and the WI-IPVAW scale were constrained to load onto this social desirability factor, using the BIDR items as social desirability markers (Ferrando, 2005, 2008). To make the model identifiable, the loadings of the BIDR were fixed to the same value. If a WI-IPVAW item loading on the social desirability factor was greater than the BIDR loadings, we considered the item to be biased by social desirability. Those items were removed from the scale.

A larger sample (Sample 2) was used to study further the psychometric properties of the WI-IPVAW scale and to crossvalidate the factorial model. The descriptive statistics, the itemtest correlations, and Cronbach's α were again computed. A CFA was carried out using the WLSMV estimation method. Several nested models were compared. Model fit was evaluated using the same combination of fit indices and the same cut-offs.

Measurement invariance across genders was also evaluated in an independent sample (Sample 3). To this end, several levels of group invariance were tested by conducting and comparing a series of multi-group CFAs. Configural, metric, scalar and strict invariance models were estimated using the WLSMV estimation method (Milfont and Fischer, 2010). Configural invariance tests whether men and women conceptualize the construct in the same manner, estimating the same factorial model for each group and allowing the structural parameters (i.e., loadings, thresholds, and item variances) to vary across groups. The metric invariance model constrains the item loadings to have the same value for both groups, testing whether men and women interpret the items in the same way. The scalar invariance model fixes the threshold parameters to the same value across groups, establishing whether the latent construct yields the same score in the items for men and women. The strict invariance model assesses whether the measurement error is equal in each group, constraining the variances of the observed variables (i.e., the items) to have the same values across groups. The models were compared following the guidelines of Cheung and Rensvold (2002), computing the change in CFI (1CFI) and RMSEA (1RMSEA) to test which of the invariance models is better supported by the data. A change

in the CFI (1CFI) and in the RMSEA (1RMSEA) ≤ 0.010 and ≤0.015, respectively, support the more restrictive model (i.e., the configural model is the most flexible model and the strict invariance the most restrictive). However, these criteria were proposed for models estimated with maximum likelihood estimation for continuous variables and, given that we used weighted least-squares estimation for categorical data, we also ran a corrected chi-square difference test (DIFFTEST; Asparouhov et al., 2006). If the fit indices comparisons and the DIFFTEST yield a similar result, then that invariance level is accepted.

The validity of the scale was assessed by relating it to other relevant IPVAW variables, namely, acceptability of IPVAW, attitudes of victim blaming in cases of IPVAW, perceived severity of IPVAW, and hostile sexism. Socio-demographic comparisons were also made, testing differences across gender, age, and education level groups.

Finally, two short versions of the WI-IPVAW scale of nine and five items were created following Goetz et al. (2013) recommendations. First, the most relevant items were selected attending to the internal consistency, the previous factorial models, and the assessments of the expert panel. The psychometric properties of the shortened scales were then studied and compared with the original WI-IPVAW scale using a different sample (Sample 4).

All analyses were computed using the statistical package R (R Core Team, 2017) and the psych library (Revelle, 2016). EFA, CFA, and multi-group CFAs were conducted with the MPlus 7.1 package (Muthén and Muthén, 2010).

#### RESULTS

#### Pilot Study: Factor Structure and Social Desirability

The psychometric properties, the latent structure and the effect of social desirability on the WI-IPVAW items were explored in a pilot study with Sample 1. Descriptive statistics revealed that most of the items were slightly displaced to the right, with means around 3–5 (e.g., "somewhat likely," "quite likely," "very likely"), and moderate negative skew (around −0.50), indicating that the participants tended to select the upper categories of the scale. The overall internal consistency of the scale was very high (Cronbach's α = 0.93), showing a strong relation between the score on the scale and the items, with item-test corrected correlations around 0.50. Deleting items did not improve the scale's internal consistency.

Before conducting an EFA, the suitability of the matrix for factor analysis was tested. Bartlett's sphericity test was significant (χ <sup>2</sup> = 2505.8, df = 465, p < 0.001) and the Kaiser-Meyer-Olkin statistic was good (KMO = 0.88), indicating that the data were adequate for an EFA. The parallel analysis based on minimum rank factor analysis using the polychoric correlation matrix revealed that three factors should be extracted, since adding more factors did not contribute to explain more variance in our data than in a random dataset. A three-factor model was thus estimated using WLSMV with the oblique OBLIMIN rotation. The model converged normally, and showed an acceptable fit (χ <sup>2</sup> = 2505.8, df = 465; CFI = 0.94; TLI = 0.92; RMSEA = 0.068; SRMR = 0.069). Although the CFI and the TLI were below the 0.95 cut-off, they were not below 0.90, and the RMSEA and SRMR suggested that the model was well-fitted. The items were grouped in three factors. The first factor groups all the items related to setting the law in motion by calling to the police or reporting the IPVAW incident (i.e., "calling the cops" factor), the second factor groups all items referring to ignoring the situation or doing nothing (i.e., "not my business factor"), and the third factor groups all items in which the respondents personally intervene to stop the situation (i.e., "personal involvement" factor). All the items presented factor loadings above 0.30 in their factor, and only three items presented cross-loadings in more than one factor. In these three cases the loadings on the main factor were above 0.50 and close to 0.30 in the secondary factor, indicating that the items were more related to the main factor (i.e., "personal involvement" factor in the first case, and "calling the cops" factor in the other two cases). The correlation between the "calling the cops" and the "personal involvement" factors was positive (r = 0.29), whereas the correlations between the "not my business" factor and the "calling the cops" and the "personal involvement" factors were negative (r = −0.55 and r = −0.28, respectively).

A CFA was conducted to test the extent of the effect of social desirability bias on the scale items. The CFA model posited the three previous content factors (i.e., "calling the cops," "not my business," "personal involvement") and a new social desirability factor. The content factors were allowed to correlate with each other, whereas the social desirability factor was not correlated with any content factor. The WI-IPVAW items loaded on their main factor and also on the social desirability factor. The BIDR items were used as social desirability markers and only loaded on the social desirability factor. In addition, the BIDR items were constrained to have the same factor loadings on this factor. The model was estimated using WLSMV, converged normally, and showed an adequate fit (χ <sup>2</sup> = 1130, df = 837; CFI = 0.93; TLI = 0.92; RMSEA = 0.049). The factor loadings are reported in **Table 2**.

Three items (e.g., "If a man insulted his partner in the street, I would say something to reprehend his action"; "If a man grabbed his partner's arm aggressively in the street, forcing her to go with him, I would call the police"; "If a new couple in my building argued and yelled constantly, I would call the police") presented factor loadings on the social desirability factor higher than the markers (λ = 0.37), and thus were removed from the scale. Ferrando (2005) recommends removing those items that present factor loadings above | 0.30| ; however, we decided to apply a more conservative criterion (i.e., removing only items that had factor loadings above the markers loading on the social desirability factor), since the internal consistency of the BIDR was moderate in the pilot study.

#### Descriptive Analyses and Reliability

Sample 2 was used to assess the psychometric properties of the scale. Descriptive statistics and item-test corrected correlations can be found in **Table 3**. The descriptive statistics were in the same line as in the pilot study, with items slightly displaced to the right. The item means were around 4, with a standard deviation around 1, meaning that the respondents tended to endorse the

TABLE 2 | Confirmatory factor analysis with social desirability markers (Sample 1).


Each cell contains the factor loadings (SE in brackets). Empty cells indicate that the item does not load on that factor. BIDR 1-8: items from the impression management subscale of the Balanced Inventory of Desirable Response Short Form. Bold values: items that presented loadings in the social desirability factor higher than the social desirability markers (i.e., BIDR 1–8).

upper intermediate categories (e.g., "somewhat likely," "quite likely," "very likely"). The skew statistics were moderate and negative for many of the items, and some of them also presented high kurtosis values, indicating that the items were not normally distributed. The item-test corrected correlations presented values above 0.40, indicating a strong relationship between the items and the total score of the scale. The overall internal consistency of the scale was again very good (Cronbach's α = 0.94), and the internal consistency of each factor was also good (Cronbach's α = 0.88, 0.84, and 0.92 for the "calling the cops," "not my business," and "personal involvement" factors, respectively).

#### Confirmatory Factor Analysis

Three models were estimated with Sample 2 to test the factor structure of the WI-IPVAW. The first model was a one-factor model in which all items loaded onto a general factor of "willingness to intervene in cases of IPVAW." The second model was the three-factor model resulting from the pilot study, with three correlated factors differentiated by the responses to the scenarios described by the WI-IPVAW items (i.e., "calling the cops," "not my business," and "personal involvement"). The third model was a bifactor model with three specific factors reflecting different intervention preferences—as in the previous threefactor model—and a general, non-specific factor, of "willingness to intervene." This general factor accounts for all the elements common to the specific factors. The specific factors account only for the core elements of their items, in this case the type of response to the scenarios described by the items. Thus, all the items loaded on their specific factor and also on the general factor. The factors were orthogonal, so they are not correlated. All models were estimated using WLSMV and the polychoric correlation matrix. All models converged normally.

The fit indices of the models are shown in **Table 4**. The onefactor model showed a poor fit to the data, presenting fit indices too far from their cut-offs. The three-factor model showed an acceptable RMSEA and a minimally acceptable CFI and TLI,



M, mean; SD, standard deviation; Min, minimum; Max, maximum; SE, standard error for the Skew and Kurtosis statistics. ritem-test, item-test corrected correlation.


CFI, comparative fit index; TLI, Tucker-Lewis index; RMSEA, Root Mean Square Error of Approximation (95% CI in square brackets).

which could be kept as the latent structure of the scale. However, adding a general dimension of "willingness to intervene" to the model substantially improved the fit of the model to the data. We therefore decided to retain the bifactor model.

The loadings of the bifactor model are displayed in **Table 5**. All the loadings for the specific factors were significant, with values above 0.30 in all the items except for items 2 and 3, whose loadings were around 0.20. The general factor loadings were all significant with values above |0.40|. Note that the "not my business" item loadings were negative in the general factor, reflecting that agreement with these items yielded a lower score on the general "willingness to intervene" factor. Overall, the general factor loadings were higher than in the specific factor. Furthermore, the percentage of common explained variance of the general "willingness to intervene" factor was 56.85%, whereas the specific "calling the cops" factor explained 23.16%, the "personal involvement" 11.04%, and the "not my business" 8.95% of the common explained variance.

#### Measurement Invariance

Having retained the bifactor model as the latent structure of the scale, the measurement invariance of the scale was tested across genders using Sample 3. Item 5 was removed from these analyses since there were not enough responses in the lower categories for either the men's or the women's groups. A stepwise approach was used, testing first the configural invariance, and then comparing it with the metric, scalar, and strict invariance models. The fit indices of the models and the model comparisons are shown in **Tables 6**, **7**.

The configural model showed a good fit to the data, indicating that men and women conceptualize the latent construct in the same manner, and was used as a base line for the model comparisons. Then it was compared with the metric invariance model, which constrained the factor loadings to be equivalent across groups; we found that both CFI and RMSEA indices

#### TABLE 5 | CFA item loadings on the bifactor model (Sample 2).


Each cell contains the factor loadings (SE in brackets). Empty cells indicate that the item does not load on that factor.



CFI, comparative fit index; TLI, Tucker-Lewis index; RMSEA, Root Mean Square Error of Approximation (95% CI in square brackets).

TABLE 7 | Measurement invariance model comparisons (Sample 3).


1CFI, change in CFI; 1RMSEA, change in RMESEA; DIFFTEST, robust chi square difference testing; df, degree of freedom of the DIFTEST; p, p-value of the DIFTEST.

improved once the factor loadings were constrained. The DIFFTEST also showed that these improvements were marginally significant (p = 0.02). This is most likely due to the reduction in the number of parameters to estimate, making the model more parsimonious, and it is not an unusual phenomenon when conducting measurement invariance analysis with categorical data (e.g., Brummelman et al., 2015; Megías et al., 2017). Given the improvement in model fit and the reduction in the number of parameters to estimate, the metric invariance was supported.

The scalar invariance model, which besides the factor loading also constrained the item thresholds to be equal across gender, was compared with the metric model. Although the reduction in the CFI and RMSEA fit indices were between the cut-offs established by Cheung and Rensvold (2002), the DIFFTEST was significant (p < 0.001). The modification indices were then used

to identify potential items to be unconstrained and test the partial scalar invariance model. The thresholds of two items (items 6 and 20) were allowed to vary across groups and we found that the partial invariance model did not differ from the metric model (p = 0.051). The partial scalar invariance model was thus supported.

Finally, the strict invariance model was tested, constraining the item variances to be equal across groups and comparing it with the partial invariance model. We found that the CFI decreased below the 1CFI = 0.01 cut-off and the DIFFTEST was significant. Thus the strict invariance model could not be supported.

#### Validity Analyses

Sample 3 was also used to conduct validity analyses. The correlations of the WI-IPVAW factorial scores with other related constructs are shown in **Table 8**. The general factor "willingness to intervene" was negatively related to acceptability of IPVAW, attitudes of victim blaming, and hostile sexism, implying that those respondents with higher scores on this factor tend to present lower levels of attitudes of acceptability, are less likely to blame victims of IPVAW, and show lower levels of sexist attitudes. On the other hand, the general factor was positively related with the perceived severity of IPVAW (those with higher scores on willingness to intervene tend to perceive IPVAW situations as more severe). Regarding the specific factors, the "calling the cops" factor showed a similar relation with these variables, although they were more moderate, whereas the "not my business" factor presented the opposite tendency: it was positively related with acceptability of IPVAW, attitudes of victim blaming, and hostile sexism, and negatively related to perceived severity of IPVAW. The "personal involvement" factor only presented a significant and negative relation to perceived severity.

A series of ANOVA were conducted with each factor to test differences across gender, age, and education level using the factor scores of the partial scalar invariance model. Regarding the general factor "willingness to intervene," significant differences were found between genders, F(1) = 23.53, p < 0.001, η <sup>2</sup> = 0.023, with a small effect size, women having higher values on this factor than men; marginal differences between age groups, F(3) = 3.09, p = 0.026, η <sup>2</sup> = 0.009; and no differences for education level, F(3) = 1.30, p = 0.274, η <sup>2</sup> = 0.004. The effect sizes of age and education levels were considered negligible, since they were


below the 0.01 cut-off for small size effects (Miles and Shevlin, 2001).

Significant differences were also found in the specific "calling the cops" factor by gender, F(1) = 21.24, p < 0.001,η <sup>2</sup> = 0.021, and age, F(3) = 3.73, p = 0.011, η <sup>2</sup> = 0.011, both with a small effect size. Women scored higher on this factor than men, as did the respondents of the upper age categories (i.e., 35–54 and 55+) in comparison with the lower category (i.e., 18–24). Education level had no significant effect on this factor, F(3) = 0.89, p = 0.444, η <sup>2</sup> = 0.002.

We found significant differences for the specific factor "not my business" by gender, F(1) = 5.45, p = 0.020, η <sup>2</sup> = 0.005, although the effect size was considered negligible. No differences were found in this factor for age, F(3) = 2.27, p = 0.079, η <sup>2</sup> = 0.006, or education level, F(3) = 2.27, p = 0.079, η <sup>2</sup> = 0.002.

Regarding the specific factor "personal involvement," significant differences were again found between genders, F(1) = 85.00, p < 0.001, η <sup>2</sup> = 0.079, with a medium effect size, and age groups, F(3) = 5.08, p = 0.002, η <sup>2</sup> = 0.015, with a small effect size. Men showed higher scores on this factor than women, and respondents in the upper age categories (i.e., 35–54 and 55+) presented higher scores than respondents in the lower age categories (i.e., 18–24 and 25–34). Again, education level had no effect on this factor, F(3) = 1.53, p = 0.197, η <sup>2</sup> = 0.005.

#### WI-IPVAW Shortened Forms

A combination of quantitative (i.e., social desirability loadings, bifactor model loadings, and whether items were invariant across genders) and qualitative criteria (i.e., the expert ratings) was used to decide which items should comprise the shortened versions of the scale (see **Table 9**). The items included were those that presented low loadings (i.e., below 0.20) on the social desirability factor used on the pilot study, with medium or high loadings (i.e., between 0.20–0.50, and above 0.50, respectively) on their specific and general factor, and that were invariant across genders. In addition to these criteria, the expert panel's assessment of the representativeness and clarity of each item was also considered.

#### Nine-Item Version of the WI-IPVAW Scale

To ensure content coverage, three items from each specific factor were selected to create a nine-item version of the WI-IPVAW scale (Smith et al., 2000), namely, items 2, 8, 9, 10, 12, 15, 16, 26, and 27. Although item 2 presented a low loading in the "calling the cops" factor, it was selected as it met the other criteria and the loading on the specific factor was close enough to the 0.20 cut-off for medium loadings (i.e., λ = 0.19). Sample 4 was then used to study the psychometric properties of the nine-item version of the scale. The internal consistency of this version was adequate (Cronbach's α = 0.77), and the item-test corrected correlations were above 0.30 for all items except for item 26, for which it was 0.28. The factor structure of the nine-item version presented an excellent fit to the data when the bifactor model was fitted using WLSMV estimation with the polychoric correlation matrix (χ 2 (18) = 33.01, CFI = 0.98, TLI = 0.97, RMSEA = 0.065 [90% CI 0.027; 0.099]). Evidence for the validity based on its relationships with other constructs is reported with correlations in **Table 10**, which are in the same direction as for the complete


Expert ratings: items rated as clear (C) and/or representative (R) by the panel of experts.

TABLE 10 | WI-IPVAW short forms relationships with other variables (Sample 4).


WI-IPVAW version. Finally, the correlation between the nineitem version and the complete scale was very strong, r = 0.92, t(198) = 32.81, p < 0.001, suggesting that both versions provided similar assessments.

#### Five-Item Version of the WI-IPVAW Scale

For circumstances in which space is very limited (e.g., large-scale surveys), a shorter version of the scale was created with a focus on the general factor. To this end, two items from the "calling the cops" and "personal involvement" factors and one item from the "not my business" factor were selected. These were the items that presented higher factor loadings on the general "willingness to intervene" factor in the nine-item version, namely, items 8, 9, 10, 12, and 27. Sample 4 was used to study the psychometric properties of this version of the scale. The internal consistency of the scale was again fair (Cronbach's α = 0.73), and the itemtest corrected correlations were above 0.30 for all items except for item 27 in this case, for which it was 0.27. A one-factor model was fitted to the five-item version of the scale since there were fewer than three items per specific factor, using WLSMV estimation. The model fitted reasonably well to the data (χ 2 (5) = 30.44, CFI = 0.96, TLI = 0.92, RMSEA = 0.150 [90% CI 0.099; 0.207]), although the residuals were below the 0.08 cut-off for a well-fitted model. The correlations between the "willingness to intervene" factor and the criterion-related variables were again in the same direction as for the complete version of the scale (see **Table 10**). The correlation between the five-item version and the complete version of the scale was high, r = 0.86, t(198) = 24, p < 0.001, although smaller than for the nine-item version.

# DISCUSSION

fpsyg-09-01146 July 13, 2018 Time: 16:9 # 12

In this paper, we described the development and psychometric properties of the long and short forms of the WI-IPVAW, a set of new self-report questionnaires assessing willingness to intervene in cases of IPVAW. Taken together, our results provide strong support for the reliability and validity of both the long and short versions of the WI-IPVAW scale.

Content validity of the WI-IPVAW was assessed during the scale development process using the ratings of a panel of experts, to ensure that the items adequately captured the different aspects of the construct. One of the advantages of the WI-IPVAW is that it also takes into account various community settings (next door house, streets, bars, etc.) where IPVAW can occur, as well as several expressions of this type of violence (e.g., verbal, threats, physical violence) in diverse situations and with different degrees of severity. The WI-IPVAW also includes a variety of potential responses to different IPVAW scenarios (e.g., talking to victims, personal involvement, calling the police, etc.). Tapping situationspecific responses across a range of settings provides greater ecological validity to this measure, and also facilitates future research on situational correlates of such attitudes (Carlo and Randall, 2002; Banyard, 2008; Banyard and Moynihan, 2011; Copp et al., 2016). Moreover, the effect of social desirability bias was controlled in a pilot study through a confirmatory factor analysis using social desirability markers (Ferrando, 2005, 2008). This analytical approach is one of the major strengths of the present study, because it allowed us to identify and remove items with higher loadings on the social desirability factor from the scale.

Regarding the internal structure of the scale, our results supported a bifactor model as the latent structure of the scale, as it presented the best fit to the data of all the models. In this model, each item loaded on one specific factor and also onto a general factor. This general factor (i.e., "willingness to intervene in cases of IPVAW") captures the common variance of all items, reflecting the shared elements of the measured construct. On the other hand, the specific factors (i.e., "calling the cops," "personal involvement," and "not my business") represent the remaining unique variance not attributable to the general factor. The model is orthogonal and thus the factors are uncorrelated, meaning that the general factor is assumed to be independent of the specific factors, and also that the specific factors are assumed to be different from and independent of each other (e.g., Chen et al., 2006; Gibbins et al., 2012). In addition, our results highlight the relevance of the general factor since most of the loadings presented higher values on the general factor than on their respective specific factor. The general factor also accounted for the largest proportion of the common explained variance, 56.85%. The "calling the cops" factor accounted for almost half of the remaining common variance, 23.16%, whereas the "personal involvement" and "not my business" specific factors explained the rest, 11.04 and 8.95%, respectively.

We also conducted measurement invariance analyses of the WI-IPVAW across genders. A partial scalar invariance model was supported, showing that men and women conceptualize the underlying latent structure in the same manner (configural invariance), that the scale unit is the same, and thus the items are interpreted similarly by men and women (metric invariance), and that the thresholds of the items are the same for both genders, as the factorial scores were comparable across gender groups (scalar invariance). However, the threshold parameters of two items (items 6 and 21) were allowed to vary across groups, implying that men and women do not share the same distribution on these items. To obtain comparable scores for men and women in the general "willingness to intervene" factor and in the specific factors, researchers and practitioners could remove items 6 and 21 from the scale. We recommend, however, using the invariant items as anchor items and treating these two items differently for each gender. To this end, we provide an Mplus syntax to compute this model in Appendix 2 (see Supplementary Material).

Regarding validity analyses based on the relationships of the WI-IPVAW with other variables, we found that the general factor (i.e., "willingness to intervene in cases of IPVAW") was significantly associated with a set of relevant variables linked to IPVAW. Thus, as expected, respondents with higher scores on the WI-IPVAW (i.e., those more willing to intervene), perceive IPVAW situations as more severe, find IPVAW less acceptable, have fewer victim-blaming attitudes, and score lower in hostile sexism. This supports the idea that willingness to intervene in cases of IPVAW reflects the personal level of tolerance and acceptance of this type of violence and suggests that attitudes toward intervention in cases of IPVAW are also linked to attitudes justifying IPVAW, such as victim blaming, and to hostility toward women (Glick et al., 2002; Taylor and Sorenson, 2005; Gracia et al., 2014; Herrero et al., 2017; Ivert et al., 2018). With respect to the specific factors, both "calling the cops" and "not my business," were related as expected (i.e., the first positively and the second negatively) with the same set of variables. For example, those scoring high in the "not my business" factor tended to perceive IPVAW as less severe and more acceptable and scored higher in both victim-blaming attitudes and hostile sexism. Interestingly, the "personal involvement" factor was related, negatively, only with the perceived severity of IPVAW, suggesting that the more severe an IPVAW situation is perceived, the more other intervention preferences are favored, as greater personal costs or negative consequences may be involved. For example, as Gracia et al. (2009) observed, reporting incidents of IPVAW to the police is more likely among those who tend to perceive these incidents as more severe.

In this study, we also developed two shortened versions of the WI-IPVAW scale. The full WI-IPVAW scale is a relatively lengthy questionnaire. The length of questionnaires often prevents their inclusion in population surveys where space is limited and expensive, or in studies where time is an issue. Large-scale surveys tend to resort to single items addressing these attitudes or use a set of questions with unknown reliability or validity (Richins, 2004; Gracia and Lila, 2015). On the other hand, shortened versions can have the drawback of limited reliability and validity, which makes it particularly important to ensure that short versions of questionnaires retain their psychometric soundness (Smith et al., 2000; Stanton et al., 2002; Kovacs et al., 2017). As Smith et al. (2000) point out, rigorous application of psychometric principles

is crucial when validating short forms. In the present study, two short nine- and six-item versions of the parent WI-IPVAW scale were constructed based on quantitative and qualitative criteria (Goetz et al., 2013), supporting the adequate transfer of validity from the parent form of the WI-IPVAW to the two short forms. The complete and short versions of the WI-IPVAW demonstrated high reliability as well as construct validity as they were strongly related to acceptability of IPVAW, victimblaming attitudes, perceived severity of IPVAW, and hostile sexism. Although some loss of reliability is inevitable, our results provide strong empirical support for the high quality of their psychometric properties of the short versions of the WI-IPVAW scale. When research or survey needs (large-scale surveys, limited space or time, etc.) require the use of short forms, our results demonstrate that both the nine- and the five-item short forms are reliable and valid alternatives to the most comprehensive and broader assessment of willingness to intervene in cases of IPVAW provided by the long version of the WI-IPVAW (both reduced versions presented a high correlation with the parent WI-IPVAW scale). For example, the nine-item WI-IPVAW short scale showed not only adequate reliability, but also allowed meaningful assessment of both the general non-specific factor expressing the willingness to intervene in cases of IPVAW, and the three specific factors reflecting different intervention preferences (adequate representation of the construct is ensured by incorporating three items from each of the specific factors of the original scale). In turn, the five-item WI-IPVAW short scale is particularly recommended for the reliable and valid assessment of the general "willingness to intervene" factor when space and/or time constraints are an issue, but this construct is still important for research or policy-making purposes. The five-item version only mapped the general factor as there were not enough items to preserve the original latent structure of the scale. The scores on the general factor of the five-item version presented a similar pattern when related to acceptability of IPVAW, attitudes of victim blaming, perceived severity, and hostile sexism.

This study is not without limitations. Although social desirability was controlled in the pilot study following the procedure proposed by Ferrando (2005), the items used as social desirability markers presented a mediocre reliability, and thus these results should be taken with caution. Regarding the measurement invariance, although the partial scalar invariance level for the WI-IPVAW across genders was supported, further research is needed to establish whether this instrument is also invariant across age and education level groups. The online sampling method is another limitation of the study, as it has some tradeoffs that limit its generalizability. Although this sampling strategy is effective for obtaining large sample sizes in a short period of time and is also cost-effective, it is more difficult to verify the socio-demographical information provided by the participants (Thornton et al., 2016; Topolovec-Vranic and Natarajan, 2016). Self-selection bias is another issue, since the respondents who agreed to participate might also be those that are more motivated. In addition, it is important to note that the WI-IPVAW was developed in the Spanish sociocultural context. Spain is among the countries with the lowest IPVAW lifetime prevalence in the EU (Vives-Cases et al., 2011; European Union Agency for Fundamental Rights, 2014; Gracia and Merlo, 2016). This is particularly interesting given that other European countries have considerably higher levels of gender equality than Spain (Gracia and Merlo, 2016). As to whether these differences in prevalence are linked to differences across countries regarding public attitudes such as willingness to intervene in cases of IPVAW, future research is needed to adapt and validate the WI-IPVAW scale to other cultural settings (Gracia and Lila, 2015; Boira et al., 2016).

The study also has practical implications. Addressing attitudes towards IPVAW, such as willingness to intervene in cases of IPVAW, and advancing in their conceptualization, measurement, prevalence, and determinants is central to monitoring social changes in such attitudes and to better informing prevention and intervention strategies (Powell and Webster, 2018). Public willingness to intervene in cases of IPVAW reflects the level of tolerance and acceptability of IPVAW, and when these attitudes are held collectively at different levels of aggregation (e.g., social groups, neighborhoods, communities, countries), they are able to create a social climate that can help to legitimize or deter this type of violence (Browning, 2002; Emery et al., 2011; Heise, 2011; Wright and Benson, 2011; Heise and Kotsadam, 2015; Voith, 2017; Marco et al., 2018). For example, a public education strategy should consider targeting those social groups or communities were IPVAW risk is higher, and these attitudes can be more commonly held (Gracia and Tomás, 2014; Gracia et al., 2015a). In this regard, the different versions of the WI-IPVAW—especially the short versions, which are more appropriate for survey type research—can be used to assess pro- or non-intervention norms at different aggregation levels, such as neighborhoods or communities, when they are considered as key targets for social and community intervention strategies addressing the prevalence of IPVAW and its correlates, such as public attitudes (Gracia, 2014; Gracia et al., 2015a; Voith, 2017). As Klein et al. (1997, p. 90) state, "we need to educate people to recognize that they have a role in helping battered women and to teach them that their behavior matters, and showed them how to get involved." In this regard, and in line with Gracia et al. (2009), public education efforts must promote attitudes that reinforce the helping role of the victim's social circle in order to increase feelings of social and personal responsibility about the high prevalence of IPVAW in our societies. Increasing the likelihood of public intervention to help IPVAW victims, not only among the general public but also within professional groups (social services, health, law enforcement, etc.), can contribute to deter and reduce this major social and public health problem (Gracia et al., 2014; Ferrer-Perez et al., 2016; López-Ossorio et al., 2016; Touza-Garma, 2017). The WI-IPVAW therefore offers a useful instrument to assess prevention policies and public education campaigns aiming to promote a more responsive social environment in cases of IPVAW.

#### ETHICS STATEMENT

This study was performed in accordance with the Declaration of Helsinki. Informed consent information was supplied and

implied through participation in the on-line survey. The study and protocol were reviewed and approved by the University of Valencia Ethics Committee.

#### AUTHOR CONTRIBUTIONS

EG conceived the study and supervised the writing of the manuscript. MM-F designed the analytic strategy, contributed to developing materials, conducted the statistical analysis, and wrote the methods and results sections of the manuscript. MM contributed to developing materials, data collection, and writing some parts of the manuscript. FS contributed to developing materials, data collection, and writing some parts of the manuscript. VV contributed to developing materials, data collection, and writing some parts of the manuscript. ML coordinated the data collection and contributed to the writing of the manuscript.

## REFERENCES


#### FUNDING

This research was supported by the Spanish Ministry of Economy, Industry, and Competitiveness (PSI2017-84764-P). MM-F was supported by the FPI program of the Spanish Ministry of Economy, and Competitiveness (BES-2015- 075576). MM was supported by the FPU program of the Spanish Ministry of Education, Culture, and Sports (FPU2013/00164). FS was supported by the FPU program of the Spanish Ministry of Education, Culture, and Sports (FPU2015/00864).

#### SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyg. 2018.01146/full#supplementary-material


in Chicago neighborhoods. Am. J. Community Psychol. 48, 373–383. doi: 10.1007/s10464-010-9362-5



**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 © 2018 Gracia, Martín-Fernández, Marco, Santirso, Vargas and Lila. 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.

# New Directions for Preventing Dating Violence in Adolescence: The Study of Gender Models

#### Chiara Santoro, Belén Martínez-Ferrer, Carmen Monreal Gimeno\* and Gonzalo Musitu

Department of Education and Social Psychology, Pablo de Olavide University, Seville, Spain

Dating violence is a huge transcultural and alarming phenomenon, directly linked with endless discrimination against women. The latest research on dating violence in adolescence shows how dating violence is persistent and common in the adolescent period as well and pinpoints the origin of gender violence from first adolescent relationships. This element takes us to considerate how recent gender violence studies and policies, increased also thanks to international efforts on this issue, are not bringing expected results, especially among young people. This mini-review aims to analyze the main characteristics of current gender studies and policies on dating violence, focusing on percentages with a woman-centered approach, which stresses the consequences of gender violence. Other gender studies, that consider gender as a relational product, stress the importance of integrating the analysis of gender models as a key instrument to understand the main causes of dating violence, providing new elements to develop effective policies against dating violence. Indeed, gender models of femininity and masculinity are based on a binary system, which is also a reciprocal recognition and identity system: gender models define female and male characteristics, roles, stereotypes, and expectation, being complementary and foreclosing at the same time. Recent studies on gender relationships, especially among the youth, allows us to propose a new dialog between dating violence studies and gender model studies, underling the need of a complete and complex understanding of gender structure, and of its tensions and contradictions, to put an end to gender and dating violence, through effective programs.

Keywords: dating violence, adolescence, gender models, gender policies, prevention

# INTRODUCTION

In the context of gender violence, dating violence is a particularly worrying phenomenon. The World Health Organization (WHO) affirms that one in three women has been a victim of physical or sexual violence by an intimate partner at some point in her lifetime and, globally, as many as 38% of all murders of women are committed by intimate partners (WHO, 2013). Recent studies highlight a progressive increase in gender violence in intimate relationships, especially in adolescent population (Karakurt and Silver, 2013; Rodríguez and Megías, 2015; Taylor and Mumford, 2016). Considering these stunning data, dating violence is considered a human rights violation and public health issue throughout the world (Campbell, 2002; Garcia-Moreno et al., 2006). This fact leads us to the need of understanding what elements are contributing to the resistance and to the increase of dating violence.

#### Edited by:

Luca Rollè, Università degli Studi di Torino, Italy

#### Reviewed by:

Cristiano Scandurra, Università degli Studi di Napoli Federico II, Italy Amparo Osca, Universidad Nacional de Educación a Distancia (UNED), Spain

\*Correspondence:

Carmen Monreal Gimeno mcmongim@upo.es

#### Specialty section:

This article was submitted to Gender, Sex and Sexuality Studies, a section of the journal Frontiers in Psychology

> Received: 14 April 2018 Accepted: 23 May 2018 Published: 12 June 2018

#### Citation:

Santoro C, Martínez-Ferrer B, Monreal Gimeno C and Musitu G (2018) New Directions for Preventing Dating Violence in Adolescence: The Study of Gender Models. Front. Psychol. 9:946. doi: 10.3389/fpsyg.2018.00946

**176**

In this mini-review, we propose an analysis of current studies on dating violence (for example, Fernández and Fuertes, 2010; Martsolf et al., 2012; Tapp and Moore, 2016), that focused the problem on percentages and statistics and on a woman centered perspective, often from a paternalistic approach. Later, we insert the study of gender models as a key element to shift attention from the study of the consequences or factors involved in gender violence to the analysis of the causes and of the structural elements that contribute to the reproduction of gender relations supported on a binary, complementary and excluding basis, starting from a gender relational perspective. To conclude, this perspective is proposed as innovative especially for its constant dialog with social movements and with the claims of collective rights and history of the different contexts, allowing the management and evaluation of gender policies in a more precise way to achieve equality between men and women.

# DATING VIOLENCE: BETWEEN VICTIMIZATION, PATERNALISM AND PERCENTAGES

Violence against women is a theme that draws the attention of state governments, international and global organizations for its pervasiveness, cultural transversality, and persistence. Gender violence is an urgent subject on the agenda, whose visibility has certainly been increased in recent years thanks to international pressure and to increasingly pervasive campaigns, allowing the observation of some of the main trends of gender studies and main gender policies. However, it has usually been perceived as a female problem. On the one hand, this fact sheds lights on the endless discrimination toward women, which involves them in a persistent struggle that constantly affects their lives. On the other hand, it strengthens a type of intervention that focuses more on the consequences than on the causes of the phenomenon, for example giving assistance to female victims of violence and trying to understand psychological, social and health consequences. There is a trend strongly built on helping women and the analysis of their condition as victims, which does not consider the relational structure in which women are inserted (Taylor and Mumford, 2016). In fact, a significant proportion of women victims of gender violence have experienced a long history of polyvictimization and revictimization (Fernández-González et al., 2017).

The need to make the phenomenon more visible, a particularly difficult operation for its main development in private space, has produced various studies that validate main international programs, which highlight the number of women victims of gender violence and dating violence (for example, WHO, 2013; European Union Agency for Fundamental Rights, 2014). These studies, fundamental for the understanding of the pervasiveness of the phenomenon, do not, however, allow a deeper analysis of the causes and of "gender order" (Connell, 1987). Gender violence is the top of the iceberg, the most extreme and definitive consequence of a patriarchal system that reproduces unequal gender relations, hierarchically ordered, complementary and excluding, based on a dual basis, masculinity and femininity gender models. This violence is instrumental, and its goal is to control and submit women who do not agree with patriarchal models.

# THE STUDY OF GENDER MODELS: FROM CONSEQUENCE TO CAUSE ANALYSIS OF DATING VIOLENCE

Starting from the second feminist wave, and from the need, no longer postponed, to include women in the dominant power systems and to tackle a deep revision. We begin to focus our attention on how the patriarchal system creates structures of domination and subordination based on the existence of a binary reference system that is essential on the basis of sexual difference; this system is based on the model of masculinity and femininity. For gender models, we mean the set of characteristics, values, attitudes, roles, expectations that are expected of a person identified biologically as a man or a woman. This concept is also known as gender schema (Monreal Gimeno and Martínez Ferrer, 2010). The principal characteristics of the gender models as key elements in the reproduction of patriarchy:


TABLE 1 | Summary of masculinity and femininity models main characteristics (Santoro, 2018).


from the very early age through an education conforming to our gender reference models are not often perceived as a restriction on individual possibilities, but as a matter of fact, as the only and possible configuration of reality. Moreover, in this socialization phase, the process reproduces a specific gender model, in an exclusionary way.

(4) Gender models are dynamic. We are not defined monolithically by the gender socialization we receive: gender identity is created in a certain gender culture, which limits the possibilities of expression and self-realization, giving us a definite set of options. At the same time, gender identity depends on our personal experiences, in the form of education we receive, on contacts we have with other realities, on our ethical choices.

# GENDER RELATIONSHIPS AND DATING VIOLENCE: A NEW PERSPECTIVE IN GENDER STUDIES

Integrating gender model analysis into the study of dating violence means shifting attention from a woman as a violence victim, to gender relations, taking into attention how and why this violence is the result of relationships of power socially and culturally built and reproduced. This perspective can only be integrated into the understanding of gender concept as relational and multidimensional. The internal differences of gender studies highlight the complexity of the gender concept and identify at least three, continually interconnected dimensions (Santoro, 2018). Gender is above all a personal identity variable, which allows the definition of others and myself. Gender also has a cultural, historical and social dimension. The structural dimension of gender, as a cultural system for organizing sexual difference. Moreover, gender is inserted into a power structure: in fact, its normative contents are constantly organized hierarchically, positively or negatively valuing, on the basis of binarism and heteronormativity (Tinat, 2016), that is shown in discrimination against no binary identity and sexual identities (Rollè and Marino, 2011; Ciocca et al., 2017).

Starting from a relational perspective, gender relations are compromise between social and cultural content, the dimension of power and personal gender identity. Including these elements in the study of dating violence means integrating a certain complexity that has two main consequences: analyze the phenomenon by going back to its possible causes, and. to make clear how gender violence, in all its forms and manifestations, is the most extreme consequence of a complementary, exclusionary and unequal relations system, which feeds on gender-normative content and also contributes to its own identity definition. Despite these historical changes, the binary reference structure of gender models of masculinity and femininity remains in force, causing movements, dynamics and contradictions that can be related to dating violence.

Firstly, there is a greater resistance of gender models by men than by women. In fact, if the model of femininity has been the subject of a profound revision process, which has led to important changes in the lives of women, that of masculinity has remained fixed. Secondly, there is a constant contradiction between practices and beliefs related to gender models. Despite women's integration in both education and in the workplace the normative models which constitute the sphere of ideas and beliefs continue to be the gender material which constitutes and regulates the society in which we live. As Monreal Gimeno and Martínez Ferrer (2010) states, gender patterns maintain certain autonomy despite the real changes in the characteristics associated with this group.

Starting from these considerations, and from the results of recent research on gender relations and dating violence, especially in adolescents, it is clear how these elements are related. The freedom of women, which extends beyond the model of femininity that sees women as mother, wife, passive, sensitive, docile and attentive to others, puts in crisis the same identity recognition of man, with whom it relates, which becomes an expression of the norm that re-establishes the predetermined order. Moreover, the same binary power system is based on excluding opposites, incommunicable, but at the same time complementary. At the same time, this difference is justified by the myth of romantic love or the "soul mate". In this discourse, man and woman, however different, are complementary and need this union to reach fullness (Cubells and Calsamiglia, 2015).

# CONCLUSION AND FUTURE DIRECTIONS

The perpetuation of violent relationships that have seen women victims of every kind of abuse, deprivation, humiliation and physical and psychological damage, over the centuries goes hand in hand with a certain gender training that is put in place in order to reproduce an order of established gender hierarchy, which relies on models of masculinity and femininity that selfconfirm and justify each other. For this reason, the dialog between studies on gender models with those on dating violence, allows us to reconstruct the network of gender relations in which we are constantly immersed to understand the tendencies, contradictions, difficulties that are at the base of the phenomenon of gender violence. Furthermore, this new understanding allows to have the elements for a direct intervention on differential socialization process, that transmit different gender values for men and women (McCarry, 2010; Santoro, 2018). In fact, gender violence is not characterized by its exceptionality, but by its complexity: considering gender as a multidimensional and relational variable, it is possible to understand what the causes of this phenomenon are, provoking an important change of perspective with respect to the development of policies aimed at preventing gender-based violence. For this reason, gender equality intervention should be started in early age, and directed to boys and girls, working on deconstruction of gender roles and gender attributes, striking gender as a vehicle of personal and continuous identity construction, a chance to express yourself starting from your own strengths and debilities, not limited by models or role. Working on gender models means a continuous process in which education institutions could strike a central role (Biemmi, 2015), giving the opportunity to find a right and

comfort space to debate, to experience, to express and create an equality culture and values. Especially, focusing on emotional education (Leathwood and Hey, 2009). and spreading dialog on real and daily discrimination cases and on instruments that allow permanent stereotypes reproduction, as mass media and social network (Tortajada, 2013).

These policies must consider the need to put in tension, to open spaces of possibilities, to eliminate the normativity of gender models that continue to be static and to influence people's lives. Similarly, and starting from a perspective that includes gender as relational, it seems fundamental to exit from a vision that binds gender violence as a "feminine" problem: the construction of unequal relationships originates from a binary basis, which sees equally involved men and women. For consequence, it's fundamental to integrate men as a direct actor and beneficiary of gender policies. It has been pointed out that there are cultural differences in aggressors in aspects related to gender models such as the acceptability of violence against women in relationships, sexist beliefs about male domination and honor, and the role of women as caregivers of the family (Lila et al., 2013; Vargas et al., 2015). At the same time, integrating the understanding of these elements to the study of dating violence means working on the development and implementation of efficient and at the same time complex measuring instruments; research should not be focused only on percentage and statistics on gender violence's victims and perpetrators, but also on the diffusion of an equality culture and, on the opposite, on gender models' resistance. Focus on ideas and beliefs states a change

#### REFERENCES


from consequences to a causes approach to gender violence, and, for instance, from restraint to prevention policies. In general, it means looking at complexity in gender studies, which increasingly embraces an approach to transversality, relationality, intersectionality, and which can, therefore, be the basis for the development of gender policies that are truly effective with respect to the integral goal of equality between men and women. Today, we can develop integral equality plan, directed to ends with gender discrimination, only taking into account and knowing our context's gender culture resistances and changes, helping men and women to reflect and reconstruct their own relationships starting from their own needs and opportunities, instead of on their own fear to not be social recognized as "good," "complete," "proper" man and woman.

# AUTHOR CONTRIBUTIONS

CS, BM-F, CM, and GM had participated in the intellectual content, the analysis of data, and the writing of the work. CS, BM-F, CM, and GM had reviewed the final version of the work and they approve it for publication.

# FUNDING

This mini-review is based on project results and funding thanks to FP7-PEOPLE-2012-IRSES N◦ 318960 GENDERCIT Project.

study on women's health and domestic violence. Lancet 368, 1260–1269. doi: 10.1016/S0140-6736(06)69523-8



WHO (2013). Global and Regional Estimates of Violence Against Women: Prevalence and Health Effects of Intimate Partner Violence and Non-Partner Sexual Violence. Available at: http://www.who.int/reproductivehealth/ publications/violence/9789241564625/en/ [accessed May 13, 2018].

**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 © 2018 Santoro, Martínez-Ferrer, Monreal Gimeno and Musitu. 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 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.

# Intimate Partner Violence and Child Custody Evaluation: A Model for Preliminary Clinical Intervention

#### Marialuisa Gennari, Giancarlo Tamanza and Sara Molgora\*

Department of Psychology, Università Cattolica del Sacro Cuore, Milan, Italy

Intimate partner violence is defined by the World Health Organization as "any behavior within an intimate relationship that causes physical, psychological, or sexual harm to those in the relationship" and it refers to a specific relationship dynamic. In recent decades, an increasing number of studies have focused on this phenomenon, considering its exponential growth over time. Many studies have focused on risk factors for violence within the couple relationship. This paper specifically analyses the association between violence and separation or divorce. Although many interventions have been developed over the years, the effectiveness of extant interventions on violent behaviors is not yet empirically supported. Since clinical experience allows to affirm that both partners can be involved in treatment for intimate partner violence especially during mandated proceedings, the present study focuses on domestic violence in separated couples involved in a child custody evaluation process. In this case, literature supports the need for individualized assessment in order to promote the best intervention according to the specific conditions of each partner, whether the battered one or the perpetrator. However, little research has been done on child custody evaluation in the presence of violent couples. The aim of the present study is to present a model of couple clinical intervention with a separated violent couple in the context of a child custody evaluation. This model can be defined as relational-intergenerational and its main aim is to understand the exchange between familial generations and to search for factors that safeguard and care for family relations. Furthermore, according also to the therapeutic assessment approach, there is an intrinsic connection between assessment and "family transformative potential." This paper presents the specific working methodology underlying this model, through the description of a single clinical case. In particular, the proposed model provides a multi-dimensional assessment comprising three levels: individual, evaluating parents' history through representations, thoughts, and feelings; interpersonal, investigating the different relations; discussion and dialogue with the parental couple about findings.

Keywords: intimate partner violence, separation, divorce, child custody evaluation, relational-intergenerational approach, therapeutic assessment, single case

#### Edited by:

Shulamit Ramon, University of Hertfordshire, United Kingdom

#### Reviewed by:

Silvia Mazzoni, Sapienza Università di Roma, Italy Carles Perez-Testor, Universitat Ramon Llull, Spain

> \*Correspondence: Sara Molgora sara.molgora@unicatt.it

#### Specialty section:

This article was submitted to Gender, Sex and Sexuality Studies, a section of the journal Frontiers in Psychology

> Received: 13 April 2018 Accepted: 26 July 2018 Published: 17 August 2018

#### Citation:

Gennari M, Tamanza G and Molgora S (2018) Intimate Partner Violence and Child Custody Evaluation: A Model for Preliminary Clinical Intervention. Front. Psychol. 9:1471. doi: 10.3389/fpsyg.2018.01471

**181**

# INTRODUCTION

Intimate partner violence (IPV) is defined by the World Health Organization as "any behavior within an intimate relationship that causes physical, psychological, or sexual harm to those in the relationship" (World Health Organization, 2012). Specifically, IPV generally refers to a specific relationship dynamic in which affection and aggression are combined (Chester and DeWall, 2018), and violent behaviors occur as an ongoing pattern of abuse (Sugg, 2015). IPV can be non-reciprocal (i.e., perpetrated by only one partner) or reciprocal (i.e., both partners are violent); in the latter case, violent behaviors can occur in different ways (Whitaker et al., 2007).

In recent decades, an increasing number of studies have focused on IPV, considering the exponential growth of this phenomenon over time (World Health Organization, 2013). Research has analyzed outcomes of IPV, focusing on the negative impact of violence on the psychological and physical well-being of partners (e.g., post-traumatic stress disorders, generalized anxiety disorder, depression, health-compromising behaviors, etc.), including over the long-term (Zlotnik et al., 2006; Bosch et al., 2017; Pickover et al., 2017; Spencer et al., 2017). This impact could be mediated by personality characteristics such as temperament traits (Yalch et al., 2017).

The wide dissemination of this phenomenon over the years and the evidence of its negative effects on partners' health underscores the importance of developing interventions for IPV. Specifically, clinicians and researchers are called to develop instruments in order to screen couples at risk for violent behaviors and to prevent the escalation of violence between partners. However, a variety of factors can prevent partners from reporting violence, thus reducing the possibility of access to services and interventions (Spangaro et al., 2016; Gennari et al., 2017).

Although many interventions have been developed over the years, the effectiveness of extant interventions on violent behaviors is not yet empirically supported (Stover et al., 2009). In particular, greater efforts have been made to provide services to support victims, whereas less attention has been paid to intervention programs for batterers (Ferrer-Perez and Bosch-Fiol, 2018). Even less attention has been paid to intervention programs for the couple (i.e., both partners together), also considering that the opportunity for couple treatment is controversial precisely due to the relational asymmetry usual present in violent relations (Beach et al., 2004; Holtzworth-Munroe et al., 2005). In this regard, Kelly and Johnson (2008) suggested the need for effective intervention programs tailored to the specific characteristics of partner violence.

Clinical experience allows us to affirm that both partners can be involved in treatment for intimate partner violence especially during mandated proceedings, that is: (a) in mandatory evaluations carried out by social services in cases of multiproblematic families; and (b) in cases of child custody evaluations. In the first case, social services, after having obtained authorization from appropriate judicial authority, launch investigations and evaluations regarding parenting skills after being alerted by different actors in the social context (school, neighborhood, sports groups) in order to protect children from variously problematic family situations. In the second case, the court requests an intervention in order to supervise the conditions of the couple's separation or divorce.

In this paper we focus on this second condition. Specifically, the aim of the present study is to present a model of couple clinical intervention with a separated violent couple in the context of a child custody evaluation through the description of a single clinical case. In this paper, moreover, we make reference to the type of intimate partner violence that involves acts of verbal and physical aggression (injuries caused by blows and/or objects) because it is one of the most common and clearly identifiable forms of violence (World Health Organization, 2012).

# VIOLENCE AND DIVORCE

Research has focused on risk factors for violence within the couple relationship, and many variables have been analyzed as predictive of violent behaviors (e.g., childhood experiences and history, socio-demographic characteristics, intrapersonal and interpersonal variables, biological factors) (see for example, Capaldi et al., 2012; Dim and Elabor-Idemudia, 2017; Goodnight et al., 2017; Chester and DeWall, 2018). In particular, this paper focuses on the association between violence and separation or divorce. Some studies underscored that violent behaviors could lead to the breakdown of the couple relationship (Davidson and Beck, 2017). At the same time, separation (or divorce) could be considered as a risk factor because it can make possible the emergence as well as the escalation of violence within the couple (Stolzenberg and D'Alessio, 2007; Toews et al., 2008; Ellis et al., 2014). Considering that divorce in itself is a critical and potentially traumatic event (Cigoli and Scabini, 2006; Parmiani et al., 2012), the presence of violent behaviors between partners makes this process more challenging, increasing the risk of maladaptive outcomes for partners (e.g., depressive symptoms, etc.) (Rutter, 2005) and for children who are exposed to a double stress (Bernet et al., 2016): parental separation and violent behaviors. Furthermore, the presence of violence in couples during the separation process could lead to more negative postdivorce outcome both in terms of general agreements between partners and the co-parenting relationship, raising important issues about child custody (Lessard et al., 2014). The exercise of parental roles also depends on the specific type of violent behaviors (Davidson and Beck, 2017; Hardesty et al., 2017).

Regarding interventions for separated or divorced violent couples, literature supports the need for individualized assessment in order to promote the best intervention according to the specific conditions of each partner, whether the battered one or the perpetrator (Beck et al., 2013; Hardesty et al., 2016). However, little research has been done on child custody evaluation in the presence of violent couples (Saunders et al., 2015). More research is needed to respond to some open questions such as whether the type of violence makes a difference and whether and when shared parenting could be practicable for violent couples (Saunders, 2015).

# VIOLENCE AND CHILD CUSTODY EVALUATION

Child custody evaluation makes it possible to obtain a clinical space within a social-judicial mandate aimed at the parental couple in order to reorganize the family relations after a separation in the best way possible (Gennari et al., 2014). This mandate is defined by the judge to whom the partners have appealed, asking that "justice be rendered" in a situation perceived to be unjust or prejudicial for oneself or for one's children.

It is important to highlight the unique characteristics of child custody evaluations to understand the possibilities of clinical intervention even when situations of intimate partner violence exist. This context is characterized by transfert (from the couple to the judge) that should be correctly understood and considered (Cigoli, 1998). The characteristics of the partners' petition to the judge, contained in the court proceeding documents, always have important meanings concerning the partners' needs and fears as well as their goals and objectives, of which they are not always conscious. In any case, the consultant as the judge's competent and trusted expert, accepts from both the judge and the couple the task of rendering justice, acknowledging and establishing rights and wrongs. This is the particular intervention setting that, differently from what usually happens, makes it possible to eviscerate and treat violence and enables the partners to entrust themselves to the consultancy precisely in the hope that the wrongs they have received can be rectified. We can thus affirm that this specific intervention setting promotes the trust that one needs to be able to expose one's pain and suffering, including that of violence. It must also be added that in these situations, the judge has preliminarily directed that the partners live separately, often imposing a certain physical distance (restraining order) between them: this is an element that gives the partners the necessary peace of mind to be able to work with the clinic, reducing the fear of violent reprisals.

The purpose of the child custody evaluation is to provide the judge with useful information for establishing the best living conditions for the children as he/she decides on custody, residence, and visitation rights between children and the parent with whom they do not live on a daily basis. Thus, this is a parenting assessment intervention. In this context, the partners often ask to have custody of the children, and for this very reason they are highly motivated to convince the consultant of their good behavior, both as a person and a parent. In this scenario, therefore, it is not uncommon for the parents to disclose incidents of violence with particular vehemence and in great detail, even when these actions did not occur. In short, we can affirm that the child custody evaluation is an assessment setting where violence is brought up very naturally and is often accentuated even as a means to getting custody of the children. In cases of violence, it is thus important for the consultant, even before evaluating the resources and problematic aspects of the partners and their relationship, to evaluate three aspects connected to violent behavior: power, model, and primary perpetrator of the violence (Jaffe et al., 2008). Distinguishing between various types of violence makes it possible to evaluate its seriousness and thus the risk for the children as well as the necessity of putting into place protective mechanisms for the child in the custody decision process, and at the same time to understand the couple dynamics of violent behaviors.

In this regard, for example, some authors (Lebow and Rekart, 2007; Jaffe et al., 2008; Kelly and Johnson, 2008) identified four types of violence in the context of child custody evaluation: (1) Abusive-controlling violence (ACV), also called battering or intimate terrorism or coercive controlling violence, that is, the use of coercive behaviors (e.g., threat, force, emotional abuse, etc.) by a partner to dominate the other inducing fear, submission, and compliance; (2) Conflict-instigated violence (CIV), also called situational or common couple violence, that is, the perpetration of violence by both partners who have limited skills in resolving conflict; (3) Violent Resistance (VR), that occurs when one partner uses violence to defend in response to abuse by the other partner (it may be a self-defense reaction or an overreaction); (4) Separation-Instigated Violence (SIV), that is, when either a man or a woman perpetrates violent behaviors as a reaction to the stress due to divorce in a relationship that has not otherwise been characterized by violence.

It is clear that in the first case (ACV)—in which men are usually the offenders and women are the victims (Kelly and Johnson, 2008)—it will be very difficult to conduct a child custody evaluation that can function as a preliminary clinical intervention able to treat intimate partner violence; in the other cases, instead, the child custody evaluation can be considered to be efficacious as a first intervention to assess and treat the partners' violent behavior.

The longstanding clinical experience of the authors of this contribution confirms the possibility of working with partners conjointly in the (2), (3), and (4) situations of violence as defined by Kelly and Johnson. These are situations in which the violence has the following characteristics:


As we shall see in the clinical case presented herein, the evaluation of the situation of violence takes place in the first assessment level with respect to the tolerance and exposure of each partner to violence. In the second assessment level, which has to do with the couple relationship, the forms of violence and the reciprocity of violent behaviors are investigated in order to verify the shared responsibility for the violence. In cases in which, from the first joint meeting with the couple, the clinician finds responsibility for the violent behavior in one partner only and the total victimization of the other, or in cases in which the impossibility of a reciprocal dialogic-interactive exchange between the partners is apparent in the first joint child custody evaluation encounter, the assessment levels will be carried out in individual, and not joint, encounters.

It must be pointed out that while there is agreement in literature on the need for an initial differential diagnosis of the type of violence occurring, to date, there do not appear to be specific instruments for such an evaluation in the child custody evaluation setting. Moreover, although many IPV screening tools have been developed over the years (Crane et al., 2017), most were evaluated only in a small number of studies (Rabin et al., 2009), so that it is up to the clinician's theoretical and methodological competency to evaluate the severity and dynamic of the couple's violent behavior.

# PROPOSAL OF A MODEL FOR WORK WITH VIOLENT COUPLES IN CHILD CUSTODY EVALUATIONS

In what way can a process of parenting assessment be considered a possible preliminary clinical intervention with the couple, in situations with IPV? To achieve this objective, we believe it is indispensable to develop a specific clinical work methodology defined as the relational-intergenerational approach to child custody evaluation (Cigoli and Scabini, 2006; Gennari et al., 2014; Ranieri et al., 2016). Its main characteristic is that it is aimed at understanding the exchange between familial generations and at searching for factors that safeguard and care for family relations. The consultant is thus called upon to perform actions that considers multiple aspects for the specific purpose of offering the judge the most complete report possible regarding the parents, their children, and their relations. A consulting framework is therefore necessary, one that traces the subjects' history, finds the meaning of events, and captures the characteristics of people and their relations so as to open up the possibility of a future organization that can safeguard the minor's development.

In this model of child custody evaluation there is an intrinsic connection between assessment and "family transformative potential," a connection that is also at the basis of a therapeutic assessment approach (Finn, 2007) and that, therefore, goes well beyond the production of a static snapshot of the participating subjects (Gennari and Tamanza, 2017). In fact, knowledge of the people and their relations is connected to the dynamic of the separation, as a transformation occurring in the way family is lived during the separation. Thus, it is not only a matter of capturing the marital and parenting dynamic, but also of assessing the capacity of the family configuration to evolve as it copes with the separation transition. Thus, the consultant cannot do without some sort of prognostic apparatus that is founded on a temporal perspective connecting past, present, and future.

Interest in the dynamic and process aspects requires a specific working methodology which, by utilizing specific skills and instruments, activates and moves the family so that, in addition to having evaluative information, it is possible to consider and activate the potential to transform and, thus, also care for the family relations. In short: without eliciting a change, to any degree and in any direction, it is impossible for the consultant to predict the family's possible evolution and to suggest to the judge what might be the family's potential in regard to the development of new forms of family relations to be defined with the separation.

The transformative potential in the relation is solicited from the parents' ongoing participation and reflection on the information coming to light during the clinical process, as well as from the possibility of ultimately agreeing with the parents on solutions to be presented to the judge, precisely due to the growing awareness of the parents themselves during the entire child custody evaluation process. It should be clarified that any soliciting of possible solutions to the parental conflict (included how to manage the violent behaviors) is, first of all, a diagnostic operation and not an obligation to transact or negotiate. In any case, it represents an opportunity: for the consultant to discern additional characteristics in the parents, while for the parents it is an opportunity to take into their own hands the parental function in its aspects of decision making and planning. From this perspective, the partners' resources are activated and valorized from the start, and directed toward behaviors and actions of constructive change.

Within this work model, assessment consists of a multidimensional investigation comprised of three different levels:


quality of individual intrapsychic functioning or the parents' personality characteristics, but from a systemic perspective of interrelation between the two parents, the child, and the relational systems of the family of origin. From this perspective, the evaluative criterion of parental adequacy will not be limited to considering the "care-giving capacities" of each parent, but will center attention, in an environment of reciprocal relation, on each parent's capacity and willingness to realize, maintain, and consolidate "parental unity" and, more broadly, family unity in order to safeguard the minor's family (New York Convention, 1989)<sup>1</sup> . With these specific characteristics, the child custody evaluation context makes it possible to offer an assessment of the parents and also of the violence in the parental couple while constituting a preliminary treatment phase of the parental conflict, including the aspect of violence that characterizes the relation.

In line with the specific aims of the present work, in the following paragraphs we will exclusively examine in depth the assessment of each partner and of the couple relation in order to explore treatability with respect to violence. Thus, we do not discuss all the assessment levels typical of the child custody evaluation.

# EVALUATING AND TREATING COUPLE VIOLENCE THROUGH THE CHILD CUSTODY EVALUATION: THE CASE OF JAMES AND MARY

Each partner signed the written informed consent form given by the clinician at the beginning of the child custody evaluation process for the processing of data for research and scientific dissemination purposes. The written informed consent form is prepared in accordance with the national law on the processing of personal sensitive information and privacy, pursuant to Article 10 of Law no. 675/96 and subsequent modifications. Since this is not a research project, ethics approval by an institution was not required as per university guidelines. Nevertheless, to preserve the confidentiality of the reported clinical case, the authors did not describe details which would make the couple recognizable.

# The First Assessment Level: Individual Characteristics of Each Partner

Literature underscores the importance of intrapersonal variables in the study and understanding of violent behavior. In particular, perpetrators of IPV often present some symptoms of psychopathology such as antisocial or borderline personality disorders (Chester and DeWall, 2018). Some psychological traits could also be considered as predictive factors/risk factors for IPV (Ulloa et al., 2016; Goodnight et al., 2017). For example, a lack or a fatigue in self-control and self-regulation is an important predictor of IPV perpetration (Finkel et al., 2009; Chester and DeWall, 2018). Furthermore, other personality characteristics (e.g., narcissism, neuroticism, etc.) were described as predictive of violence, both for perpetration and victimization although

Considering that IPV is a social process, is very important to assess the partners' interpersonal adjustment (i.e., skills, attitudes, and behaviors that an individual uses when entering a relationship with others). Indeed, lacks and difficulties in interpersonal adjustment between partners (e.g., dehumanization, infidelity, rejection, etc.) could be associated with a violent relationship (Chester and DeWall, 2018).

The history of the individual partners' families of origin is also important for understanding violence in the couple relation. Literature highlights that there is an intergenerational transmission of violence—the so-called "cycle of violence" (Widom, 2017). Specifically, childhood victimization and abuse or childhood exposure to domestic violence are predictors of IPV during adulthood both for perpetrators and victims (Ehrensaft et al., 2003; Dim and Elabor-Idemudia, 2017; Lieberz et al., 2017). In this connection, it thus becomes important to specifically capture the partners' representations of their history, the characteristics of their internalized parental models, and, more in general, their representations of the upbringing context as well as the internalized characteristics of the parental relations experienced by the individual partners as children (Delsol and Margolin, 2004; Skuja and Halford, 2004).

As shown in **Supplementary Figure 1** and **Supplementary Table 1**, James' individual assessment reveals a personality (Questionario di Adattamento Interpersonale— QAI; Di Nuovo, 1998) characterized by pathological impulsivity and narcissism and a problematic mood profile (Profile of Mood States—POMS; McNair et al., 1971; Farnè et al., 1991) with respect to five of the six states investigated (fatigue, tension, depression, vigor, and anger) while the sixth (confusion) is at the limit. It is interesting to observe the scores under the range of normality relative to anxiety and stress in social situations. Undoubtedly, the conflictual separation that James is experiencing influences his moods, yet these aspects have characterized James for a long time, as the story of his adolescence, as well as his marital history, reveal. James narrates that his childhood is characterized by economic well-being and considerable physical comfort; nevertheless, he talks about a constant lack of affection and care in family relations. From James' story, in fact, an almost total absence of significant relationships and relations between each family member emerges: both between the parents, who often betrayed each other even in the presence of the children, as well as between the siblings, who grew up without interactions and significant relationships. In the family context that James experienced, the needs of the family's components did not have space either to be expressed or addressed, except for the sake of appearances. This affective deficiency came to be structured in James in terms of unresolved needs and great anger and rancor toward his family of origin, still today at the basis of rancorous interactions with his family and a feeling of being entitled to compensation for what he feels he missed. The feeling of affective emptiness is well-expressed by

with some gender differences (Larson et al., 2015; Talbot et al., 2015; Ulloa et al., 2016). Finally, couple violence is connected to mood states. In particular, emotional instability could predict IPV (Talbot et al., 2015), and, at the same time, violence improves a deep emotional instability in the people involved (Beach et al., 2004; Pickover et al., 2017).

<sup>1</sup>https://www.ohchr.org/en/professionalinterest/pages/crc.aspx

high levels of narcissism as well as of impulsivity, which appears to characterize James' relational modalities in his relations with his peers, in his family, and in the couple relation. Moreover, a concern for the social dimension and appearances turns out to be excessive, notwithstanding that these very aspects were abundantly criticized by James in his narrative about his family of origin. His internalized models of marital relations show his father to be domineering, with no significant relation with his own family of origin nor with his wife's family of origin. The mother is described as a woman interested only in her own well-being and disinclined to establish sincere and honest relations with others. The models of marital relations experienced by James are characterized by the possibility of buying affection with money, by the right to use what is given against the other person to keep the other person him/her tied to oneself, in order to meet one's needs, even by force. James' aggressive and violent behaviors can thus be traced throughout his relational history.

In Mary's story of her childhood and adolescence, affective relational dimensions are present even if, as she herself says, there was aspects of painful emotional nature. One example is the experience of anorexia in adolescence and young adulthood and an obsessive-compulsive disorder that caused her to get out of bed repeatedly during the night to take a shower, consequences of repeated intrafamilial episodes of violence. Mary talks about a life experience characterized by a male chauvinist mentality in which the female figure is considered as an object to be relegated exclusively to the domestic domain, with no right to express her own needs and desires. Mary's story reveals an experience of precocious adultization (she had to take responsibility for managing the home, her father, and her older brother during her mother's prolonged absences caused by depression and joint problems of the back) and a context particularly lacking in emotional care, which led to the development of an unstable mood (still present in constructs of tension, fatigue, and confusion) and a high score for depression and lack of strength and vigor (POMS). From the point of view of personality traits, there is a diffuse difficulty in coping connected to a lack of assertiveness, that is, to inadequacy experienced in responding to the events and requirements of the context. Her indices for stress and anxiety in social situations are below the threshold of the normal range (QAI), indicating a inhibited and defensively intractable aspect. During the interviews with the consultant, Mary also struggles to stay in dialogue: she is more preoccupied with finishing what she has to say than establishing a dialogic exchange in keeping with the questions posed to her. The parental model experienced is characterized by an abuse of power on the part of the father with respect to the mother (who is forced by her husband to stop working and to meet all of his demands), extensive use of corporal punishment with the children when parental rules were transgressed, and, finally, the parents' legitimization of the brother's violent behavior toward Mary, the younger sibling and a female. Regarding her parents' life together as a couple, Mary says that she would not have wanted to have a married live like theirs, because she would not have been able to bear being an objectified, enslaved, and used woman like her mother. The family relations were characterized by the absence of relationships with the paternal relatives, while with the mother's family there were more interactions and contacts.

# The Second Assessment Level: The Nature and Characteristics of the Couple Relation

The second level of assessment aims to identify the characteristics and quality of a violent couple relationship, beyond the partners' individual differences. The theoretical model presented above assumes that the individual perspective is not sufficient for explaining the couple relationship and, in particular, violent behaviors within the couple. It considers the couple relationship as something unique and specific, as a third part with its own characteristics that must be investigated with specific instruments (Cigoli and Scabini, 2006; Cigoli et al., 2014). The couple relationship originates and is built upon the histories of individuals, according to each partner's needs, desires, and aims regarding the relation. Thus, the couple relationship assessment attempts to capture the specific modalities of establishing and being a couple, underlying those dynamics that are shared and involve both partners as they contribute to the construction of the same relation (Cigoli et al., 2014).

Literature underscores the importance of capturing the specific violence dynamic within the couple (Chester and DeWall, 2018).

The relation between James and Mary turns out to have been conflictual from the start; it originated as an opportunity to experience a different way of life from the one the partners had lived in their respective families of origin. Mary hoped to feel liberated and more valued as a person. James chose Mary because she was a simple girl, a "housewife," appreciating her meekness and sweet nature which compensated for the internalized representation of his mother as distant and emotionally unexpressive. The bond, as described by the exspouses, is continued until it takes on the characteristic of reciprocal dependency and fusion in which each partner devoted him/herself to meeting the other's needs and healing the pain each experienced in the family of origin. We thus can recognize a relation that originated with the aim of medicating, compensating, and making up for the deficiencies experienced in the partners' history as children.

The relational dynamic between the partners, characterized by control and dependency, gave rise over time to a progressive escalation of conflict and aggression that was accentuated by and eventually exploded with the arrival of children. Parental responsibilities, in fact, limited the time and possibility of attending to and indulging reciprocal needs and requests, and the promise of attention began to go unfulfilled. The fact that each partner had limited resources for analyzing the new family configuration in a realistic way, by scaling back personal requests, for example, brought this couple to a condition of stalemate and profound crisis. Each partner had become disillusioned in their initial expectations, and a feeling of real betrayal characterized the experiences of both spouses. Mary felt that she was not appreciated and sought out by her husband, no longer the center of his attention; James says that his wife's unhealthy and obsessive control had increased, and arguments characterized by violent verbal and physical attacks by both partners had intensified.

Mary decides to separate from her husband after the discovery that he had a stable extramarital relationship; nevertheless, the couple continues to live together for 3 years without reaching an agreement on a separation. In this period, the reciprocal violence, both physical and verbal, further increases despite the partners' repeated attempts to restore the peace. Both highlight the strength of the bond that unites them still today, the impossibility of accepting the end of the relation, and assert that they cannot tolerate that their partner could have a new relationship and that the children would have to live with someone else.

The feasibility of separation for this couple appears to be an impossible goal that is greatly complicated by each partner's unconscious actions to prevent the other one from leaving. The outcome is an even more unstable and violent relational situation. It is more than evident, still today, that the spouses are not able to separate despite the violent behavior and the reciprocal harm inflicted in 14 years of their relationship as a couple. In brief: James and Mary are living the impossibility to accept to lose the Other (spouse and children) because he/she has the function to care the one's own aspects of pain. For this reason partners are moved to force and obligate the Other, also through violent behavior, to answer one's own vulnerabilities and frailties.

# The Third Assessment Level: The Search for Resources for Abandoning Violent Relational Modalities and Achieving the Marital Separation

The information gathered with the prior assessment levels allows us to capture the specific meanings of the violence in James' and Mary's story.

Since childhood, both experience violence as a relational modality. Many aspects of their couple relation is also characterized by demands for salvific support which, not having been fully reciprocated since children arrival, fostered aggressive, and violent behaviors. Nevertheless, the partners are unaware of these aspects, even if they are present in their narratives; that is, they are not able to grasp to what extent some of their personal characteristics and relational modalities are dysfunctional and are thus the cause of reciprocal suffering and the failure of the relation. Each partner, in fact, is only focused on his or her own pain, on his or her own needs. The capacity for grasping the other's struggles is non-existent; the possibility of evaluating how much their own requests are unrealizable and idealistic is completely missing. The representation of the other appears limited to an instrumental function with regard to one's own needs. The literature clearly highlights the risk of exclusively egosyntonic relational modalities, both for relational failure as well as for situations of violence.

It is therefore the consultant's responsibility, in our view, to "be able to read" and to "help to read" the violence and failure of the relationship in their less conscious meanings as impossible and idealized requests, in addition to promoting a perspective on reading events that also contemplates the other, assigning value to him/her, considering his/her needs and fragilities. Sharing with each partner what emerges from the preliminary assessment levels is therefore the premise for being able to treat the violence. It is a matter of enhancing each partner's reflective functioning and mentalization capacity, which, we know, are often reduced in violent situations (Stover and Coates, 2016). In situations of violence found in separations, the process of elaboration is indispensable for a true emotional separation. Separation, in fact, allows the partners to grow because it introduces them to the process of individuation, allows them to experience possible modalities of independence from the other one, and leads to a complete psychological separation. In reciprocally violent couples, the difficulty of the process is even clearer because the primary processes of individuation and separation have never been accomplished (Gray, 2004).

The clinical work of assessment does not happen because of a spontaneous request on the part of the couple: their willingness, therefore, to reflect, rethink, and put into discussion their relational modalities and behaviors is quite limited. Nevertheless, there are some aspects that make this work possible. The first is precisely the thirst for justice with which each partner arrives in the consultant's setting: being able to demonstrate that one is right and finally finding someone to acknowledge one's claims with respect to the partner. These purposes predispose the spouses to narrate their personal stories and make their own case. This aspect, together with the need for a space where they could feel accepted and express their pain, made it possible in their case to open a space for elaboration.

The work with James involves reflection on his difficulty with "being alone" and the deep and longstanding conflict in his relations with his family of origin. All of this does not allow him to find his own stability. His family history, moreover, has determined the absence of an affective reference point that has prevented him from perceiving support: the relations that did not "nourish" from an emotional standpoint did not allow him to construct and make use of a relational model of reciprocal exchange: from this also derive the aspects of dependence (which manifest in the counterdependent modality) that lead him to establish relations in which he demands too much from his partner (but also from others) to be able to satisfy and fulfill his own affective needs. In the profile of the affectively dependent personality, one also finds mood swings that sometimes lead him to lose control of his own reactions toward people who oppose him. Work on the deep anger that he feels when encountering a relational frustration is revealed to be necessary if he is to manage relations with others more calmly.

The work with Mary addresses two aspects: her affective dependence and the difficulty of managing her anxiety and sense of incompetence and helplessness. Having grown up alone, without any protection or guidance, exposure to violence and a precocious adultization determined the need to constantly lean on someone in order to feel whole and worthy; indeed, it is precisely this need that does not allow Mary to end her couple relation and, in general, induces her to stay strongly anchored to relations, even dysfunctional ones. The second aspect, closely connected to the first, is the deep social and psychological aloneness in which she finds herself. This dimension of isolation, in fact, leads her to accentuate even more her physical, economic, and relational dependency.

Working through what emerged from the preliminary assessment phases proceeds through a necessary, if difficult, rereading of the couple relation that allows the partners to become aware of the needs and desires that were not met in the marital relation and the events that made it impracticable to move beyond them. In fact, if the pact of trust underpinning the partners' choice enters into crisis, and the affective theme that had given rise to the bond is not adequately addressed and worked through by the couple, an emotional blockage occurs. This blockage pervades the couple relation, amplifying, and exasperating aspects of conflict and violence and making the process of separation more difficult. Indeed, everything that remains unacknowledged regarding the relation's characteristics will inevitably be transferred and projected on to the parenting dimension, with the consequent and unavoidable involvement of the children. Only by working through the end of the couple bond, therefore, can the parenting dimension be relaunched (Gennari et al., 2014).

The work on James' and Mary's relation was made difficult by the fact that the partners' experience of mourning and loss has particular characteristics: in fact, it is a bereavement where the object is alive and often very present in the real lives of the partners (Losso, 2003). This amplifies conflictual feelings and affections activated by the separation event. Very briefly, the process of critical reflection on the part of the two spouses/parents, to be functionally useful to the process of elaborating the relation, must necessarily entail an assessment of the experience of the relation by means of two psychological actions. On the one hand, the partners are called upon to assume specific responsibilities. This entails feeling part of the history and events that led to the end of the relation that resulted from their own principal modalities of living, acting, and investing in the relation. The goal is to abandon a vision fixated on one's own pain and the lived experience of having suffered an injustice in order to reach a position of shared fragility and responsibility. On the other hand, one must be able to acknowledge the good received from the other and the relation (Cigoli and Scabini, 2006). In this way, in one's relational experience it becomes possible to acknowledge not only the other partner's debts toward oneself but also the credits as a way of offsetting losses and suffering. Thus, the separation relation can find closure with the perception of co-responsibility, and not only of failure.

This is primarily an ethical matter that has to do with the feeling of injustice invoked by both partners at the beginning of the child custody evaluation and makes it possible to put into perspective the experience of being wronged and one's demands for reparation from the other partner. If this does not happen, an unnecessarily persecutory, fragmentary, damaging, and disintegrative approach will prevail, and the conditions for an attack on the family's relations overall, and not only on the martial bond, will take shape. James' and Mary's story exemplifies this aspect: the escalation of violence is the outcome of the non-comprehension of the relation's failure; it is the cipher of the impossibility of accepting that the other has not been able to satisfy one's needs and requests since the beginning of the relation. In fact, themes of loss and quite significant narcissistic wounds emerge preventing the separation (Losso, 2003).

In the clinical work with James and Mary present together, the aim is to share the affective dependency that characterized both spouses, although in different ways. It also entails eliciting from each partner a request for support from the other, asking him/her for total dedication and care for the wounds experienced as children. Moreover, the possibility of putting into perspective the idealized representation that each one has constructed of the other's role is explored: with James, this involves putting into perspective his expectations of Mary as wife and mother so that he can revisit and possibly recuperate the relationship with his own mother, but also so that he can adopt a position of greater acceptance of Mary's characteristics and modalities. With Mary it is important to put into perspective the salvific omnipotence constantly required of James as husband and father.

An additional aspect that allows the couple to work in terms of reflection and personal development is each partner's desire to be able to protect and care for the children. It is precisely for the good of the children that the parents are able to move beyond their marital strife and become involved and motivated in the work of discussion and change with respect to themselves. The child is often of such immense value that it becomes possible to discuss even very solid and rigid relational positions and modalities: the child enables the parents to tune into an object other than their own pain; the child's struggles and needs are more legitimate than those of the other spouse and sometimes even than one's own. In fact, the child is the possibility of an opening onto a different future, one that is more positive, and this often motivates the parents to make efforts and pursue goals that are unthinkable if they are focused only on themselves. In short, the child can be an important engine for personal change. But children, as Lemaire (2002) emphasizes, constitute a living testimony of the other parent's presence and make it impossible to erase all traces of the relation, as the partners wish they could do. The children and their needs are thus the starting point that makes it possible to revisit one's representation of one's partner and his/her value as a parent (Cigoli and Scabini, 2006).

In this regard, discussion with James addresses his operative competence in the relation with his children, his capacity for containment and control, but also the absence of a more affective and supportive parental component. This explains the children's fear of him and their jealousy toward his companion, whom they see as receiving all their father's attention and affection. Moreover, James's difficulties tolerating and, as a result, valorizing at least partially his children's mother are discussed. This aspect also becomes crucial for interrupting the relational dynamic that characterizes the oldest child who, in order to protect his mother, has rigidified into a position of rejection and defiance toward his father.

Work with Mary focuses for a long time on how her unstable emotivity involves her children who participate excessively and directly in her malaise and fragility. Her dysfunctional involvement in the conflicts of her husband's family is also problematic since it foments the conflict with her children's father. As a parent, she needs to focus on relational and child-rearing modalities with her children that are less confusing and more evenhanded. Thus, space is made for the children to find in their mother an autonomous and solid parental figure, able to guide and protect them, interrupting Mary's current dynamic in which she is experienced and perceived by her children as a peer in need of protection.

It is precisely in the presence of both parents that it becomes possible, in a setting of cooperative work, to delineate the children's needs, needs that in the first place point to the importance of the joint exercise of parental functions (Emery, 2012). Indeed, some studies have emphasized that the negative effects of divorce as well as of violent behavior between partners on children's adjustment could be mediated by stable and cooperative co-parenting, which reduced the perceived parental distress (Molgora et al., 2014; Bernet et al., 2016).

# CONCLUSION

The aim of the present contribution was to present a clinical model of intervention with divorced couples experiencing IPV in the context of the child custody evaluation. Professionals often use traditional work modalities with no adjustment made for the presence of violence (Saunders, 2015). We believe, instead, that in these situations it is important to design a specific intervention that makes it possible, first of all, to evaluate what type of violence is present in the couple (Kelly and Johnson, 2008) and then to work not only in order to eliminate the violent behavior, but also to transform family relations in a more radical way.

In this sense, the model we have described, within a legal framework, could be described as a clinical intervention because it focuses not only on the evaluation of each partner ("the best parent"), but above all on the relationship between the partners. Indeed, bonds cannot be dissolved and erased, but only transformed (Cigoli and Scabini, 2006) and so, even if the partners are no longer a married couple, they will be a parental couple forever. The goal of this intervention is to make each parent an active protagonist able to collaborate, support, trust, and legitimize each other in their parental role. This process of acknowledgment and legitimization of the other is possible only by working to strengthen capacities for mentalization and reflective functioning (Aschieri et al., 2016; Stover and Coates, 2016). This increase the possibility that the partners will acknowledge their own part of the responsibility for the conflict and the violence, and thus initiate actions aimed at a constructive management of the conflict itself, finding the resources needed to care for the children. We know, in fact, that divorce, especially in its more conflictual and violent forms, often risks creating a family scenario in which suffering, reciprocal annihilation, and demands made between the spouses saturate every space, leaving in the background the children's developmental needs and requirements.

In this scenario, a crucial role is played by the dimension of time. In the clinical work, in fact, the consultant, in addition to considering the present, also focuses on the past (historical dimension) and the future. In particular, the future does not only have to do with possible family configurations after the divorce but also with the developmental trajectories of the children involved in the separation. Identifying spaces for change in the parent-child relations and the adoption of future scenarios centered on the children's needs and requirements becomes a primary objective in this intervention model.

This is what happened in the story of James and Mary where both, precisely due to the work of assessment described above, were able to move beyond reciprocally vindictive relational modalities centered on their own childhood needs to collaborative and more reflective modalities oriented principally toward responding to their children's needs. Such a problematic situation required that both partners move away from their reciprocal aggressive and violent relation and concentrate together on the need to help and support their children. From this perspective, it became possible for James to spontaneously undertake a path of personal psychotherapy, and, at the end of the child custody evaluation, both parents agreed to be involved in a joint program of parenting support. This outcome well represents the act of justice that both partners must institute and pursue to give a new reason for hope and redemption to their family relations.

Despite the innovative focus of this method, it presents some limitation. In particular, the work method presented reveals itself to be useful in those specific situations of violence that erupt from the post-separation event and which, therefore, are not prevalent in the couple relationship. The method also showed a certain efficacy in situations of reciprocal violence, even if in different forms and intensities between the partners, as long as these are not extreme situations that put the partners' lives at risk. This assessment proposal has not been sufficiently tested in cases of extreme forms of violence or in situations in which one partner is clearly dominated and victimized by the other. Therefore, in these situations, the proposed assessment method cannot be considered reliable and effective.

# ETHICS STATEMENT

Ethics approval for this research was not required as per the Università Cattolica del Sacro Cuore's guidelines and Italian laws and regulations. Written informed consent was obtained from all research participants and form the parents/legal guardians of all children. The written informed consent form is prepared according to the national law on the processing of personal sensitive data and privacy, pursuant to Article 10 of Law no. 675/96 and subsequent modifications. This was given to the parents by the clinician at the beginning of the child custody evaluation process, for the collection and processing of data for research and scientific dissemination purposes.

# AUTHOR CONTRIBUTIONS

MG contributed to the writing of the manuscript's clinical case paragraphs. GT supervised the entire manuscript. SM contributed to the writing of the manuscript's first five paragraphs. All authors reviewed and approved manuscript for publication.

# SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyg. 2018.01471/full#supplementary-material

# REFERENCES


**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 © 2018 Gennari, Tamanza and Molgora. 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.