Tracey E. Rizzuto1*
Ague Mae Manongsong1
Ally St Aubin2Shinhee Jeong3Christiane Spitzmueller4
Juan M. Madera2- 1School of Leadership and Human Resource Development, Louisiana State University, Baton Rouge, LA, United States
- 2Conrad N. Hilton College of Global Hospitality Leadership, University of Houston, Houston, TX, United States
- 3Department of Human Resource Development, University of Texas at Tyler, Tyler, TX, United States
- 4Office of the Vice Chancellor for Academic Affairs, University of California Merced, Merced, CA, United States
Introduction: This two-part multimethod study develops and validates a Faculty Leadership Linguistic Inquiry Word Count (FL-LIWC) dictionary for use in university settings.
Methodology: In Study 1, an integrative literature review centered on Yukl’s (2012) hierarchical taxonomy of leadership behavior and Delphi panels are used to establish a four-component definition of Faculty Leadership and to generate a 354-word LIWC dictionary reflecting leadership behaviors performed by university faculty. Study 2, exploratory, confirmatory, and structural path models using textual data extracted from external review letters (ERLs) from 981 promotion and tenure cases across five universities is employed to conduct a multi-step validation analysis of the dictionary’s accuracy, nomological network and potential for gender difference.
Results: This study demonstrates that FL-LIWC is a theoretically coherent and empirically robust tool for measuring faculty leadership and is relatively free of potential biases associated with gender difference.
Discussion: It provides greater understanding of how linguistic reflections in ERL text may be used in university setting to provide additional insight into leadership behaviors that are of emerging importance to faculty performance assessment and academic advancement decisions.
1 Introduction
United States (US) institutions of higher education have been the target of intense criticism in recent years, from the reported loss of confidence in university leadership (Camera, 2023) to questions of performance evaluation bias in longstanding traditions of promotion and tenure (P&T) review processes (Nietzal, 2024). Connections among criticisms are central to this paper which responds to the need for multi-institutional research on faculty leadership and performance evaluations (Jamali et al., 2022) and investigates whether external review letters (ERLs) that are commonly used in the P&T processes may contribute useful insight and new standardized approaches to evaluating an important yet ambiguous, subjective, and difficult to measure construct: faculty leadership.
This two-part multimethod study establishes and validates a Faculty Leadership Linguistic Inquiry Word Count (FL-LIWC) dictionary for use in university settings. In Study 1, an integrative literature review centered on Yukl (2012) hierarchical taxonomy of leadership behavior and Delphi panels are used to (a) establish a four-component definition of Faculty Leadership and (b) generate a 354-word LIWC dictionary that reflects leadership behaviors performed by university faculty. In Study 2, textual data extracted from the external review letters (ERLs) of 981 promotion and tenure cases across five universities is employed to analyze exploratory, confirmatory, and structural path models in a multi-step validation of FL-LIWC dictionary’s accuracy and nomological network. It further explores its potential for mitigating known sources of gender bias that may affect P&T decision making processes (Masters-Waage et al., 2024). Findings from this study contribute a novel standardized approach to the systematic measure of faculty leadership and offers guidance for how institutions may benefit from using the FL-LIWC dictionary for leadership assessment and development practices.
1.1 Faculty leadership and its emerging importance in academia
The higher education literature provides ample evidence that effective and capable leaders are critical to university productivity, morale, and progress (Scott et al., 2008). Yet, the concept of leadership within the academic context (i.e., faculty leadership) is relatively poorly understood, often amorphously defined and hinged upon research that lacks rigorous, transparent methodologies (Gibbs et al., 2008; Gonaim, 2016; Nguyen et al., 2020). Despite the lack of construct clarity, there is a growing consensus that leadership in universities is in crisis (Camera, 2023). Gonaim (2016) describes a global deficiency in the preparation of faculty leaders and inadequate attention devoted to their development based on the belief that leadership roles are temporary (Fullan and Scott, 2009). While much of the research on faculty leadership focuses on formal administrative roles conducted by senior faculty (Jing and Avery, 2008; Rowe et al., 2005), a deeper appreciation of what it means to lead in the academy is taking hold that transcends traditional bureaucratic models and formal titles (Fletcher, 2004): “New models of leadership recognize that effectiveness in knowledge based environments depend less on the heroic actions of a few individuals at the top and more on collaborative leadership practices distributed throughout an organization” (pp. 647).
The growing demand for more effective leadership and newer conceptualizations of distributed leadership practice triangulates on faculty leadership as an emerging performance concern for universities (Hempsall, 2014; Kezar and Lester, 2009; Kezar et al., 2011). But how faculty leadership is defined, how it is assessed, and its influence on faculty performance evaluations and P&T review processes is not well understood. These questions are the focus of the present inquiry and the motivation for developing and evaluating the FL-LIWC. This tool provides a theoretically-grounded conceptual definition of university faculty leadership, an operational measure that can be rigorously assessed, and a standardized approach to using institutional textual data to empirically reflect leadership behavior in decision making processes (e.g., leadership development, performance assessment, and career advancement).
1.2 External review letters in the promotion and tenure review process
Promotion and tenure (P&T) review is a critical step in the career advancement of faculty. It often determines a scholar’s ability to continue one’s career track, be forced to relocate to another institution, or pursue a different career path entirely (Berenbaum, 2019; Schimanski and Alperin, 2018). The bulk of attention to faculty career advancement and P&T review focuses on research productivity and teaching evaluations (e.g., Collins and Liang, 2015; Gottlieb et al., 2022). Research productivity, often measured in terms of research output (e.g., number of publications, citations, grant dollars), is one of the most heavily weighed factors in faculty performance evaluation at research-intensive institutions (Collins, 2015). It is typically assessed alongside other important performance factors like teaching evaluations, mentorship, and leadership service (National Academies of Sciences, Engineering, and Medicine (NASEM), 2023). This body of performance is contextualized and evaluated by field experts from outside institutions through ERLs, which provide an independent and holistic evaluation of a candidate’s scholarly impact and contributions. Despite these steps to make P&T review objective and meritocratic, research suggests that irrelevant demographic factors, such as gender, affect ERLs and validity in P&T decision outcomes.
For example, recent research by Masters-Waage et al. (2024) examined over 1,500 P&T decisions across five US universities and revealed a double standard in how scholarly productivity is regarded. They found that underrepresented minority (URM) faculty with below average h-indexes were judged more harshly than non-URM faculty with similarly low h-indexes. This discouraging finding is consistent with prior research that points to race and gender biases in faculty lived experiences (Turner et al., 2011). However, what is new and encouraging about this study’s conclusion is that differential treatment of URM women was mitigated through language reflected in ERLs. Specifically, the more external reviewers highlighted candidates’ scholarship in their letters, the less the double standard penalty affected URM women. This finding suggests that ERL language shapes how faculty performance is contextualized and may indicate the need for new approaches to performance evaluation that mitigate known biases and improve the accuracy of P&T review processes.
Although ERLs have a long tradition in P&T review processes, the linguistic content, value, and influence on performance evaluation is gaining attention due to new analytic software that allows for systematic content coding of ERLs. Linguistic inquiry word count (LIWC, Chung and Pennebaker, 2012) is a computer analytic framework that quantifies psychological constructs embedded in textual data. It classifies words into multiple linguistic/psychological categories based on the usage of the words referenced in a dictionary file, and then calculates the proportion of words in each category relative to the total number of words in the entire text. LIWC dictionaries define target words in reference to specific interest categories (e.g., positive emotions, social concerns). These linguistic categories can be used for simple sentiment analysis to study affective states (Tausczik and Pennebaker, 2010), but more recently it is gaining traction as a tool for inferring more complex psychological sentiments. Jamali et al. (2022) suggest that more rigorous methodological assessments of faculty leadership and performance could enhance communication and consistency in the cultivation of leadership capacities. The FL-LIWC dictionary offers a novel standardized framework to analyze existing ERL documentation in a manner that derives new evaluative insight and utility that may serve to enhance accuracy in P&T review processes. The following two studies hypothesize and psychometrically assess FL-LIWC content and construct validity as well as explore the potential for gender difference in its criterion measures and associations with faculty performance.
2 Study 1: creating a faculty leadership definition and LIWC dictionary
The purpose of Study 1 is to establish an evidence-based theoretical definition and target word list that can be used to operationalize a LIWC dictionary that measures leadership behaviors performed by university faculty.
2.1 Theoretical underpinnings of faculty leadership in academia
Early conceptualizations of organizational leadership point to the transactional and transformational nature of leadership behavior (e.g., Bass, 1985; Burns, 1978), while later theories emphasize the importance of balancing multiple stakeholders’ interests (Parmar et al., 2022). More recent literatures focus on behaviors performed by effective leaders and seek to uncover behavioral styles (Benmira and Agboola, 2021) which lies in contrast to classic leader trait theories that articulated common attributes among great leaders of their time (Carlyle, 1907; Galton, 1869). From business to education, leadership is generally understood as the ability to inspire, guide and influence others, but the specific actions that define leadership within a given context vary in their application and impact (Black, 2015). For instance, corporate leadership may prioritize maximized profits, competitive innovation, and financial growth (Alateeg and Alhammadi, 2024; Orieno et al., 2024), while faculty leadership often emphasizes the fostering of learning culture and behaviors that target collaboration, mentorship, and human resource development (Kezar et al., 2011; Wieczorek and Lear, 2018). These differences substantiate the need for context-specific understandings of leadership (Oc, 2018).
Current research on leadership behaviors performed by faculty in university settings range in topics from administrative onboarding, leadership training quality, to advancement barriers (e.g., Kezar and Lester, 2009; Pate and Angell, 2013; Tsoh et al., 2019). These studies identify an array of desired traits, behaviors, attitudes, and abilities for faculty leaders (e.g., Gonaim, 2016; Johnson, 2023), but a parsimonious definition of what university faculty leadership entails has yet to emerge (Black, 2015). Critics contend that a unifying definition would provide a foundation for standardized evaluation criteria and tools to develop the skills needed to balance administrative, teaching, and research duties (Black, 2015; Rosari, 2019).
Yukl et al. (2002) hierarchical taxonomy of effective leadership is one of the most widely researched cohesive models for comparing leadership effectiveness across settings (Borgmann et al., 2016; Yukl and Mahsud, 2010). It has been applied to topics related to strategic leadership, flexibility, creativity, and the capacity for change (Drew, 2010), but to date has received little attention in higher education except for a small sample of studies that investigate faculty development challenges (Buchen, 1998; Lawler and King, 2000; Kezar and Lester, 2009). Yukl (2012) framework is comprised of four meta-categories (task-oriented, relations-oriented, change-oriented, and external behaviors). These meta-categories provide a comprehensive understanding of how leadership behaviors drive transactional effectiveness through task performance, strategic thinking applied to change management, and transformation through interpersonal influence and by maintaining diverse stakeholder relationship.
2.2 Study 1 material and method: integrative literature review
Study 1 adopts a two-step qualitative approach to address our research aims. First, an integrative literature review methodology (Torraco, 2005) is employed to synthesize literature critiques and conceptualize understandings of university faculty leadership. We center our literature search and review on Gary Yukl (2012) taxonomy of effective leadership behavior. We adopt this framework as the organizing theory to categorize faculty leadership descriptions and definitions published in peer-reviewed literatures across multiple disciplines to determine the extent of alignment with the theorized categories in Yukl (2012) framework. Second (section 2.4), we outline procedures for the development and validation of the FL-LIWC dictionary, which employs a Delphi panel process (Dalkey and Helmer, 1963) of subject matter experts to evaluate the content validity of the faculty leadership definition and the derived FL-LIWC target word list.
2.2.1 Search process
Article selection for the integrative literature review took place in multiple phases and involved four graduate assistants. Articles published between 1986 and 2024 were systematically identified using Google Scholar and a centralized university search engine that allowed us to scan several research databases simultaneously: ERIC, APA PsycInfo, Human Resources Abstracts, Louisiana State University Libraries Discovery (Academic Search Complete and All Providers). A combination of Boolean operators and key phrases were queried across all article fields in each database including: “faculty leadership in higher education,” “academic faculty leadership,” “leadership traits in academic affairs,” “leadership and higher education,” “effective leadership behaviors in academia,” “leadership in higher education,” “faculty leadership,” “leadership competencies OR characteristics,” and “leadership in academia,” This procedure identified 61 articles to be considered for inclusion in the study.
2.2.2 Inclusion and exclusion criteria
Next, abstracts from the 61 articles were screened according to the following criteria: (1) empirical studies (qualitative, quantitative, or mixed method) with designated methodology sections, (2) based in the United States (US), and (3) study samples focused on higher education faculty (full-time, tenured, tenure-track). We targeted empirical and conceptual works that adhered to the standards of the peer review process, ensuring methodological rigor and sound theory-making. Newspaper articles, book or article reviews, and conference presentations were excluded because we could not ensure these pieces underwent the same scrutiny as academic peer-reviewed journals or dissertations. All abstracts were screened for relevance. Full article texts were reviewed if they met the inclusion criteria. In total, 19 articles were retained for the final analysis: 10 qualitative, 5 quantitative, 3 literature reviews, and 1 conceptual paper with the majority published within the past 10 years (2011–2024), with three articles published between 2000 and 2010 and two articles published between 1986 and 1999. All articles included in the integrative literature review analysis are listed in Appendix 1.
2.2.3 Taxonomic analysis and annotated matrices
Following established integrative literature review procedures (e.g., Estes and Jia Wang, 2008; Torraco, 2005), data from the selected articles were organized in a Microsoft Excel matrix that contained research title, author(s), journal, sample, methodology, theoretical framework, and derived leadership behaviors. Six coders were trained to ensure a shared understanding of the leadership framework. The articles were independently coded, categorizing behavioral alignment with Yukl (2012) taxonomy. An additional “other” category was created for behaviors that did not align. All data were organized into a matrix with leadership descriptors from each article according to their taxonomic classification. Coding allocations were collectively discussed until consensus was met regarding the categorization with Yukl (2012) taxonomy.
2.3 Taxonomic findings and discussion
This integrative literature review drew from a variety of theoretical frameworks and research methodologies, including comparative studies that extended to samples in United Kingdom and Australia, highlighting the global relevance of faculty leadership research. Our analysis demonstrates literature coherence with Yukl (2012) four-component (15 subcomponent) framework, providing qualitative secondary literature support for the content validity of the Faculty Leadership construct. Through this analytic process, 275 leadership behavior terms and descriptors were identified in relation to the Faculty Leadership construct.
Task-oriented Behavior (Task Leadership) pertains to actions engaged to initiate clear structures, expectations, and performance standards that are required to fulfill a transactional exchange between leader and follower. This component featured prominently in the literature and was associated with 42 descriptors. Specific subcomponent behaviors include planning, which had 11 descriptors (e.g., organizing, coordinating), clarifying with 12 descriptors (e.g., communicative, transparent), and monitoring (e.g., adjusting, assessing) and problem solving (e.g., mediating, decision making), each with 10 descriptors. These leadership descriptors underscore the importance of well-organized and productive work environments within academia (e.g., Gonaim, 2016; Keith and Buckley, 2011; Zacher and Johnson, 2015).
Relations-oriented Behavior (Relational Leadership) is the expression of caring and consideration toward followers’ needs, openness to input from others, and friendly, approachable, and equitable treatment toward others. Specific subcomponents include supporting behaviors, which was associated with 24 descriptors (e.g., considerate, caring), developing behaviors with 9 descriptors (e.g., providing feedback, mentoring), recognizing behaviors with 3 descriptors (e.g., rewarding, appreciating), and empowering behaviors with 8 descriptors (e.g., participative, authorizing). Combined, these subcomponents provide the foundation for positive workplace relationships, supportive work environments, and encouragement for both personal and professional growth (Bryman, 2007; Gonaim, 2016; Yukl, 2012).
Change-oriented Behavior (Change Leadership) emphasizes a leaders’ attempt to initiate and encourage change. Defining subcomponent actions include envisioning change with 10 descriptors (e.g., visionary acumen, foresight), encouraging innovation with 11 descriptors (e.g., challenging decisions, seeking different perspectives), and advocating change with 2 descriptors (e.g., confronting, speaking up for) and facilitating collective learning with 11 descriptors (e.g., creating a sense of community, team building). Collectively, leaders who display these subcomponent behaviors are better equipped to manage organizational changes and foster a culture of continuous improvement and resilience (Fernandez and Shaw, 2020).
External Behavior (External Leadership) facilitates a follower’s performance by providing relevant information about outside events, getting necessary resources and assistance, and promoting the reputation and interests of the work unit. Three distinct subcomponent behaviors include networking with 8 descriptors (e.g., making connections, partnering), external monitoring with 3 descriptors (e.g., being situationally aware, adjusting tactics), and representing with 6 descriptors (e.g., enhancing reputation, lobbying).
Through our analysis, we discovered several leadership qualities that did not fit neatly within Yukl (2012) hierarchical taxonomy. A total of 34 descriptors emerged and were coded as Non-Categorized Leader Behaviors. Among them were character traits associated with ethical leadership (e.g., fairness, honesty, transparency) and expressions of leadership style (e.g., serenity, modesty, tenacity). These descriptors highlight the complex interplay between leadership behavior and leader attributes, point to a preference for leadership that emphasizes authenticity and transparency (Johnson, 2023) and demonstrates high moral character (Agarwal and Bhal, 2020), and provide insight into how higher education values and implements leadership. Further research is needed to delineate and synthesize these various non-categorized descriptors.
2.4 Faculty leadership (FL-LIWC) dictionary development
Building upon the 275 leadership behavioral descriptors, we sought further content validity evidence of the Faculty Leadership meta-component domains, and to expand the word list of terms describing leadership behaviors performed by university faculty to ensure word count balance across each (approximately 23 words per subcomponent). To do this, a Delphi panel comprised of five faculty leadership and career advancement subject matter experts (SMEs) from four universities were emailed definitions of the four Faculty Leadership meta-components1 (task, relation, change and external leadership) and 15 subcomponents. The panel were asked to (a) assess the framework’s content sufficiency and (2) generate terms and phrases associated with the meta-components and subcomponents. These terms were cross referenced with the original 275 leadership descriptors--retaining words common to both generation processes, adding new words endorsed by SMEs, omitting words that were not orthogonal or were flagged by SMEs as irrelevant to the faculty leadership content domain. The resulting target word list was re-reviewed by the panel to ensure the word list’s content adequacy. To further strengthen content validity and the practical utility of FL-LIWC word list dictionary (Rezvanifar et al., 2022), a second Delphi panel comprised of university administrators and diversity scholars were asked to forecast interpretive user reactions (i.e., stigma, bias, clarity) to the target word list, followed by re-review. The final product resulted in a FL-LIWC dictionary containing 354 target words (see sample words in Supplementary Appendix 2).
3 Study 2: validating the faculty leadership (FL-LIWC) dictionary
The goal of Study 2 is to examine aspects of construct and criterion-related validity of FL-LIWC and test hypotheses related to its functioning. Construct validity refers to the extent to which a measurement accurately assesses the construct it is intended to measure (Campbell and Fiske, 1959). Supplementing the content validation steps described in Section 2.4, an exploratory factor analysis is conducted in Study 2 to assess the FL-LIWC component structure. Next, using a separate data set, measurement models of two Faculty Leadership components structures are compared to confirm the factor structure and assess FL-LIWC’s empirical alignment with the construct’s theory. Construct validity, the appropriateness of inferences made based on a measure (Cronbach and Meehl, 1955), explores whether FL-LIWC produces expected empirical relationships with other factors. Structural path models test the hypothesized nomological network of expected relationships to assess measurement convergence (convergent validity). Discriminant validity is inferred by examining theoretically unrelated (divergent) factors produced by the same measurement method (ERL-based LIWC estimates) as well as the average variance extracted (AVE; Kock, 2019).
3.1 Faculty leadership factor structure and nomological network
Task Leadership pertains to actions engaged to initiate clear structures, expectations, and performance standards that are communicated from leader to follower (i.e., junior faculty member, graduate assistant, student). Conway (1999) conceptualizes leadership task performance as a hybrid of task and contextual job performance, such as the ability to problem-solve and clarify job roles important for effective transactional leadership (Bass, 1985). Effective task leadership for university faculty is often demonstrated across a range of academic products, such as funded grants, high-quality publications, well-balanced service duties and evidence of teaching effectiveness. We hypothesize that linguistic reflections of academic productivity will be positively associated with FL-LIWC.
The second definitional domain, Relational Leadership, is the expression of caring and consideration toward followers’ needs, openness to and support for others. For university faculty, the cultivation of high-quality interpersonal relationships (Branson et al., 2016) and the effective mentoring of students and faculty peers (Reynolds et al., 2023) are thought to be associated faculty relational leadership. Therefore, we anticipate that ERL mentorship language would be positively associated with FL-LIWC.
Change Leadership pertains to one’s ability to inspire and encourage innovative thought, facilitate action and advocate for change and is most closely affiliated with transformation leadership (Bass, 1985). Disrupting the status quo can be necessary for institutional growth and improvement, but it can also be controversial. It can generate negative reactions and catalyze resistance among those invested in maintaining existing academic policies, practices, and norms (Anttonen and Chaskes, 2002). For this reason, we anticipate significant associations between FL-LIWC and language that characterizes faculty as having strong political drive and affiliations.
Lastly, External Leadership is the act of facilitating others’ achievements by listening to multiple stakeholder interests (Parmar et al., 2022) and promoting the reputation and interests of the professional entities they represent. University faculty to contribute their expert knowledge and skills beyond the walls of the academy to benefit local communities, professional associations, and society at large. This type of performance is often termed, “service.” We anticipate that linguistic ERL reflections of service will be positively associated with FL-LIWC.
Hypothesis 1: FL-LIWC will (a) demonstrate convergence with theoretically related ERL measures of productivity, mentorship, politics, and service, and (b) structurally discriminant from a theoretically unrelated ERL LIWC measure (i.e., present focused language).
3.2 Gender difference in faculty leadership and FL-LIWC measures
Meta-analytic findings presented by Koenig et al. (2011) suggest that people generally view leaders as masculine. Eagly (1987) maintains that gendered views of leadership may be explained through the mechanisms of social role congruity theory which suggests that gender stereotypes create different behavioral expectations for men and women in leadership roles. She explains that society assigns masculine traits, like assertiveness and decisiveness, to leadership roles, thus creating role incongruity for women leaders who are stereotypically associated with traits like empathy and nurturing; trait that are often seen as incongruent with leadership roles (Eagly and Karau, 2002; Williams and Dempsey, 2014).
Academic environments have been criticized for systemic gendered structures that stall careers (glass ceiling) and lead to academic departures (leaky pipelines; Acker, 2012; Britton, 2017; Ray, 2019) and an underrepresentation of women and minoritized faculty in academia (Eagly and Karau, 2002; Hannum et al., 2015; Hoyt and Murphy, 2016). Gendered expectations of leadership may inject biases that perpetuate obstacles to faculty career advancement for women (Eagly and Karau, 2002). These obstacles may persist even when women faculty adopt traditionally masculine behaviors. For example, Makarova et al. (2019) found that women who adopt traditionally masculine assertive and authoritative agentic qualities may be penalized for acting inconsistent with their gender roles (Makarova et al., 2019). Drawing on this bed of empirical literature and role congruity theory (Eagly and Karau, 2002), the discord between societal expectations of gender roles and the demonstration of faculty leadership behavior may differentially affect men and women faculty and result in gender difference in FL-LIWC scores.
However, measurement error from biases within any single measure can be minimized when it is used in combination with other measures from multiple sources. Faculty Leadership, operationally measured using FL-LIWC, is only one aspect of faculty performance, among oa multitude of other performance and scholarly impact measures taken into account in P&T reviews. Possible measurement error associated with gender difference in FL-LIWC is likely to be mitigated by other faculty performance criteria, productivity being one of the most heavily weighed factors in faculty performance evaluation at research-intensive institutions (Collins, 2015).
Hypothesis 2: (a) Women faculty will have less faculty leadership language (lower FL-LIWC) expressed in ERLs than men faculty. (b) Gender difference in FL-LIWC will be mitigated by ERL measures of faculty performance (productivity, politics, and service).
3.3 Study 2 materials and method
3.3.1 Participants
This study employed cross-sectional data from 5 research intensive (R1) universities over a 7-year period (2015–2022) containing 981 promotion and tenure cases for faculty members seeking promotion to associate professor (with tenure) or full professor. This dataset is supplemented with linguistic analyses of external review letters (ERLs) written for P&T candidates and publicly available data scraped from the internet (e.g., h index). These data were randomly divided into two demographically comparable subsamples which chi-square analyses determined to be statistically similar, ruling out sample representation differences between the two subsamples (Table 1). Subsample 1 (N = 317 P&T candidates) was used to conduct an exploratory factory analysis (EFA) and Subsample 2 (N = 664 P&T candidates) was applied to confirmatory faculty analyses and structural models used to test the study’s hypotheses.
Table 1. Sample and subsample comparisons.
3.3.2 Procedures and measures
All LIWC dictionary variables were constructed using LIWC-22 generated calculations of the proportion of the words in each dictionary category to the total number of words articulated in ERLs provided for each P&T faculty candidate. Scores were averaged for each category per candidate, and then transformed into standardized z-scores for analysis. Table 2 displays all variable descriptives and correlations.
Table 2. Variable descriptives and correlations (N = 981).
Following the Delphi panel process described in Section 2.4 for the generation of target list terms for the creation of a LIWC dictionary, three additional new variables were created to examine the construct validity of the FL-LIWC measure and linear regression estimates of these additional criterion-related factors. The three ERL-based LIWC variables created for this study are as follows: Productivity reflects commonly performed behavioral functions that are often associated with faculty research productivity (e.g., publish. Edit, research). Mentorship reflects commonly performed mentorship, student support and teaching behaviors (e.g., tutor, sponsor, guide). Service is a variable that reflects common forms of faculty service duties and assignments within universities, communities, and professional associations (e.g., advise, administrate, recommend).
The last two variables incorporated in this study are pre-established LIWC dictionary variables that are embedded within the LIWC-22 software (Boyd et al., 2022). Politics is a broadly constructed variable that includes words often used in political or legal discourse (e.g., policy, govern, preside). Present Focus Language reflects the use of present-tense grammar in ERL text (e.g., is, are, I’m, can). The prevalence of Present Focus Language is thought to be theoretically unrelated to the Faculty Leadership construct and FL-LIWC measure and therefore is introduced into this study design to establish discriminant validity (see Supplementary Appendix 2 for examples of target words included in each ERL-based LIWC dictionary variable).
3.4 Results and discussion
3.4.1 Exploratory factor analysis
Using Subsample 1, an exploratory factor analysis (EFA) was implemented to evaluate the FL-LIWC component structure. As Watkins (2018) suggest, it is necessary to verify that the measured variables are sufficiently intercorrelated to justify factor analysis. Bartlett’s test of sphericity yields a statistically significant chi-square (X2 = 13.187, d.f. = 6, p = 0.042) and the Kaiser-Meyer-Olkin (KMO) valued is within the prescribed range between 0 and 1, both meeting requisite standards to justify the EFA. A principal components extraction with varimax rotation was performed to correct for the likelihood of shared variance between correlated factors. Based on our evaluation of the scree plot and eigenvalue under one rule, a two-component structure emerged (55.6% variance extracted) producing the respective factor loadings. External Leadership (−0.585) and Change Leadership (0.836) loaded on Component 1, and Relational Leadership (0.532) and Task Leadership (0.859) loaded on Component 2. Loading averages exceeded the desired standard 0.70 threshold and all were well above the recommended minimum threshold of 0.40 (Hinkin, 1998; see Supplementary Appendix 3).
3.4.2 Confirmatory factor analysis
Further examination of the FL-LIWC factor structure and construct validity were analyzed using Subsample 2 in a series of covariance-based SEM (AMOS) models. Standard fit indices were applied to evaluate the goodness of fit of the confirmatory factor analysis (CFA): comparative fit index (CFI), Root Mean Square Error of Approximation (RMSEA), and the chi-square test. Model fit is deemed acceptable if CFI values are greater than 0.90 and RMSEA is smaller than 0.50 (Nye, 2023). The maximum likelihood method was applied, which requires the data to be normally distributed. As shown in the Supplementary Appendix 3, the data were normally distributed with the absolute values of skewness below 2 and the absolute values of kurtosis less than 5 (Ghiselli et al., 1981).
A confirmatory factor analysis (CFA) of the FL-LIWC component structure was conducted employing Chen et al. (2006) guidelines for distinguishing bifactor constructs from second-order hierarchical constructs. First, we tested a bifactor structural equation model to confirm the empirically derived EFA findings. As shown in Appendix 4, CFA Model 1a (CFA Model 1a, X2 = 3.124; d.f. = 1; CFI = 0.932; RMSEA = 0.057) contains two first-order factors, both allowed to freely covary with covariances constrained to 1.0. We compared this bifactor model fit with CFA Model 1b (X2 = 2.637; d.f. = 2; CFI = 0.980; RMSEA = 0.022), a hierarchical second-order construct model where FL-LIWC serves as a second-order latent variable that has paths connecting to the two first-order factors described in Model 1a. The structure of Model 1b aligns with the unidimensional conceptualization of effective leadership behavior theorized by Yukl (2012). As reflected in Appendix 4, Model 1a and Model 1b exhibit comparably acceptable fit and the chi-square difference between the two nested models reveal no significant difference. In deference to theory, we adopt a unidimensional model of Faculty Leadership which suggests that the FL-LIWC dictionary can be conceived of as a single global construct that linguistically reflects the core defining behaviors of university faculty leadership.
Next, in the model series (Model 2: Construct Validity), a first-order structural model is used to estimate FL-LIWC’s associations with theoretically related (convergent) and unrelated (divergent) nomological network factors. Convergent validity estimates the degree to which measures of the same or related variables fall on the same factor, while discriminant validity represents the degree to which divergent factors differ from other latent variables in the network that could account for more variance (Campbell and Fiske, 1959). Although there are no strict psychometric thresholds to establish construct validity, correlations with related measures reflect convergent validity, while relationships to other measures should be low to establish discriminant validity (Lezak et al., 2012).
3.4.3 Construct analysis of nomological network
In Model 2a, indicators freely correlate with first-order factors corresponding to FL-LIWC, which has correlational linkages drawn to four theoretically related aspects of faculty performance: productivity, mentorship, politics, and service. Model 2b is structurally similar but it is a congeneric model that includes Present Focus Language, a discriminant factor that is hypothesized to be theoretically unrelated to FL-LIWC. As expected, congeneric Model 2b (Model 2b X2 = 72.154; d.f. = 21; CFI = 0.906; RMSEA = 0.061) exhibited slightly weaker fit than Model 2a and produced a non-significant regression coefficient for Present Focus Language (β = −0.019, p = 0.699). In addition, all expected convergent relationships in Model 2a (X2 = 44.253; d.f. = 16; CFI = 0.944; RMSEA = 0.052) yield significant regression coefficients when regressed on FL-LIWC: Productivity (β = 0.75, p < 0.001), Mentorship (β = 0.60, p < 0.001), Politics (β = −0.29, p < 0.001), and Service (β = 0.64, p < 0.001). All structural models are summarized in Appendix 4.
These models provide some support for convergent validity and initial (but not sufficient) evidence of discriminant validity. To strengthen support for discriminant validity, the average variance extracted (AVE) is prescribed as a reliable for assessing construct divergence in factor-based structural equation models (Kock, 2019). As advised by Hair et al. (1998) and demonstrated by Wang et al. (2024), for any two constructs, A and B, the AVE for A and the AVE for B both need to be larger than the shared variance (i.e., square of the correlation) between A and B. Thus, we compared the AVE of each construct with the shared variance between constructs. The shared variance between FL-LIWC and Present Focus Language is 0.005. This value is less than the AVE and AVE-sq for both Faculty Leadership (AVE 0.025; AVE-sq 0.150) and Present Focus Language (AVE 0.102; AVE-sq 0.319). This finding demonstrates that a theoretically divergent construct (Present Focus Language) measured by the same LIWC method is statistically unrelated to FL-LIWC, providing evidence of discriminant validity.
3.4.4 Hypothesis testing and criterion-related validity
The last structural model (Model 3: SEM Mediation Hypothesis) is constructed to estimate FL-LIWC’s nomological network relationships and test the study’s proposed hypotheses. In Model 3, the four faculty leadership indicators were loaded on one latent measure of FL-LIWC. Faculty gender (0 = men, 1 = women) is the primary independent variable and has direct paths to each of the four performance factors (Productivity, Mentorship, Politics and Service). Direct paths are also drawn connecting gender to FL-LIWC so that direct and indirect effects can be estimated. Results from this model are presented in Table 3.
Table 3. Model 3 goodness of fit indices and estimates.
Model 3 demonstrates good fit with the data (X2 = 66.423; d.f. = 17; CFI = 0.909; RMSEA = 0.066; ECVI = 0.212) and so path estimates were analyzed for hypothesis testing. Significant coefficients were observed for all but one performance factor when regressed on FL-LIWC. Specifically, Political language did not significantly predict FL-LIWC (β = −0.02, p = 0.812). However, general support was found for the remaining nomological network. More Faculty Leadership language was expressed (i.e., higher FL-LIWC) in ERLs when P&T candidates had a high prevalence of Productivity language (β = 0.46, p < 0.001), Mentorship language (β = 0.16, p = 0.036), and Service language (β = 0.28, p < 0.001). These findings lend partial support for Hypothesis 1 and provide evidence of FL-LIWC’s construct validity.
The next set of analyses aimed to understand the role of gender on FL-LIWC controlling for faculty performance (Hypothesis 2). Independent samples t-test comparisons were conducted for each of the four Faculty Leadership and for global Faculty Leadership, the mean FL-LIWC scores across all four components. As shown in Table 1 and as hypothesized (Hypothesis 2a), men faculty have significantly more leadership language reflected in their ERLs than women faculty in general (t (523.41) = 4.59, p = 0.001; Men = 0.47 and Women = 0.43). This gender difference is statistically pronounced for Task Leadership (t (585.61) = 4.69, p = 0.001; Men = 0.83 and Women = 0.69) and External Leadership (t (531.38) = 2.71, p = 0.001; Men = 0.49 and Women = 0.45). This pattern is consistent with prior research that substantiates the perception of leadership as a predominant masculine phenomenon.
Given that faculty leadership is only one type of performance that may be considered in the P&T review process, Hypothesis 2b states that gender difference associated with FL-LIWC would be attenuated when considered in combination with other faculty performance factors. That is, the influence of gender on FL-LIWC would be mediated by other measures of faculty performance. To test this hypothesis, we examine the direct and indirect effects of gender on FL-LIWC through the mechanism of performance to explore the extent to which gender differences in FL-LIWC are predicted by gender alone or potentially controlled through the intervening influence of faculty performance.
As shown in Table 3, significant direct effects are observed for Gender on Service (β = 0.10, p = 0.004) and Political (β = 0.12, p = 0.002) ERL language, each in the positive direction. This finding indicates that these leadership behaviors are expressed more frequently for women faculty than for men faculty. While there are also direct effects of Productivity (β = 0.46, p = 0.001) and Service (β = 0.28, p = 0.001) language on FL-LIWC such that more ERL discussion of production and service predicts more leadership discourse, there are no significant direct or indirect effects of faculty Gender on any of the FL-LIWC indicators. This suggests that all gender differences observed in FL-LIWC are fully explained through the mechanism of the four faculty performance factors (Productivity, Mentorship, Political, and Service). When taking faculty performance into account, the gender differences observed in the FL-LIWC variable disappears. This finding suggests that FL-LIWC is relatively free of potential error biases associated with gender (A summary of all tested models is presented in Supplementary Appendix 4.)
4 Discussion
This two-part multimethod study provides a theoretically-derived and empirically-supported conceptual definition and operational measure (i.e., FL-LIWC) of Faculty Leadership in university settings. The integrative literature review and Delphi panels conducted in Study 1 provide a comprehensive qualitative approach to establishing the content validity of the Faculty Leadership dictionary FL-LIWC measure. Study 2 applied rigorous psychometric testing and empirical analysis of the FL-LIWC measure to establish aspects of content, construct and criterion-related validity. In doing so it also explored practical issues pertaining to gender bias and sources of measurement error known to be a concern in leadership assessment and in P&T review processes (e.g., Masters-Waage et al., 2024). Subsample 1 was applied in an exploration factor analysis (EFA) to explore the underlying component structure of FL-LIWC measures. A confirmatory factor analysis (CFA) using Subsample 2 confirmed a theoretically aligned unidimensional construct. Construct validity was examined using a series of covariance-based structural models (SEM AMOS) to test hypothesized convergent and divergent associations within the FL-LIWC nomological network, and discriminant validity was evidenced using average variance extracted (AVE) comparisons. Finally, the effect of gender difference in the FL-LIWC dictionary was investigated using a mediating structural model that demonstrated that, even though different amounts and types of faculty leadership language are expressed in ERLs for men and women faculty, these differences are largely explained through the mechanism of faculty performance. This finding reflects Jamali et al.’ (2022) conclusion that faculty performance is differentially affected by faculty leadership style. No direct or indirect gender effects were expressed on FL-LIWC. This finding suggests that the FL-LIWC dictionary produces a measurement that is relatively free of gender difference. The cumulative contribution of this research is a validated FL-LIWC dictionary and a standardized process for measuring textual linguistic reflections of leadership behavior performed by university faculty.
4.1 Conclusion
The FL-LIWC dictionary offers a robust and theoretically sound tool for evaluating faculty leadership through ERL text. It provides a novel method for capturing aspects of faculty performance that are often overlooked in traditional P&T processes. It stands to contribute to fair assessment of leadership behaviors and provide a deeper understanding of faculty leadership engagement in higher education. The study advocates for institutions to formally incorporate standardized approaches to leadership assessment to strengthen faculty development and advancement in higher education.
4.2 Implications
Findings from this study suggest that the FL-LIWC dictionary is a theoretically sound and psychometrically robust measure of a newly emerging and important behavioral consideration in higher education. By establishing a validated tool to assess aspects of faculty leadership through ERL text, we hope to offer academic institutions a novel way to capture new insight about leadership behaviors often associated with faculty performance (Jamali et al., 2022); behaviors that have not formally be documented in P&T reviews but certainly have been voiced and reflected in performance evaluations provided by external reviewers engaged in the P&T process. Furthermore, the dictionary offers performance insight free of gender difference, which minimizes the potential impact of gender bias from this measurement source. The FL-LIWC dictionary contributes scientific knowledge toward articulating and measuring leadership behaviors that are often conflated conceptually and in evaluative practice with traditional measures of academic faculty performance. FL-LIWC provides an opportunity for further theoretical differentiation of the faculty leadership construct. From a practical standpoint, this study introduces a new way of assessing and evaluating meaningful behaviors for faculty advancement in higher education. Institutions should examine how faculty leadership is currently identified, both formally (e.g., administrative roles) and informally (e.g., task, relational, change, and external behaviors), how is it valued in relation to expected faculty description (i.e., core or contextual job function), and then consider whether or how such behaviors should be weighed in university decision making practices related to leadership development, performance evaluation, and career advancement. For inclusion in P&T processes, universities should clearly communicate informal leadership behavior expectations, formally specify such behaviors as aspects of performance to be reflected in ERLs when inviting external reviewers, and codifying how FL-LIWC measures should be taken into account in the P&T decision making process. Universities that acknowledge the importance of informal faculty leadership behaviors, and use measurement tools like FL-LIWC to evaluate and reinforce ehaviors, may empower faculty to engage in such behaviors and return greater value to their institutions.
4.3 Limitations and future directions
This study employed a robust methodology using field setting data collected from a volume of P&T cases across multiple higher education institutions. It used two subsamples to explore and validate the FL-LIWC Dictionary, and rigorous modeling to investigate practical applications and gender bias implications for its use in faculty advancement processes. Despite the many strengths of the methodology, a few limitations are worth noting. First, we followed normative standards for establishing validity and acceptable model fit (Lezak et al., 2012; Nye, 2023). While these standards are typically deemed sufficient, particularly for critiquing newly established measurement instruments, acceptable standards may not be stringent enough when using this tool to make high-stakes personnel decisions.
First, future researchers should seek to improve and refine the FL-LIWC dictionary to strengthen its psychometric functioning in accordance with preferred fit standards (CFI values above 0.95 and RMSEA values below 0.06). While most of our SEM models exceeded the preferred standards, not all did. In addition, future researchers should re-evaluate construct validity using a more substantial multitrait-multimethod (MTMM) methodology in complement to the average variance extracted (AVE) comparisons. The low shared variance we observed between FL-LIWC, and Present Focus Language might not provide robust enough, especially when the AVE values are small. These two steps would strengthen confidence in the tool’s accuracy and evaluative utility.
Next, although the study finds that gender differences in FL-LIWC are mediated by performance, other potential biases in the LIWC dictionary variables might influence these results. For example, most of the study variables were obtained using the same method: ERL text frequencies transformed into LIWC dictionary variables. Common method variance (CMV) is systematic error that can occur when the same method is used to measure constructs in a study. Common method bias (CMB) occurs when CMV affected the relationship between the variables. Both pose limitations to this study. ERL writer biases would be similarly expressed in FL-LIWC measures as they would be in LIWC measures of performance, and ERL writers who view leadership as men would also likely have gender biased perceptions of performance, particularly given the recent research demonstrating negative effects of gender bias on performance ratings (Braddy et al., 2020). Although there are common techniques for detecting CMB effects, Schwarz et al. (2017) find inconsistent accuracy in discerning CMB effects and differential impacts on resulting measurements and structural models. Future research should employ research designs that integrate multisource and multimethod data to isolate the CMB effects from relational influences.
Lastly, while a standardized measure of faculty leadership may offer a new tool and added insight into this important emergent higher education concern, it is still unclear the extent to which faculty leadership formally plays in making evaluative judgment in P&T decisions. Future research should explore the extent to which perceptions of faculty leadership and faculty leadership roles affect P&T outcomes to better understand the potential of the FL-LIWC dictionary to influence the overall faculty advancement process.
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Author contributions
TR: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. AM: Formal analysis, Methodology, Supervision, Writing – review & editing. AS: Methodology, Project administration, Resources, Writing – review & editing. SJ: Conceptualization, Formal analysis, Investigation, Methodology, Writing – review & editing. CS: Data curation, Funding acquisition, Project administration, Resources, Writing – review & editing. JM: Data curation, Funding acquisition, Investigation, Methodology, Resources, Supervision, Writing – review & editing.
Funding
The author(s) declare that financial support was received for the research and/or publication of this article. We acknowledge our grantors, including the Alfred P. Sloan Foundation and the National Science Foundation (NSF grant no. 1409928 to R. Khator and NSF grant no. 2100034 to JM). The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.
Acknowledgments
We also thank our entire team within the P&T collaborative research consortium and our External Advisory Board members that support the consortium.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declare that Gen AI was used in the creation of this manuscript. ChatGPT was used to help condense my own original text writing. I used it to improve conciseness and brevity. “Summarize this 1000 word statement into 200 words”.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/feduc.2025.1572933/full#supplementary-material
Footnotes
^The non-categorized leadership behaviors component was omitted from the FL-LIWC dictionary development given the need for further research and understanding of the unclassified terms.
References
Acker, S. (2012). Chairing and caring: gendered dimensions of leadership in academe. Gend. Educ. 24, 411–428. doi: 10.1080/09540253.2011.628927
Agarwal, S., and Bhal, K. T. (2020). A multidimensional measure of responsible leadership integrating strategy and ethics. Group Org. Manag. 45, 637–673. doi: 10.1177/1059601120930140
Alateeg, S., and Alhammadi, A. (2024). The impact of organizational culture on organizational innovation with the mediation role of strategic leadership in Saudi Arabia. J. Stat. Appl. Probab. 13, 843–858. doi: 10.18576/jsap/130220
Anttonen, R. G., and Chaskes, J. (2002). Advocating for first-year students: a study of the micropolitics of leadership and organizational change. J. First-Year Exp. Students Transition. 14, 81–98. doi: 10.65005/154230702813695491
Benmira, S., and Agboola, M. (2021). Evolution of leadership theory. BMJ Lead. 5, 3–5. doi: 10.1136/leader-2020-000296
Berenbaum, M. R. (2019). Impact factor impacts on early-career scientist careers. Proc. Natl. Acad. Sci. 116, 16659–16662. doi: 10.1073/pnas.1911911116,
Black, S. (2015). Qualities of effective leadership in higher education. Open J. Leadership 4, 54–66. doi: 10.4236/ojl.2015.42006
Borgmann, L., Rowold, J., and Bormann, K. C. (2016). Integrating leadership research: a meta- analytical test of Yukl’s meta-categories of leadership. Pers. Rev. 45, 1340–1366. doi: 10.1108/PR-07-2014-0145
Boyd, R. L., Ashokkumar, A., Seraj, S., and Pennebaker, J. W. (2022). The development and psychometric properties of LIWC-22. Austin, TX: University of Texas at Austin.
Braddy, P. W., Sturm, R. E., Atwater, L., Taylor, S. N., and McKee, R. A. (2020). Gender bias still plagues the workplace: looking at derailment risk and performance with self-other ratings. Group Org. Manag. 45, 315–350. doi: 10.1177/1059601119867780
Branson, C. M., Franken, M., and Penney, D. (2016). Middle leadership in higher education: a relational analysis. Educ. Manage. Adm. Leadersh. 44, 128–145. doi: 10.1177/1741143214558575
Britton, D. M. (2017). Beyond the chilly climate: the salience of gender in women’s academic careers. Gend. Soc. 31, 5–27. doi: 10.1177/0891243216681494
Bryman, A. (2007). Effective leadership in higher education: a literature review. Stud. High. Educ. 32, 693–710. doi: 10.1080/03075070701685114
Buchen, I. H. (1998). Servant leadership: a model for future faculty and future institutions. J. Leadersh. Stud. 5, 125–134.
Campbell, D. T., and Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait-multimethod matrix. Psychol. Bull. 56, 81–105. doi: 10.1037/h0046016,
Carlyle, T. (1907). On heroes, hero-worship and the heroic in history (written in 1841). US: Houghton Mifflin Press.
Chen, F. F., West, S. G., and Sousa, K. H. (2006). A comparison of bifactor and second-order models of quality of life. Multivar. Behav. Res. 41, 189–225. doi: 10.1207/s15327906mbr4102_5,
Chung, C. K., and Pennebaker, J. W. (2012). “Linguistic inquiry and word count (LIWC):pronounced Luke, and other useful fact” in Applied natural language processing: Dentification, investigation and resolution. eds. P. M. McCarthy and C. Boonthum-Denecke (USA: Hampton University), 206–229.
Collins, P. H. (2015). Intersectionality’s definitional dilemmas. Annu. Rev. Sociol. 41, 1–20. doi: 10.1146/annurev-soc-073014-112142
Collins, L., and Liang, X. (2015). Examining High Quality Online Teacher Professional Development: Teachers' Voices. Available online at: https://www.semanticscholar.org/paper/Examining-High-Quality-Online-Teacher-Professional-Collins-Liang/a3692eb95fda3b8107262a5a081f540b268d6965 (Accessed November 30, 2025).
Conway, J. M. (1999). Distinguishing contextual performance from task performance for managerial jobs. J. Appl. Psychol. 84, 3–13.
Cronbach, L. J., and Meehl, P. E. (1955). Construct validity in psychological tests. Psychol. Bull. 52, 281–302.
Dalkey, N., and Helmer, O. (1963). An experimental application of the Delphi method to the use of experts. Manag. Sci. 9, 458–467. doi: 10.1287/mnsc.9.3.458,
Drew, G. (2010). Issues and challenges in higher education leadership: Engaging for change. The Australian educational researcher. 37, 57–76.
Eagly, A. H. (1987). Sex differences in social behavior: A social-role interpretation. US: Lawrence Erlbaum Associates, Inc.
Eagly, A. H., and Karau, S. J. (2002). Role congruity theory of prejudice toward female leaders. Psychol. Rev. 109, 573–598. doi: 10.1037/0033-295X.109.3.573,
Estes, B., and Jia Wang, (2008). Integrative literature review: workplace incivility: impacts on individual and organizational performance. Hum. Resour. Dev. Rev. 7, 218–240. doi: 10.1177/1534484308315565
Fernandez, A. A., and Shaw, G. P. (2020). Academic leadership in a time of crisis: the coronavirus and COVID-19. J. Leadersh. Stud. 14, 39–45. doi: 10.1002/jls.21684
Fletcher, J. K. (2004). The paradox of postheroic leadership: an essay on gender, power, and transformational change. Leadersh. Q. 15, 647–661. doi: 10.1016/j.leaqua.2004.07.004
Fullan, M., and Scott, G. (2009). Turnaround leadership for higher education. San Francisco, CA: Jossey-Bass.
Ghiselli, E. E., Campbell, J. P., and Zedeck, S. (1981). Measurement theory for the behavioral sciences. New York: W.H. Freeman and Co.
Gibbs, G., Knapper, C., and Piccinin, S. (2008). Disciplinary and contextually appropriate approaches to leadership of teaching in research-intensive academic departments in higher education. High. Educ. Q. 62, 416–436. doi: 10.1111/j.1468-2273.2008.00402.x
Gonaim, F. (2016). A department chair: a life guard without a life jacket. High Educ. Pol. 29, 272–286. doi: 10.1057/hep.2015.26
Gottlieb, M., Chan, T. M., Yarris, L. M., Linden, J. A., and Coates, W. C. (2022). Promotion and tenure letters: a guide for faculty. AEM Educ Train. 6:e10759. doi: 10.1002/aet2.10759,
Hair, J. E., Anderson, R. E., Tatham, R. L., and Black, W. C. (1998). Multivariate data analysis. 5th Edn. Upper Saddle River, NJ: Prentice-Hall.
Hannum, K. M., Muhly, A. M., Shockley-Zalabak, P. S., and White, J. (2015). Women leaders within higher education in the United States: supports, barriers, and experiences of being a senior leader. Adv. Women Leadersh. 35, 65–75. doi: 10.21423/awlj-v35.a129
Hempsall, K. (2014). Developing leadership in higher education: perspectives from the USA, the UK and Australia. J. High. Educ. Policy Manag. 36, 383–394. doi: 10.1080/1360080X.2014.916468
Hinkin, T. R. (1998). A brief tutorial on the development of measures for use in survey questionnaires. Organ. Res. Methods 1, 104–121.
Hoyt, C. L., and Murphy, S. E. (2016). Managing to clear the air: stereotype threat, women, and leadership. Leadersh. Q. 27, 387–399. doi: 10.1016/j.leaqua.2015.11.002
Jamali, A., Bhutto, A., Khaskhely, M., and Sethar, W. (2022). Impact of leadership styles on faculty performance: moderating role of organizational culture in higher education. Manag. Sci. Lett. 12, 1–20. doi: 10.5267/j.msl.2021.8.005
Jing, F. F., and Avery, G. C. (2008). Missing links in understanding the relationship between leadership and organizational performance. Int. Bus. Econ. Res. J. 7, 67–78. doi: 10.19030/iber.v7i5.3256
Johnson, N. N. (2023). “Stabilizing and empowering women in higher education: aligning, centering, and building” in Stabilizing and empowering women in higher education: Realigning, Recentering, and rebuilding. eds. H. Schnackenberg and D. Simard (US: IGI Global), 1–18.
Keith, S. J., and Buckley, P. F. (2011). Leadership experiences and characteristics of chairs of academic departments of psychiatry. Acad. Psychiatry 35, 118–121. doi: 10.1176/appi.ap.35.2.118,
Kezar, A., Gallant, T. B., and Lester, J. (2011). Everyday people making a difference on college campuses: the tempered grassroots leadership tactics of faculty and staff. Stud. High. Educ. 36, 129–151. doi: 10.1080/03075070903532304
Kezar, A., and Lester, J. (2009). Supporting faculty grassroots leadership. Res. High. Educ. 50, 715–740. doi: 10.1007/s11162-009-9139-6
Kock, N. (2019). From composites to factors: bridging the gap between PLS and covariance- based structural equation modeling. Inf. Syst. J. 29, 674–706. doi: 10.1111/isj.12228
Koenig, A. M., Eagly, A. H., Mitchell, A. A., and Ristikari, T. (2011). Are leader stereotypes masculine? A meta-analysis of three research paradigms. Psychol. Bull. 137, 616–642. doi: 10.1037/a0023557,
Lawler, P. A., and King, K. P. (2000). Faculty development: leadership strategies for success. J. Contin. High. Educ. 48, 12–20. doi: 10.1080/07377366.2000.10400401
Lezak, M. D., Howieson, D. B., Bigler, E. D., and Tranel, D. (2012). Neuropsychological assessment. 5th Edn. New York, NY: Oxford University Press.
Makarova, E., Aeschlimann, B., and Herzog, W. (2019). The gender gap in STEM fields: the impact of the gender stereotype of math and science on secondary students' career aspirations. Front. Educ. 4:60. doi: 10.3389/feduc.2019.00060
Masters-Waage, T., Madera, J., Henderson, E., Gutierrez, A., Edema-Sillo, E., Werner, C., et al. (2024). Underrepresented minority faculty face double standard in promotion and tenure decisions. Nat. Hum. Behav. 8, 2107–2118. doi: 10.1038/s41562-024-01977-7
National Academies of Sciences, Engineering, and Medicine (NASEM), (2023). Advancing antiracism, diversity, equity, and inclusion in STEMM organizations: Beyond broadening participation. Available online at: https://nap.nationalacademies.org/catalog/26803/advancing-antiracism-diversity-equity--and-inclusion-in-stemm-organizations-beyond (Accessed on February 4, 2025)
Nguyen, D., Harris, A., and Ng, D. (2020). A review of the empirical research on teacher leadership (2003–2017): evidence, patterns and implications. J. Educ. Adm. 58, 60–80. doi: 10.1108/JEA-02-2018-0023
Nietzal, M.T. (2024). Black and Hispanic faculty disadvantaged in promotion and tenure process, study finds. Forbes. Available online at: https://www.forbes.com/sites/michaeltnietzel/2024/08/02/study-finds-black--hispanic-faculty-disadvantaged-in-promotiontenure-process/ (Accessed November 30, 2025).
Nye, C. D. (2023). Reviewer resources: confirmatory factor analysis. Organ. Res. Methods 26, 608–628. doi: 10.1177/10944281221120541
Oc, B. (2018). Contextual leadership: a systematic review of how contextual factors shape leadership and its outcomes. Leadersh. Q. 29, 218–235. doi: 10.1016/j.leaqua.2017.12.004
Orieno, O. H., Udeh, C. A., Oriekhoe, O. I., Odonkor, B., and Ndubuisi, N. L. (2024). Innovative management strategies in contemporary organizations. Int. J. Manag. Entrepreneurship Res. 6, 167–190. doi: 10.51594/ijmer.v6i.727
Parmar, B. L., Wicks, A. C., and Freeman, R. E. (2022). Stakeholder management & the value of human-centered corporate objectives. J. Manag. Stud. 59, 569–582. doi: 10.1111/joms.12716
Pate, J. C., and Angell, L. R. (2013). Factors community college faculty consider important to academic leadership. Ky. J. High. Educ. Policy Pract. 2, 1–13. Available online at: https://uknowledge.uky.edu/kjhepp/vol2/iss1/1
Ray, V. (2019). A theory of racialized organizations. Am. Sociol. Rev. 84, 26–53. doi: 10.1177/0003122418822335
Rosari, R. (2019). Leadership definitions application for lecturers’ leadership development. Journal of Leadership in Organizations. 1, 17–28.
Reynolds, G., Samuels, K. L., Din, C., and Turner, N. (2023). Doing leadership development through mentoring in a social learning space: the case of the inaugural leadership learning lab. J. Manag. Dev. 42, 234–252. doi: 10.1108/JMD-04-2022-0082
Rezvanifar, F., Shariat, S. V., Shalbafan, M., Salehian, R., and Rasoulian, M. (2022). Developing an educational package to improve attitude of medical students toward people with mental illness: a Delphi expert panel, based on a scoping review. Front. Psychol. 13:117. doi: 10.3389/fpsyt.2022.860117,
Rowe, W. G., Cannella, A. A., Rankin, D., and Gorman, D. (2005). Leader succession and organizational performance: integrating the common-sense, ritual scapegoating, and vicious-circle succession theories. Leadersh. Q. 16, 197–219. doi: 10.1016/j.leaqua.2005.01.001
Schimanski, L. A., and Alperin, J. P. (2018). The evaluation of scholarship in academic promotion and tenure processes: past, present, and future. F1000Res 7:1605. doi: 10.12688/f1000research.16493.1,
Schwarz, A., Schwarz, C., Rizzuto, T., Roldán, J., and Barrera, R. (2017). Examining the impact and detection of the “urban legend” of common method Bias. DATA BASE for Advan. Information Syst. 48, 93–119. doi: 10.1145/3051473.3051479
Scott, G., Coates, H., and Anderson, M. (2008). Learning leaders in times of change: Academic leadership capabilities for Australian higher education. Strawberry Hills, NSW: Carrick Institute for Learning and Teaching in Higher Education Ltd.
Tausczik, Y. R., and Pennebaker, J. W. (2010). The psychological meaning of words: LIWC and computerized text analysis methods. J. Lang. Soc. Psychol. 29, 24–54. doi: 10.1177/0261927X09351676
Torraco, R. J. (2005). Writing integrative literature reviews: guidelines and examples. Hum. Resour. Dev. Rev. 4, 356–367. doi: 10.1177/1534484305278283
Tsoh, J., Kuo, A., Barr, J., Whitcanack, L., Merry, I., Alldredge, B., et al. (2019). Developing faculty leadership from ‘within’: a 12-year reflection from an internal faculty leadership development program of an academic health sciences center. Med. Educ. Online 24:1567239. doi: 10.1080/10872981.2019.1567239,
Turner, C. S. V., González, J. C., and Wong, K. (2011). Faculty women of color: the critical nexus of race and gender. J. Divers. High. Educ. 4, 199–211. doi: 10.1037/a0024630
Wang, Z., Jiang, H., Jin, W., Jiang, J., Liu, J., Guan, J., et al. (2024). Examining the antecedents of novice STEM teachers’ job satisfaction: the roles of personality traits, perceived social support, and work engagement. Behav. Sci. 14:214. doi: 10.3390/bs14030214,
Watkins, M. W. (2018). Exploratory factor analysis: a guide to best practice. J. Black Psychol. 44, 219–246. doi: 10.1177/0095798418771807
Wieczorek, D., and Lear, J. (2018). Building the "bridge": teacher leadership for learning and distributed organizational capacity for instructional improvement. Int. J. Teach. Leadersh. 9, 22–47.
Williams, J. C., and Dempsey, R. (2014). What works for women at work: Four patterns working women need to know. US: NYU Press.
Yukl, G. (2012). Effective leadership behavior: what we know and what questions need more attention. Acad. Manag. Perspect. 26, 66–85. doi: 10.5465/amp.2012.0088
Yukl, G., Gordon, A., and Taber, T. (2002). A hierarchical taxonomy of leadership behavior: integrating a half century of behavior research. J. Leadersh. Org. Stud. 9, 15–32. doi: 10.1177/107179190200900102
Yukl, G., and Mahsud, R. (2010). Why flexible and adaptive leadership is essential. Consult. Psychol. J. Pract. Res. 62, 81–93. doi: 10.1037/a0019835
Keywords: ERL, tenure and promotion, mixed methodologies, faculty leadership, LIWC
Citation: Rizzuto TE, Manongsong AM, St Aubin A, Jeong S, Spitzmueller C and Madera JM (2025) Measuring leadership language: faculty leadership LIWC dictionary validation study. Front. Educ. 10:1572933. doi: 10.3389/feduc.2025.1572933
Edited by:
Sereyrath Em, University of Cambodia, CambodiaReviewed by:
Saban Bon, Mae Fah Luang University, ThailandSara Omair, National University of Modern Languages, Pakistan
Copyright © 2025 Rizzuto, Manongsong, St Aubin, Jeong, Spitzmueller and Madera. 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.
*Correspondence: Tracey E. Rizzuto, dHJpenp1dEBsc3UuZWR1