DOES IDENTITY MANAGEMENT BEGET WORK-LIFE BOUNDARY MANAGEMENT? AN EXAMINATION OF LESBIAN, GAY & BISEXUAL EMPLOYEES By Danielle Gardner A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Psychology – Doctor of Philosophy 2021 ABSTRACT DOES IDENTITY MANAGEMENT BEGET WORK-LIFE BOUNDARY MANAGEMENT? AN EXAMINATION OF LESBIAN, GAY & BISEXUAL EMPLOYEES By Danielle Gardner For employees facing identity-based stigma, bringing one’s whole self to work may expose oneself to experienced prejudice; accordingly, the present dissertation examined theoretical models unpacking the notion that those seeking to conceal a stigmatized identity at work may subsequently separate their work and life spheres, regardless of their ideal boundaries between their work and home lives. Across two multi-wave studies, the current work demonstrates a novel relationship between stigmatized identity management and enacted boundary management for lesbian, gay and bisexual workers (Study 1), supporting the prediction that those engaged in distancing identity management strategies are more likely to segment their work and home lives. Study 2 provided limited support of differential boundary management congruence when comparing heterosexual and sexual minority employees, though the importance of boundary management congruence in predicting a number of work-relevant attitudes was underscored, and disparities in work-relevant attitudes across heterosexual and sexual minority employees emerged. Theoretical implications for work-life and diversity literatures, as well as practical learnings for organizations are discussed. I dedicate this dissertation to my parents, Dan and Sharon Gardner. I could not have made it this far without you. iii ACKNOWLEDGEMENTS Beginning with my parents, I do not have words to adequately thank Dan and Sharon Gardner for their endless support. I likely can never repay the sacrifices you made to allow me a boundless education, though I will try. Outside of academics, you were always at every play, every basketball game, every football game watching me cheerlead from the sidelines- your unwavering support has no doubt shaped me in ways I can never comprehend, and I am and will be forever grateful. I would also like to thank my feline pals (Mick, Ollie, Bailey, Kirby and Bandit), who throughout my years have provided unmeasurable comfort and taught me the importance of compassion. I would like to deeply and sincerely thank my advisor, Dr. Ann Marie Ryan, for her developmental support on this project, in addition to her mentorship throughout my entire graduate career. I cannot imagine a better advisor if I tried, and I am proud and grateful to be a product of your guidance and support. Additionally, I would like to thank my committee members, Dr. Kevin Ford, Dr. Fred Leong and Dr. Chris Nye, for their helpful suggestions and insightful comments throughout this dissertation process and beyond- I view this final project as a true culmination of my graduate school learnings, of which each of you played an important role. Finally, I would like to thank my partner, Ryan, for his support and encouragement through every program milestone, every tough moment, and every moment of celebration. You truly bring levity and joy when the days seem long and uncertain, so I am forever grateful that I can rely on you always- thank you. iv TABLE OF CONTENTS LIST OF TABLES........................................................................................................................vii LIST OF FIGURES........................................................................................................................ix CHAPTER 1: INTRODUCTION ................................................................................................... 1 CHAPTER 2: STIGMA & IDENTITY MANAGEMENT ............................................................ 4 Sexual Orientation Stigma .......................................................................................................... 4 Identity Management .................................................................................................................. 7 CHAPTER 3: WORK-LIFE BOUNDARY MANAGEMENT .................................................... 11 Boundary Theory ...................................................................................................................... 11 Border Theory ........................................................................................................................... 12 Preferred Versus Enacted Boundary Management ............................................................... 13 Linking Identity Management & Boundary Management ........................................................ 15 CHAPTER 4: HYPOTHESES...................................................................................................... 18 Study 1 ...................................................................................................................................... 18 Study 2 ...................................................................................................................................... 26 CHAPTER 5: STUDY 1 METHOD ............................................................................................. 29 Procedure .................................................................................................................................. 29 Participants ................................................................................................................................ 29 Measures (Time 1) .................................................................................................................... 31 Identity Management ............................................................................................................ 31 Boundary Management (Enacted and Preferred) .................................................................. 31 Diversity Climate Perceptions .............................................................................................. 33 Job Title & Work Autonomy ................................................................................................ 33 Demographic Variables ........................................................................................................ 34 Measures (Time 2) .................................................................................................................... 34 Subjective Wellbeing ............................................................................................................ 34 Withdrawal ............................................................................................................................ 35 Job Satisfaction ..................................................................................................................... 35 CHAPTER 6: STUDY 1 RESULTS............................................................................................. 36 Hypothesis Testing.................................................................................................................... 36 Mediational Analyses................................................................................................................ 40 Exploratory Analysis ................................................................................................................ 41 Discussion ................................................................................................................................. 42 CHAPTER 7: STUDY 2 METHOD ............................................................................................. 46 Procedure & Measures .............................................................................................................. 46 Participants ................................................................................................................................ 46 v Analytic Technique ................................................................................................................... 47 CHAPTER 8: STUDY 2 RESULTS............................................................................................. 49 Hypothesis Testing.................................................................................................................... 49 Combined Path Model .............................................................................................................. 51 CHAPTER 9: DISCUSSION........................................................................................................ 53 Theoretical and Practical Implications...................................................................................... 56 Limitations and Future Directions ............................................................................................ 58 Conclusion ................................................................................................................................ 60 APPENDICES .............................................................................................................................. 62 APPENDIX A: Tables and Figures .......................................................................................... 63 APPENDIX B: Study Measures ............................................................................................... 89 REFERENCES ............................................................................................................................. 96 vi LIST OF TABLES Table 1. Study 1 means, standard deviations, scale reliabilities and intercorrelations. ................ 63 Table 2. Hierarchical regression results of boundary management preferences and autonomy operationalized via self-report as predictive of enacted boundary management. ......................... 64 Table 3. Hierarchical regression results of boundary management preferences and autonomy operationalized via O*NET scores as predictive of enacted boundary management. .................. 65 Table 4. Linear regression results of distancing and affirming identity management predicting enacted boundary management. .................................................................................................... 66 Table 5. Multiple regression results of affirming and distancing identity management strategies as predicting enacted boundary management. .............................................................................. 67 Table 6. Linear regression results of distancing identity management as predictive of enacted life-to-work and work-to-life segmentation. ................................................................................. 68 Table 7. Hierarchical regression results of boundary management preferences and enacted boundary management as predictive of subjective wellbeing. ..................................................... 69 Table 8. Hierarchical regression results of boundary management preferences and enacted boundary management as predictive of job satisfaction. .............................................................. 70 Table 9. Hierarchical regression results of boundary management preferences and enacted boundary management as predictive of withdrawal. .................................................................... 71 Table 10. Linear regression results of diversity climate perceptions predicting affirming and distancing identity management. .................................................................................................. 72 Table 11. Linear regression results of diversity climate perceptions predicting subjective wellbeing, job satisfaction and work withdrawal. ........................................................................ 73 Table 12. PROCESS Model 4 results of diversity climate perceptions predicting enacted boundary management through distancing identity management. ............................................... 74 Table 13. PROCESS Model 4 results of distancing identity management predicting subjective wellbeing through enacted boundary management....................................................................... 75 Table 14. PROCESS Model 4 results of distancing identity management predicting withdrawal through enacted boundary management. ...................................................................................... 76 Table 15. Linear regression results of boundary management congruence operationalized via polynomial regression predicting subjective wellbeing, withdrawal, and job satisfaction. .......... 77 vii Table 16. Study 2 polynomial regression results predicting subjective wellbeing. ...................... 78 Table 17. Study 2 polynomial regression results predicting withdrawal. ..................................... 79 Table 18. Study 2 polynomial regressions results predicting job satisfaction. ............................. 80 Table 19. Means, standard deviations, and intercorrelations between Study 2 variables. ............ 81 Table 20. PROCESS Model 4 results of sexual orientation predicting subjective wellbeing through boundary management congruence. ................................................................................ 82 Table 21. PROCESS Model 4 results of sexual orientation predicting withdrawal through boundary management congruence............................................................................................... 83 Table 22. PROCESS Model 4 results of sexual orientation predicting job satisfaction through boundary management congruence............................................................................................... 84 Table 23. Standardized structural model results for relationships between sexual orientation and Study 2 outcomes through boundary management congruence indices. ...................................... 85 viii LIST OF FIGURES Figure 1. Mapping of reviewed identity management techniques across a general spectrum of identity management. .................................................................................................................... 86 Figure 2. Proposed model predicting enacted boundary management via identity management, boundary management preferences, and job characteristics, as it relates to individual outcomes depending on ideal segmentation preferences. ............................................................................. 87 Figure 3. Proposed model linking sexual orientation to level of congruence between preferred and enacted boundary management as predictive of attitudinal outcomes. .................................. 88 ix CHAPTER 1: INTRODUCTION “Where I work isn’t accepting of alternative lifestyles, so I don’t talk about family other than my children. I don’t talk about my partner except with a few people I trust (Participant 35 of Sawyer, Thoroughgood & Ladge, 2017).” “There was a time when my girlfriend was here, and she wanted to visit my office and that stressed me out. It’s just the questions that would come after. I feel like she is obviously gay, so people would ask questions. I don’t want to lie to people (Participant 40 of Sawyer et al, 2017).” Broadly, work-life boundary management has been conceptualized as the degree to which one segments (i.e. keeps separate) or integrates (i.e. brings together) their work and non-work spheres. However, for those with home lives or identities that are stigmatized within their workplace, how are they to manage such boundaries? As exhibited in the above quotes, to the extent that bringing one’s home life to the office exposes an employee to stigma at work, that employee may be limited in the degree to which they are comfortable integrating all life aspects; accordingly, while all employees are faced with the consideration of how and to what degree to integrate life spheres, it becomes apparent that some workers face greater barriers than others. It is this notion that is the basis for the present dissertation. In this two-study dissertation, I argue that individuals of invisible stigmatized identities (sexual orientation, in the present case) face an additional constraint to the manner in which they are able to manage their work and non-work boundaries, beyond the constraints faced by those without such identities: their identity management. That is, I suggest that the extent to which an individual distances (vs. affirms) their 1 identity at work may impact the manner in which they enact boundaries between their work and non-work spheres. I examine this notion across two studies. First, I explored a potential link between lesbian, gay and bisexual (LGB) workers’ identity management and subsequent enacted boundary management through a multi-wave survey paradigm. Through this effort, I aimed to show how identity management is a constraining factor (alongside other theoretically informed organization-level and job-level constraints) impacting how LGB employees manage their work and non-work spheres. In my second study, I evaluated whether compared to heterosexual employees, LGB employees face greater misfit between their preferred and enacted boundary management as a result of this identity management constraint; I hypothesized that this increased misfit may negatively impact attitudes (i.e. job satisfaction, withdrawal, subjective wellbeing) relevant to the work experience. In conducting these studies, I contribute theoretically in multiple manners. First, via the examination of individuals of minority sexual groups, I aim to highlight the additional constraints workers of invisible stigmatized identities (e.g. those of stigmatized faiths, those with concealable disabilities) broadly face when considering work-life boundary enactment. Therefore, the present work should contribute to the literature on workplace diversity by identifying and uncovering mechanisms surrounding work-life processes for stigmatized workers. Further, by examining the unique processes of individuals of such (often, less studied) groups, I seek to address broader criticisms of the work-life literature as focusing disproportionately on White heterosexual women (Aycan, 2008), via the integration of identity- based theoretical perspectives. 2 Beyond contributing via the identities of focus, the present work seeks to offer greater conceptual clarity surrounding the process of boundary management broadly. While some recent work has differentiated between one’s boundary management preferences and their enacted boundary management (e.g. Powell & Greenhaus, 2010; Carlson, Ferguson & Kacmar, 2016), the relationship between these variables has been examined almost exclusively linearly (i.e. preferences as predictive of enactment). The current research seeks to extend understanding by uncovering greater nuance between these related concepts; beyond examining preferences as an antecedent of enacted boundary management, I examine whether situational constraints (i.e. limiting work context factors) moderate that relationship. Further, I examined the interaction of preferences and enactment (operationalized via both hierarchical and polynomial regression) as predictive of work-relevant attitudes. Accordingly, the present work advances a more refined understanding of work-life boundary management via 1) the examination of a group currently underrepresented in this particular literature (i.e. sexual minorities), 2) the identification and investigation of moderating boundary conditions to an established linear relationship (i.e. boundary management preferences to enacted boundary management), and 3) the evaluation of the interaction between preferred and enacted boundary management and its associated consequences at work. This dissertation will be organized as follows: first, I will outline literature related to sexual orientation stigma and identity management at work. Next, I will discuss two theoretical perspectives related to work-life boundary management, before considering identity management and boundary management in tandem. Upon presenting logic for each hypothesis across both studies, I will present each study’s methodology, analytical approach, and results. I will conclude this dissertation with discussion of theoretical and practical study implications. 3 CHAPTER 2: STIGMA & IDENTITY MANAGEMENT Individuals may possess characteristics that are stigmatized in nature. As defined by Goffman (1963), stigma refers to the situation in which an individual with an attribute faces societal rejection and discreditation as a result of perceptions of that attribute. Jones and colleagues (1984) further developed our understanding of stigma through the introduction of a framework identifying six characteristics (i.e. controllability, stability, visibility, disruptiveness, peril, and aesthetics 1) defining the degree to which a stigmatized characteristic is likely discredited, and becoming perhaps the seminal framework for understanding stigma broadly. Below, I discuss how each of these characteristics potentially describe the stigma surrounding individuals of minority sexual orientations, before connecting this stigma to individuals’ identity management at work. Sexual Orientation Stigma Individuals of minority sexual orientations (e.g. gay, lesbian, bisexual) may face stigma within the United States working context. Considering the first characteristic within Jones and colleagues’ (1984) framework, sexual orientation may be stigmatized particularly when it is viewed as controllable in nature. The notion of perceived controllability as linked to stigmatization has been demonstrated with respect to weight (e.g. Tiggeman & Anesbury, 2000; Hebl & Kleck, 2002) and disability (e.g. Lyons, Volpone, Wessel & Alonso, 2017; Hebl & Kleck, 2002), among other characteristics (e.g. mental illness, alcoholism, Hegarty & Golden, 2008). With respect to sexual orientation, while growing evidence suggests that sexual 1 Given the invisible nature of sexual orientation, the sixth component (aesthetics) is not particularly relevant when considering this characteristic and is therefore not discussed further. 4 orientation is indeed biological and not up to individual discretion (Mustanski, Chivers & Bailey, 2002), research suggests that individuals vary to the extent they perceive sexual orientation as a choice under one’s control (Hegarty, 2002). It follows then that those who do perceive sexual orientation as controllable would be more likely to reject and discredit those of minority sexual orientation identities. Perceptions of sexual orientation stability (a component that has been examined with respect to a variety of physical and psychological ailments; Weiner, Perry & Magnusson, 1988; Ruybal & Siegel, 2019) likely relate to perceptions of sexual orientation controllability (Weiner et al., 1988). That is, those who believe that individuals have control over their sexual orientation likely perceive the characteristic as changeable and under the individual’s discretion. However, the same evidence that suggests the biological determinants and influences of sexual orientation would also likely suggest the characteristic as relatively static in nature. Despite this evidence, it is the extent to which this perception of sexual orientation’s controllability pervades that may determine to what degree sexual orientation is accordingly stigmatized. Concerning the level of sexual orientation’s visibility, this characteristic is largely considered invisible (rather than visible) in nature (Ragins, Singh & Cornwell, 2007), as are characteristics such as religious affiliation and certain medical conditions (Ragins, 2008). As an individual’s sexual orientation is not immediately discernable from their outward appearance (as is comparatively the case for an individual’s race and gender presentation), it is then up to the individual to decide if and to what degree they would like their identity to be known. Therefore, individuals of sexual minority identities might attempt to avoid stigma by continuing to conceal their orientation from being publicly known. 5 Two final characteristics identified by Jones and colleagues (1984) that help to understand the stigma surrounding sexual orientation are disruptiveness and peril. Broadly, disruptiveness concerns the extent to which a stigmatized attribute is seen as hampering the possibility of successful interactions with others, and has been evaluated with respect to HIV- status (Crandall & Coleman, 1992) and mental illness (Feldman & Crandall, 2007; Perry, 2014). Sexual orientation should not strongly be linked to disruptiveness, beyond the extent to which people feel uncomfortable interacting with those who differ from themselves on a stigmatized attribute (Hebl, Tickle, Heatherton, 2000). Similarly, sexual orientation should not be particularly linked with peril, although the extent that minority sexual identities are perceived as challenging the status quo and subsequently threatening and devaluing the traditional might suggest potential perceptions of peril (Lyons, Pek & Wessel, 2017). Symbolic threat theory would support this notion (Riek, Mania & Gaertner, 2006), suggesting that individuals of minority sexual identities might be devalued to the extent that heterosexuals perceive them as a threat to the values they hold true. Acknowledging this review, the stigma surrounding minority sexual orientation identity can likely be explained through external perceptions of controllability, instability, and potential peril. However, it is the visibility characteristic that is of primary focus for the proposed set of studies. As sexual orientation is invisible in nature, individuals of stigmatized orientations are often faced with how to present their group membership at work so as to minimize experienced prejudice and discrimination. In contrast to visible stigmatized characteristics, individuals of minority sexual identities are able to decide if and to what extent they desire to make their orientation known, a process known as identity management. Below, I review and integrate three 6 frameworks of identity management, before presenting logic linking this process to work-life boundary management. Identity Management Recognizing that individuals of particular identities face societal stigma, one question that may follow is how such individuals manage this stigma within their working lives. These circumstances may be further complicated for those with invisible (rather than visible) stigmatized identities, as the decision of how to manage one’s identity is predicated on the decision of whether to make their identity externally known. Given the possible costs associated with facing stigma, people with stigmatized invisible identities are likely to strategically consider whether, when, and how to reveal their characteristic (Ragins, 2008). Accordingly, a number of researchers have attempted to understand the antecedents and characteristics of identity management for individuals of such demographics, and I will review a number of such perspectives here with an eye toward integration (see Figure 1). One perspective regarding identity management for invisible identities comes from Clair, Beatty and MacLean (2005), who hypothesize that individual and contextual circumstances may predict one’s decision to “pass” or “reveal” their invisible social identity at work. One end of the identity management spectrum as discussed by Clair and colleagues (2005) concerns the degree to which one decides to “pass” with respect to their stigmatized identity at work. “Passing” is defined as the circumstances in which an individual is externally classified incorrectly as someone without a discredited or devalued social identity, allowing that individual to access the privileges afforded to the dominant, non-stigmatized group. Examples of passing would include those with invisible illnesses passing as healthy, those of minority sexual orientations passing as heterosexual, and those of stigmatized religious identities passing as either areligious or of the 7 dominant and non-stigmatized faith. This umbrella of “passing” is further differentiated into three specific tactics: fabrication (i.e. occurring when an individual deliberately lies regarding their standing on an identity), concealment (i.e. involving the active prevention of external information acquisition concerning the identity in question), and discretion (i.e. occurring when an individual does not engage in active concealment, but rather avoids the topic surrounding their identity characteristic). In contrast, the degree to which one “reveals” their invisible stigmatized identity at work would represent the counter end of the identity management spectrum discussed by Clair and colleagues (2005), and involves individuals making their standing on some stigmatized characteristic externally known. Like “passing,” the overall strategy of “revealing” was further distinguished into three more specific strategies: signaling (i.e. the extent to which one hints at their invisible identity, allowing those external to the self to infer group membership without explicitly expressing their identity), normalizing (i.e. occurring when individuals disclose their identity and attempt to make their difference perceived as commonplace), and differentiating (i.e. occurring when individuals present their identity as equally valid to those that are not stigmatized in nature). Anderson, Croteau, Chung and DiStefano (2001) present a similar framework describing the spectrum on which one could manage their invisible stigmatized identities, suggesting that individuals could employ one of four strategies. On one end of their spectrum is “passing,” and is defined by these authors as involving the fabrication of information in order to be perceived as a member of the dominant, non-stigmatized group (similar to the “fabrication” tactic within Clair and colleagues’ framework). Next, the authors outline the “covering” strategy, which involves the censoring of information in order for the individual’s identity to not be discovered (similar to 8 the “concealment” tactic within Clair and colleagues’ framework). Third, the authors define the “implicitly out” strategy as one that allows some discussion of an individual’s identity without explicitly labeling oneself as a member of that group (similar to Clair and colleagues’ “signaling” tactic). Finally, Anderson and colleagues (2001) define the opposite end of the spectrum from “passing” as “explicitly out,” involving honest communication about one’s invisible identity. A final conceptualization of identity management related to those reviewed previously outlines the extent to which one affirms (e.g. openly discusses their identity with coworkers) as compared to distances (e.g. actively conceals their identity from coworkers) from their stigmatized identity at work (Button, 2004; Lyons, Wessel, Ghumman, Ryan & Kim, 2014). Similar to the “passing” strategy outlined by Clair and colleagues (2005) and the “covering” strategy identified by Anderson et al. (2001), distancing involves the deemphasis and salience reduction of an individual’s identity within a given context (Lyons, Zatzick, Thompson & Bushe, 2017; Shih, Young & Bucher, 2013). In contrast, the affirming identity management strategy involves the public pronouncement of one’s identity and the positive aspects with which that identity is associated (Lyons et al., 2017). Beyond simple disclosure (as is framed within the “explicitly out” strategy outlined by Anderson and colleagues, 2004), identity management affirmation is most similar to Clair and colleagues’ (2005) differentiation strategy, as both are conceptualized as involving the positive promotion of one’s identity on top of the predicated disclosure. Figure 1 maps each of the three reviewed perspectives along a single identity management continuum, allowing for comparison across each perspective’s respective components. 9 Via the integration and presentation of multiple identity management frameworks, it becomes clear that while there are a number of ways one could consider managing their stigmatized identity at work, all conceptualizations generally fall along a continuum with a positive affirming orientation at one end, and a concealing or distancing orientation at the other. Accordingly, I apply this broadest framework of identity management in the present studies, differentiating between those employing “distancing” management techniques as compared to those enacting “affirming” management techniques with respect to their identity at work. Acknowledging this review of stigma and identity management as it relates to employees’ sexual orientation, I next will discuss literature summarizing work-life boundary management broadly. Upon reviewing these concepts, I present my logic theoretically linking identity management with enacted work-life boundary management for employees of minority sexual orientations. 10 CHAPTER 3: WORK-LIFE BOUNDARY MANAGEMENT Workers of all identities are tasked with managing their respective life roles inside and outside of the workplace; this has been the primary notion underlying research examining work- life boundary management. Given that the present studies aim to understand how LGB employees’ identity management may inform their work-life boundary management, I first review two theoretical perspectives (i.e. boundary theory and border theory) crucial to understanding boundary management broadly, before outlining logic surrounding the hypothesized relationship between identity management and boundary management Boundary Theory Ashforth, Kreiner and Fugate’s (2000) discussion of boundary theory recognizes that individuals occupy multiple roles, and accordingly are tasked with transitioning between roles on a daily basis. These role transitions are conceptualized as boundary-crossing activities, whose engagement are informed by the extent to which role boundaries are flexible (i.e. the degree to which the spatial and temporal boundaries are pliable) and permeable (i.e. the degree to which a role allows one to be physically located in one role but psychologically and/or behaviorally involved in another) in nature. The extent then that these boundaries are indeed flexible and permeable is thought to relate to the ease with which one transitions from one role to another. Ashforth and colleagues (2000) suggest that these boundary characteristics (i.e. flexibility and permeability) inform where individuals lie upon a continuum of work-life boundary management, known as the segmentation-role integration continuum (Nippert-Eng, 1996). At one end of this continuum, role segmentation represents the extent to which role identities and their respective contexts are completely separated in nature, as a function of one’s inflexible and 11 impermeable role boundaries. One example of strong role segmentation would be seen for members of highly stigmatized roles (e.g. exotic dancer, sex worker) who may conceal their employment from friends or neighbors (Ashforth & Kreiner, 1999). Ashforth and colleagues (2000) posit that individuals engaging in role segmentation are able to reduce blurring between roles (allowing for clarity regarding when one is occupying which role within a given moment), compartmentalize roles psychologically, and face less cross-role interruptions. At the same time, segmentation poses greater barriers for cross-role transitions, as the roles in question are potentially both psychologically and physically differentiated from one another. Conversely, role integration (the opposite end of the continuum from segmentation) is denoted by roles that are weakly differentiated, not strictly tied to physical and temporal circumstances, and easily facilitate cross-role interruptions. Roles that are highly integrated (rather than segmented) may further be associated with similarity in associated identities, as well as physical location; in line with this notion, one example of role integration may be seen with employees who work from home, as their work and non-work lives may be particularly intertwined given the similarity in location in which either role is enacted. Given the similarity between spheres then, one would expect individuals of integrated roles to easily jump from one role to another. However, role integration is concurrently associated with fuzzy boundaries between roles, such that the roles themselves are blurred within one another, potentially causing anxiety from lack of clarity and inhibition of full role disengagement. Border Theory Clark (2000) presents similar ideas within her work-family border theory, although placing a greater theoretical focus on the notion of “borders” (conceptually akin to Ashforth et al.’s definition of “boundaries”) between the separate work and life spheres. Borrowing from 12 Lewin’s notion of separate psychological domains for work and family, Clark (2000) suggests that the degree of interaction between domains (i.e. segmentation vs. integration) depends on the strength of the border between them. Similar then to Ashforth and colleagues’ (2000) notion of cross-role transitions, Clark (2000) suggests that individuals are border-crossers tasked with moving between different contexts demarcated by borders varying on their respective flexibility and permeability. Clark (2000) argues for a more active conceptualization of work-life boundary management than what was previously represented in the literature, noting that individuals proactively shape their spheres toward their individual preference of balance. This is in contrast to the notion that individuals’ boundary management is solely a reactive process, in which one’s circumstances are wholly deterministic of the boundaries that are constructed between roles. Clark’s (2000) theory concludes with a series of propositions outlining how an individual’s border strength, role identification, and the awareness and commitment of external domain members (i.e. those who cohabit the work or life role) influence the extent to which an individual’s boundary management is associated with subsequent role satisfaction and balance. Preferred Versus Enacted Boundary Management Acknowledging individual differences in how work and non-work spheres are managed (i.e. segmentation vs. integration; Ashforth et al., 2000; Clark, 2000; Nippert-Eng, 1996), it is further worthwhile to differentiate one’s ideal boundary management strategies from those that are actually implemented. “Enacted boundaries” are the actual demarcations created by individuals between their respective life domains, whereas “boundary preferences” represent the ideal boundaries individuals wish to create between differing life spheres (Ammons, 2013). The consideration of enacted versus preferred boundary management is important, as it recognizes that while enacted boundary management is likely influenced by boundary management 13 preferences (Clark, 2000), other contextual influences also impact how one manages their work and non-work boundaries. Indeed, Nippert-Eng (1996) makes a point to highlight the social constraints of work and family that inhibit individual discretion, effectively narrowing the range of potential boundary management options available to the individual (Ammons, 2013). This is further in line with Clark’s (2000) conceptualization of boundary management, as while she argues that individuals play an active role in constructing borders between their work and non- work spheres in line with their individual preferences, she acknowledges that they are only able to enact this active role to the extent that the context allows. Recent research has built upon this differentiation of boundary management preferences from enacted boundary management, although important nuance may be needed to fully understand the phenomena at play. For instance, Ammons (2013) conducted a series of interviews aimed at understanding the alignment between workers’ boundary management preferences and enacted boundaries, finding that men and parents of young children have better alignment as compared to women and those without caregiving duties. While an important start to our understanding of these related constructs, the limited sample size (N = 23) and focus on qualitative analyses allows for room for quantitative verification. For studies that have taken a more quantitative approach to the relationship between preferred and enacted boundaries, the conceptualization and examination has been almost exclusively linear (i.e. preferences as predictive of enactment). For example, Carlson and colleagues (2016) evaluated the extent to which boundary management preferences predicted four proposed boundary management tactics, largely finding support for the proposed associations. Similarly, Powell and Greenhaus (2010) displayed a significant relationship between preferred and actual segmentation between the work and family domain, as a part of 14 their larger examination of gender differences in the work-to-family interface. While the mentioned works are useful demonstrations of this linear relationship, limiting the conceptualization to the linear may not fully represent what is suggested for these variables by theory. Accordingly, via the examination of the interaction between preferred and enacted boundary management, I aim to bring forth attention to the fit or congruence between these two variables as important for worker experience. Further, although the support found for the relationship between preferred and enacted boundary management is a useful step toward understanding differences in the manners in which individuals segment or integrate their respective life spheres, a more nuanced path forward would further underscore circumstances in which these relationships are particularly strong versus weak. Indeed, though multiple theoretical frames have discussed situational factors constraining the link between preferences and enacted boundaries (e.g. Nippert-Eng, 1996; Clark, 2000), this has yet to be adequately empirically examined. Accordingly, the present study puts forth multiple work-relevant constraints potentially inhibiting the degree to which employees may enact boundaries in line with their preferences (e.g. freedom to make decisions and work structure as indicators of work autonomy). Even further, the current research puts forth a unique constraint faced by employees with invisible stigmatized identities: their identity management at work. Below I present logic outlining how and why identity management may be an important and presently underexamined antecedent of enacted boundary management for stigmatized workers. Linking Identity Management & Boundary Management In considering this review of boundary management, I suggest that employees with stigmatized identities may be limited in the manner in which they feel comfortable constructing 15 and enacting boundaries between the work and non-work interface; to the extent that employing integrative boundary management tactics may expose themselves to stigma at work, workers of such identities may feel more inclined to segment (or separate) their respective life spheres, irrespective of their ideal boundary management preferences. Such logic has previously been applied to those in family-unfriendly cultures (Desrochers, Sarget & Hostetler, 2012). Specifically, Desrochers and colleagues (2012) suggest that organizations can differ in the extent to which employee family needs are appropriately responded to within company policy and practice. They described family-friendly workplaces as characterized by those in which work boundaries are permeable to family domain demands; this is in contrast to family-unfriendly organizations, in which employees are implicitly encouraged to segment family from work and are discouraged from bringing family issues into the workplace. The authors note the one-way nature of this encouraged segmentation, in that while employees are implicitly kept from bringing life matters to work, work-to-family boundary management may be more integrative in nature, such that employees are encouraged (or at the very least, not discouraged) to continue addressing work-relevant matters within their home life. I extend this logic to apply to stigma surrounding minority sexual orientation. To the extent that employees of such identities anticipate facing stigma within their organization as a result of their sexual orientation, perhaps as a result of a weak diversity climate (in line with Desrochers and colleagues’ discussion of family-unfriendly cultures) or previous experiences with stigma elsewhere, those employees may find themselves circumstantially more likely to segment their non-work lives from their work lives. However, as described by Desrochers and colleagues (2012), this is likely to primarily affect the permeability of the boundary separating life from work, while leaving the permeability of the boundary between work to non-work less 16 constrained and potentially more aligned with individual preferences. Such a distinction is in line with our understanding of identity management broadly; given that the current focus is on an individual’s management of a stigmatized life characteristic at work, it then follows that the life- to-work barrier would be more likely to be impacted by identity management concerns as compared to the work-to-life barrier (assuming that the individuals’ work context is not stigmatized within their life sphere and the employee is subsequently more free to bring their work life home without fear of prejudice). Said another way, a sexual minority employee may feel that they are unable to talk about their home life when at work given the potential stigma they may experience from doing so, but comparatively feel no constraints in talking about work at home as they are not facing any stigma from doing so in that sphere. Acknowledging this review, I now outline the hypotheses I examined over two studies. In my first study, I investigated the relationship between LGB employees’ identity management and subsequent enacted boundary management; in Study 2, I compared heterosexual employees’ alignment between their preferred and enacted boundary management with LGB employees’ alignment, exploring how this potential differential congruence may relate to work-relevant attitudinal outcomes. 17 CHAPTER 4: HYPOTHESES Below I detail the logic surrounding each of my nine hypotheses, as based on the previously reviewed literature. Note that Hypotheses 1 through 7 are tested via a sample of LGB employees, while Hypotheses 8 and 9 are tested in a separate study utilizing a sample of both heterosexual and sexual minority workers. The two-study design was pursued as the construct of identity management is exclusively conceptualized for members of stigmatized groups, so to test a model including identity management for heterosexual workers would be inappropriate; if indeed the concept of identity management was applied to heterosexual identity, we would likely find little variance in management tactics given the lack of stigma associated with the identity in question (i.e. most would likely “affirm” to some degree, as there are no stigma-related costs to doing so). Accordingly, the current dissertation seeks to understand boundary management as predicted by identity management via a within-group examination of LGB workers (Study 1), and a between-group examination of LGB workers compared to heterosexual workers (Study 2). In this way, I sought to both demonstrate the unique constraints faced by those of sexual minority identities at work while further underscoring how this constraint may lead to worsened outcomes compared to those not faced with the concern. Figure 2 represents the hypotheses tested in Study 1, while Figure 3 demonstrates the hypotheses tested in Study 2. Study 1 As suggested by both boundary theory (Ashforth et al., 2000) and border theory (Clark, 2000), individuals vary in the degree to which they either segment or integrate their respective spheres (Nippert-Eng, 1996). Clark’s (2000) active conceptualization suggests that individuals can intentionally shape roles in line with their preferences for their personal conceptualization of 18 work-life balance, a notion broadly in line with the overall philosophy of differential psychology. Indeed, previous works have demonstrated a link between boundary management preferences and enacted boundaries (e.g. Powell & Greenhaus, 2010; Carlson et al., 2016), so to this end I hypothesized a relationship between one’s boundary management preferences and one’s enacted boundary management at work: H1: Individual segmentation preferences will relate to enacted boundary management, such that those with preferences for segmentation will enact boundary segmentation more so than those with preferences for integration, who will enact greater boundary integration. However, while individual preferences are indeed expected to influence the manner in which one manages their work and non-work boundaries, it is likely that individuals are not always able to perfectly enact their preferences given their respective contexts (Mellner, Aronsson & Kecklund, 2014). If given full discretion without any imposition, it is expected that one would likely enact boundary management behaviors aligned with their preferences; however, as individuals operate in circumstances with their own respective norms and constraining qualities, more often than not individuals are limited in the full range of behaviors they are practically able to enact. Such assertions are in line with conclusions drawn from the person- situation debate, suggesting that the extent to which individual differences predict behavior in a given context depends on the strength of that context (Judge & Zapata, 2015). That is, in strong situations characterized by rules, structure, and cues, individuals are thought to be less likely to impose individual discretion as compared to when they are in weak situations, involving considerably less structure (Mischel, 1977). Therefore, I predicted that the relationship between 19 boundary management preferences and enacted boundary management will depend on an individual’s work circumstances. A pivotal work characteristic potentially informative of the extent to which one may behave in line with one’s preferences is the level of relative freedom one experiences at work, also known as job autonomy. Identified by Hackman and Oldham as a core job characteristic deterministic of job satisfaction for all workers (Hackman & Oldham, 1976), job autonomy has been defined as the degree to which the job provides substantial freedom, independence, and discretion to the employee in both scheduling work and in determining procedures to be used in carrying out the work (Hackman & Oldham, 1975). It is likely that the greater job autonomy one experiences at work, the more likely one is to act in line with their overall preferences and desires. Two indicators of present focus will be used to represent the degree to which one’s job allows personal autonomy: (1) freedom to make decisions (i.e. how much decision-making freedom, without supervision, does the job offer?), and (2) level of work structure (i.e. to what extent is this job structured for the worker, rather than allowing the worker to determine tasks, priorities and goals?). One would expect that when individuals are afforded full freedom to decide how to approach and conduct their work, so too are they able to determine the manner in which they manage their work and non-work spheres (Clark, 2000). For example, an individual who has full freedom to dictate their work schedule may be particularly able to schedule their calendar to accommodate their life circumstances in line with their preferences. In contrast, those with little freedom to make work-relevant decisions are more likely to be constrained by their environment, and therefore less likely to enact boundary management as aligned with their preferences given the decreased ability to shape their work circumstances. Similar predictions 20 can be made when considering one’s level of work structure; as structure is one of the core determinants of situational strength as suggested by Mischel (1977), one would expect that those in structured work environments would be more constrained in what behaviors they are able to enact than are those in unstructured work environments, as they are afforded less personal discretion given stricter environmental expectations-to the extent then that lessened structure affords individual ability to shape work potentially in line with one’s boundary management preferences, I hypothesized: H2: The relationship between segmentation preferences and enacted boundary management will depend one’s level of work autonomy, such that the relationship will be stronger for those with greater (as compared to lesser) autonomy at work. Acknowledging that individuals are not always able to manage work and non-work boundaries according to their preferences as a result of environmental and contextual constraints, I propose that individuals of invisible, stigmatized identities face an additional constraint that may influence how they enact boundary management: their identity management at work. That is, as compared to individuals without invisible identities in need of stigma management, I suggest that individuals of such identities face supplementary circumstances that may influence how they manage their work-life boundaries, as a function of the degree to which they affirm or distance from their identity at work. As identity management largely concerns the degree to which individuals seek to make their stigmatized identity known and salient, the extent that boundary management behaviors openly reflect one’s standing on a stigmatized characteristic may inform which behaviors one seeks to engage. Given that integration and increased boundary crossover can be associated with information from one sphere being revealed or becoming increasingly salient to individuals of 21 another sphere (e.g. talking to coworkers about your weekend, talking to your significant other about your boss; Desrochers et al., 2012; Sawyer et al., 2017), it follows that individuals of stigmatized identities may be limited in what boundary management behaviors they seek to adopt, as dependent on the degree to which they seek to make their stigmatized identity known at work. For example, integration of work and life spheres would theoretically allow for ample crossover between work and life domains (Ashforth et al., 2000; Clark, 2000); perhaps then an individual who is seeking to keep characteristics associated with their home life (e.g. sexual orientation) concealed from coworkers would avoid such boundary management tactics so as to maintain privacy and avoid stigma. In contrast, an individual engaging in affirming identity management may feel more inclined to manage their boundaries in line with their boundary management preferences, as they would not share this same fear of maintaining their identity as private. Therefore, I predict that one’s identity management may inform one’s work-life boundary management: to the extent that an individual engages in identity distancing strategies at work, so too may they potentially segment their work and life spheres so as to reinforce their identity concealment. Note that my conceptualization is concerned with boundary enactment, and not boundary management preferences; I suggest that the relationship between identity management and enacted boundaries occurs irrespective of boundary management preferences, such that individuals may enact boundaries influenced by distancing identity management concerns in spite of their preferred and ideal manner of work and non-work boundary management. Therefore, I predicted: 22 H3: Engagement in (a) distancing identity management will significantly and positively relate to enacted boundary segmentation, whereas (b) engagement in affirming identity management will be unrelated to enacted boundary segmentation. However, as suggested by Desrochers and colleagues (2012) with respect to family- unfriendly cultures, such management of stigma is more likely to affect the degree to which one integrates their life into their work, rather than the degree to which work is integrated into the home sphere. Given the directionality of the stigma (i.e. life is stigmatized at work, but work is not stigmatized at home), one might expect that the constraint of identity management most aptly apply to circumstances in which employees are faced with bringing life elements to work. Accordingly, I suggested that the strength of the relationship between an individual’s identity management and their subsequent enacted boundary management will depend on the direction of the boundary of consideration: H4: The strength of the relationship between distancing identity management and enacted boundary segmentation will differ depending on the direction of the boundary, such that the relationship will be stronger for life-to-work segmentation, and weaker for work-to- life segmentation. Acknowledging my prediction that identity management may inform boundary enactment in spite of preferred boundary management tactics, the possibility of misalignment between boundary management preferences and enacted boundaries is both possible and likely. To the extent then that individuals’ boundary management is misaligned with their boundary management preferences, I suggest that individuals may be more likely to experience negative work-relevant outcomes. One may reference the fit literature to support this notion. Broadly, research on fit concerns the effect of congruence between one’s desires and/or preferences with 23 some other entity (e.g. one’s organization, one’s workgroup, one’s boss). The alignment (or, misalignment) between one’s preferences and one’s organizational reality has been meta- analytically linked to outcomes including job satisfaction, organizational commitment, and turnover intent (Verquer, Beehr & Wagner, 2003), in line with the theoretical notion that fit can provide individual need fulfillment and subsequent satisfaction (Cable & Edwards, 2004). More specific to the work-life literature, Rothbard, Phillips and Dumas (2005) found that congruence between individual segmentation preferences and access to policies that enable their desired boundary management was linked to both job satisfaction and organizational commitment. Along the same lines, I predict that enacted boundary management can be linked to worsened individual affective outcomes when that enacted management is misaligned with boundary management preferences. That is, given literature suggesting that fit between preferences and reality can provide need fulfillment that misfit cannot provide (Cable & Edwards, 2004), I suggest that individuals who are enacting behaviors incongruent with their desires are more likely to exhibit lower wellbeing, greater withdrawal, and lower job satisfaction as compared to employees enacting behaviors in line with their preferences. Therefore, I hypothesized: H5: When enacted segmentation is high (low) while segmentation preferences are low (high), enacted segmentation will (a) negatively relate to subjective wellbeing, (b) positively relate to withdrawal, and (c) negatively relate to job satisfaction; in contrast, when enacted segmentation is high (low) and segmentation preferences are high (low), enacted segmentation should display a (a) positive relationship with subjective wellbeing, (b) a negative relationship with withdrawal, and (c) a positive relationship with job satisfaction. 24 Finally, I considered the role of perceived organizational diversity climate both as a determinant of identity management, as well as a direct antecedent of work-relevant outcomes for employees of hidden and stigmatized identities. Broadly, diversity climate perceptions can be understood as individual views of an organization’s value for diversity, as well as perceptions concerning that organization’s approach to diversity management (Roberson, 2012). With respect to identity management, past work generally supports the notion that individuals manage their stigmatized identities differently depending on the degree to which they perceive their organization to be accepting (Clair et al., 2005; Chrobot-Mason, Button & DiClementi, 2001; Button, 2001; King, Reilly & Hebl, 2008). Specifically, perceptions of a strong diversity climate may communicate to individuals that their disclosure will not be reacted to negatively (i.e. identity safety), and therefore individuals may perceive fewer costs associated with implementation of affirming identity management tactics as compared to individuals in organizations with weaker diversity climates, who may perceive greater risk (Clair et al., 2005). These notions have largely been supported empirically. For instance, Clair and colleagues (2005) identified diversity climate perceptions as a crucial contextual condition associated with an individual’s decision to pass or reveal their stigmatized identity. Further, Button (2001) found that positive organizational climate for sexual minorities was associated with less frequent employment of counterfeiting and avoiding identity management strategies. Finally, Chrobot- Mason and colleagues (2001) found that lesbian and gay workers were more likely to engage in disclosure-based identity management strategies when they perceived their organizational climate as affirming. Consequently, I hypothesized a relationship between individual perceptions of diversity climate and subsequent identity management: 25 H6: Perceptions of organizational diversity climate will relate to worker’s identity management, such that diversity climate will (a) positively relate to affirming identity management and (b) negatively relate to distancing identity management. Beyond affecting identity management, I suggest that perceived diversity climate can have a direct impact on individual work-relevant attitudes, including subjective wellbeing, work withdrawal, and job satisfaction. Past work has demonstrated relationships between diversity climate perceptions and individual-level outcomes (e.g. Mor Barak et al., 2016 for a meta- analysis; Stewart, Volpone, Avery & McKay, 2011; McKay, Avery, Tonidandel, Morris, Hernandez & Hebl, 2007), with associations often explained via psychological contract or person-fit perspectives (Kristof-Brown, Jansen, & Colbert, 2002; Robinson & Rosseau, 1994). More specifically, individuals (particularly those of stigmatized identities) are thought to experience better fit with organizations when diversity climate perceptions are high rather than low, given perceptions that they are valued by the organization. Alternatively, when stigmatized employees perceive a weak diversity climate, they have demonstrated greater turnover intentions (Stewart et al., 2011; McKay et al., 2007), lower job satisfaction (Driscoll, Kelley, & Fassinger, 1996; Madera, Dawson, & Neal, 2013), and greater occupational stress (Driscoll et al., 1996), likely due to perceived misfit, feelings of exclusion, and organizational mistrust (Mor Barak et al., 2016). Accordingly, I hypothesized similar associations within the present study: H7: Perceptions of organizational diversity climate will (a) positively relate to subjective wellbeing, (b) negatively relate to withdrawal, and (c) positively relate to job satisfaction. Study 2 While Study 1 attempted to take a within-group approach to understanding the link between identity management and enacted boundary management for LGB employees, Study 2 26 sought to display how these processes may lead to disparate outcomes for LGB employees as compared to heterosexual employees. Broadly, in considering the link between boundary management preferences and enacted boundary management, I have argued that workplace contextual circumstances (i.e. freedom to make decisions and work structure as indicative of job autonomy) may moderate the strength of this relationship to the extent that context constrains one’s ability to enact preferences. Assuming that individuals’ work constraints are equivalent across sexual orientations (i.e. the jobs that heterosexual employees hold are similarly constraining to the jobs that LGB employees hold; though, some evidence does suggest some occupational differentiation by sexual orientation challenging this assumption, Tilcsik et al., 2015), I suggested that LGB employees may be further constrained via their identity management than are heterosexual employees without this concern. Considering the notion that additional constraints limit the extent to which individuals are able to enact boundaries in line with their preferences, I suggested that LGB employees may see less congruence between their preferred and enacted boundary management than do heterosexual employees (see Figure 3): H8: Heterosexual employees will experience greater congruence between their preferred and enacted boundary management than will lesbian, gay or bisexual employees. Finally, similar to the logic presented for Hypothesis 5, I suggest that the degree to which boundary management preferences and enacted boundaries are aligned may relate to important work-relevant attitudes, including subjective wellbeing, withdrawal, and job satisfaction. From a fit perspective, congruence between preferences and behaviors may provide greater need fulfillment than can misfit (Cable & Edwards, 2004), with meta-analytic evidence supporting this notion of alignment as predictive of job satisfaction, organizational commitment and turnover intent (Verquer et al., 2003). Accordingly, I predict the following: 27 H9: Employees’ degree of congruence between their preferred and enacted boundary management will relate to work-relevant attitudinal outcomes, such that greater congruence will be associated with (a) greater subjective wellbeing, (b) less withdrawal, and (c) greater job satisfaction. While seemingly similar to what is predicted in H5, H9 employs a slightly different approach via a more global consideration of congruence as a separate concrete variable itself predictive of outcomes, whereas H5 examines multiple specific and hypothesized intersections of preferred and enacted boundary management as related to work-relevant attitudes. Therefore, the examination of both H5 and H9, although highly related, provides slightly different yet complementary perspectives to the same general notion of congruence as it potentially relates to important work-relevant attitudinal outcomes. 28 CHAPTER 5: STUDY 1 METHOD Procedure Study 1 was a multi-wave survey across two time points. At Time 1, respondents were surveyed regarding their identity management tactics, enacted and preferred boundary management styles, perceptions of their organization’s diversity climate, job title, and autonomy perceptions. At Time 2 (one week later), participants provided responses to the outcome variables (i.e. subjective wellbeing, withdrawal, job satisfaction). Job context variable scores (i.e. freedom to make decisions and level of work structure) were derived from the Occupational Information Network (i.e. O*NET, a free online database that is the United States’ primary source of occupational information), as linked via the participant’s job title provided at Time 1. Participants A total of 494 participants were recruited through the Qualtrics Panel Service to complete this study’s first survey. Research on samples recruited on similar platforms have found participants to be more attentive (Hauser & Schwarz, 2016) and more diverse than those recruited through traditional student subject pools (Buhrmester, Kwang, & Gosling, 2011). Further, these recruitment methods have been identified as particularly useful for reaching disadvantaged or marginalized populations (Smith, Sabat, Martinez, Weaver & Xu, 2015); accordingly, this sampling strategy appears appropriate for the present research. Inclusionary criteria were identification as either gay, lesbian, or bisexual, current employment at least part- time in non-self-employed circumstances, residence in the United States, and correct selection of response options within two attention checks (i.e. “Please select Disagree/Slightly Disagree”). Participants who provided non-sensical qualitative responses (N = 12, 2.4%) were removed, 29 leaving a final Time 1 sample of 482 participants. Of these 482 participants, gay workers comprised the largest sexual orientation subgroup (48.7%), followed by bisexual (38.0%) and lesbian (13.2%) participants. Each of the 482 Time 1 participants received an invitation to complete the Time 2 survey one week after completion of the Time 1 survey. Of the 482 Time 1 participants eligible for Time 2, 225 participated in the second survey (46.7%). Those reporting sexual orientation at Time 2 differing for that reported at Time 1 were removed from analyses (N = 9, 4.0%), leaving a final two-time-point sample of 216 LGB employees. Of this final sample, 53.7% identified as gay, 31.5% as bisexual, and 14.8% as lesbian. Participant gender breakdown was 69.0% male, 30.6% female and 0.5% non-binary. Participants were on average 50.01 years old (SD = 14.08), working on average 37.93 hours per week (SD = 11.56), and employed at their current organization for an average of 12.87 years (SD = 11.04). The majority of participants reported working outside of the home (57.4%), with 42.3% working from the home (note this data was collected in January and February 2021 during the COVID-19 pandemic). Racial/ethnic breakdown for this sample was the following (participants could self-identify as multiple options): 87.5% White, 6.0% Black/African American, 3.2% East Asian, 3.2% Latinx, 2.3% South Asian, 0.5% Native Hawaiian/Pacific Islander, 0.5% American Indian/Alaska Native, and 1.4% “Other racial identity.” Participants were compensated monetarily directly via Qualtrics panels, with participant acquisition costs of $5.50 per completed survey. 30 Measures (Time 1) Identity Management The extent to which participants engaged in distancing and affirming strategies was assessed via scales from Button (2004). Both measures were assessed using a seven-point Likert- type scale (1 = “Strongly Disagree” to 7 = “Strongly Agree”). The affirming subscale was comprised of ten items (α = .90), an example of which is “Most of my coworkers know that I am gay.” The distancing subscale was comprised of six items (α = .87), an example of which is “I avoid situations where heterosexual coworkers are likely to ask me personal questions.” It is important to note that while distancing and affirming identity management have been conceptualized as opposite ends of the same continuum (as has been presented in each of the three reviewed frameworks), most often for measurement purposes these strategies have been assessed independently as unique scales (as have been developed here by Button; Lyons et al., 2014 as an additional example). Further supporting this notion of separation, confirmatory factor analyses of a two-factor and single-factor structure showed superior fit when treating affirming and distancing items as separate factors (One-factor fit: RMSEA = 0.176, CFI = 0.632, TLI = 0.576, SRMR = 0.136; Two-factor fit: RMSEA = 0.109, CFI = 0.860, TLI = 0.837, SRMR = 0.071; accordingly, this differentiated approach treating the two strategies as separate variables was employed in the present study. Boundary Management (Enacted and Preferred) Perhaps the most widely used measure assessing segmentation preferences is the four- item scale created by Kreiner (2006). In line with Powell & Greenhaus (2010), I adapted this measure to fit the purposes of the present study, differentiating between enacted and preferred boundary management, as well as the directionality of the boundaries in question. Kreiner’s 31 original four-item measure of boundary management preferences displayed strong internal consistency, but focused exclusively on the management of work to non-work contexts. To more fully capture the full spectrum of boundary management preferences by including the management of non-work to work contexts, I adapted and added an additional four items based on Kreiner’s (2006) original preferences scale, for a final eight-item measure assessed via a seven-point Likert-type scale (1 = “Strongly Disagree” to 7 = “Strongly Agree”). An example item is, “I prefer to keep work life at work (α = .88).” I followed Powell and Greenhaus (2010) in further adapting this measure to assess enacted boundary management. By removing all references to “preferences” and “liking” within each item, I was left with eight items that describe the manner in which individuals truly manage their boundaries. An example of item adaptation is “I prefer to keep work life at work (segmentation preferences)” to “I keep work life at work (enacted segmentation).” When transforming Kreiner’s (2006) original four items in this way within their study, Powell and Greenhaus (2010) conducted a confirmatory factor analysis which ultimately provided evidence that a multiple-factor structure that differentiated between preferences and enacted segmentation was appropriate. This eight-item adapted scale of enacted boundary management was assessed via a seven-point Likert-type scale (1 = “Strongly Disagree” to 7 = “Strongly Agree”; α = .93). To confirm the dimensionality of these measures, I conducted multiple confirmatory factor analyses to understand model fit when these 16 items are modeled as 1) a single-factor, 2) a two-factor model differentiating between preferred and enacted boundary management, and 3) a four-factor model differentiating between preferred and enacted boundaries across both the work to life and life to work directions. It was this four-factor model that displayed the best fit (RMSEA = 0.101, CFI = 0.927, TLI = 0.911, SRMR = 0.050), with the single-factor model 32 (RMSEA = 0.251, CFI = 0.522, TLI = 0.449, SRMR = 0.159) and the two-factor model (RMSEA = 0.204, CFI = 0.687, TLI = 0.635, SRMR = 0.152) each displaying poor model fit; therefore, this four-factor model differentiating 1) preferred life to work boundaries, 2) preferred work to life boundaries, 3) enacted life to work boundaries, and 4) enacted work to life boundaries was supported here. Diversity Climate Perceptions Participant perceptions of their organization’s diversity climate was captured via a four- item measure created by McKay, Avery and Morris (2008). Assessed via a five-point Likert-type scale (1 = “Strongly Disagree” to 5 = “Strongly Agree”), this measure displayed strong internal consistency (α = .89). An example item is, “Top leaders demonstrate a visible commitment to diversity.” Job Title & Work Autonomy Participants reported their current job title via a dropdown menu of 353 job options derived from O*NET. The selected job title was used to obtain occupation scores of “freedom to make decisions” and “level of work structure” (both as indicators of overall work autonomy) as sourced from O*NET. Participants’ freedom to make decisions at work ranged from 0 (i.e. “No freedom”) to 100 (“A lot of freedom”), and were developed based on responses of sampled workers within that particular occupation. Similarly, participants’ level of work structure also ranged from 0 (i.e. “No freedom”) to 100 (“A lot of freedom”), and were also developed based on responses of sampled workers within that particular occupation. As these two indicators displayed a strong relationship with one another (r = .82, p < .001), I combined them into a single variable representing work autonomy as sourced from O*NET (α = .90). To ensure these indicators indeed related to participants’ own perceptions of their work autonomy, participants 33 completed Thompson and Prottas’ (2005) four-item job autonomy scale, a five-point Likert-type measure with an example item of “I have the freedom to decide what I do on my job (α = .81).” Demographic Variables Various demographic variables were assessed at Time 1, including participant gender, race, sexual orientation, age, number of years employed at current organization, remote work status (i.e. currently working from home vs. currently working outside of the home), and the degree to which the COVID-19 pandemic impacted the manner in which participants conducted their work (five-point Likert-type scale; 1 = “No impact” to 5 = “Extreme Impact”). Note that results were analyzed both with and without level of COVID-19 impact as a control variable, though results are presented here without the variable included as results did not differ. Remote work status was found to correlate with both autonomy operationalizations, such that working from home was associated with higher autonomy, though controlling of this variable additionally did not alter results. Measures (Time 2) Subjective Wellbeing Subjective wellbeing was measured via Diener, Emmons, Larsen and Griffin’s (1985) Satisfaction with Life Scale. This five-item measure was assessed using a seven-point Likert- type scale, (1 = “Strongly Disagree,” to 7 = “Strongly Agree), has displayed strong internal consistency in past usage (e.g. α = .83, Pavot, Diener, Colvin & Sandvick, 1991), and was similarly reliable in the present study (α = .92). An example item is, “In most ways, my life is close to ideal.” 34 Withdrawal Participants completed a 13-item measure from Hanisch and Hulin (1990), in which they indicated the frequency with which they engage in each of the 13 behaviors on a five-point, Likert-type scale (1 = “Never” to 5 = “Always”). Past usage of this scale has displayed strong internal consistency (e.g. α = .85, Boswell & Olson-Buchanan, 2004), with similarly reliability noted here (α = .91). An example item is, “Wanted to leave work early.” Job Satisfaction Participant job satisfaction was assessed via the three-item Overall Job Satisfaction scale (Cammann, Fichman, Jenkins & Klesh, 1983). This scale demonstrated strong internal consistency in the current study (α = .92), as had been seen in previous use of the measure. (e.g. α = .88, Shaw, 1999; α = .90, Harari, Thompson & Viswesvaran, 2018). An example item is, “In general, I like working at my job.” 35 CHAPTER 6: STUDY 1 RESULTS Table 1 provides the descriptive statistics and intercorrelations between study variables and participant demographics. Note that higher scores on both boundary management variables indicate greater preferred/enacted segmentation, with lower scores indicating greater boundary integration. In examining Table 1, distancing identity management appears to be positively related to both boundary management preferences and enacted boundary management, such that those higher in this identity management strategy are reporting higher desired and enacted segmentation between work and life spheres. In contrast, affirming identity management appears unrelated to either boundary management preferences or enacted boundary management. Unexpectedly, the two operationalizations of work autonomy (through O*NET and through self- report) are not significantly correlated. Participant demographics were largely uncorrelated with study focal variables, though White participants did report higher distancing identity management and work autonomy operationalized via O*NET than did non-White participants- additionally, age correlated with the three study outcomes, such that older workers demonstrated higher job satisfaction, higher subjective wellbeing, and lower withdrawal than did younger workers, consistent with meta-analytic evidence linking age with job attitudes (Ng & Feldman, 2010). Hypothesis Testing Hypotheses 1 and 2 outline the predicted relationship between boundary management preferences and enacted boundary management (H1), as well as the proposed moderation of this relationship as dependent on level of work autonomy (H2). As the work autonomy scores operationalized through O*NET were not significantly correlated with self-reported autonomy as 36 expected, I present results testing this hypothesis using both variable operationalizations. Hierarchical regressions with centered main effects terms in the first step, and an interaction term created by multiplying the centered main effect terms in the second step, were used to evaluate these hypotheses. Across both operationalizations of work autonomy (Tables 2 and 3), boundary management preferences were a significant predictor of enacted boundary management at all model steps, supporting H1. However, no evidence was found to support the proposed moderation, as across both operationalizations neither the main effect of work autonomy nor the interaction term of work autonomy with preferences were significant (Tables 2 and 3). Therefore, H1 was supported, while H2 was not supported. Hypotheses 3 proposed a relationship between identity management and enacted boundary management, such that those engaging in distancing boundary management strategies would be more likely to segment their life spheres (a), while affirming identity management would be unrelated to enacted segmentation (b). When identity management strategies are examined as individual predictors of enacted boundary management (Table 4), we find, in line with expectations, distancing identity management has a significant positive relationship with enacted boundary management, such that those engaging in the strategy are more likely to segment their life spheres; in contrast, affirming identity management does not appear related to enacted boundary management- this pattern of relationships suggests support for H3a and H3b. However, when examined within a multiple regression of both identity management strategies included as predictors of enacted boundary management, both strategies demonstrate significant and positive relationships with the outcome (Table 5). Acknowledging that affirming identity management is both uncorrelated with enacted boundary management in Table 1 and shows no direct relationship with the outcome when examined as a single predictor (Table 4), I 37 suspect the presence of a suppressor effect in this multiple regression analyses. Tzelgov and Henik (1991) define a suppression situation as one in which the prediction of a dependent variable is improved via the addition of a predictor that is uncorrelated with the outcome but correlated to another predictor in the model, resulting in increases in either the predictive value of the previously unrelated variable or overall greater variance explained in the model. Such a definition seems to describe the current situation, as affirming identity management appears uncorrelated to enacted boundary management, while significantly and negatively correlated to distancing boundary management (Table 1)- this correlation likely reflects some element of shared variance between the two identity management predictors, perhaps characterized as one’s awareness of their identity at work. Therefore, putting aside the multiple regression results as influenced by a potential suppressor effect, individual regression results of each boundary management strategy alone predicting enacted boundary management supports the assertions of H3a and H3b. Hypothesis 4 suggests that this relationship between distancing identity management and enacted boundary management will depend on the direction of the boundary, such that the relationship will be stronger for life-to-work segmentation, and weaker for work-to-life segmentation. Because the proposed moderator (i.e. boundary directionality) is inherently linked to measurement of the dependent variable (i.e. four items measure work-to-life boundary enactment, while a different set of four items measure life-to-work boundary enactment), using a traditional hierarchical regression to examine this moderation is not possible here. Instead, I conducted two linear regressions of 1) distancing identity management predicting enacted life-to- work segmentation, and 2) distancing identity management predicting enacted work-to-life segmentation (Table 6). 38 Both analyses demonstrate significant and positive relationships between predictor and outcome, though whether the magnitudes of the relationships significantly differ cannot be determined here. To accordingly test this notion, I estimated the linear regressions’ corresponding 95% confidence intervals via bias corrected bootstrapping (1000 re-samples; life- to-work 95% CI = [0.08, 0.38]; work-to-life 95% CI = [0.02, 0.34]). Cumming (2009) suggests that confidence intervals that overlap by less than 50% are considered significantly different from one another at p < .05; as the upper bound of the work-to-life weight (.34) exceeded the sum of the averaged confidence interval overlap (.08) with the life-to-work lower bound estimate (.08), the confidence intervals are considered to overlap by greater than 50% and accordingly do not significantly differ. Therefore, H4 is not supported, as the strengths of the relationships between distancing identity management and enacted boundary management is of comparable magnitude in both directions (i.e., work-to-life and life-to-work). Hypothesis 5 outlined a proposed interactive effect of enacted boundary management and boundary management preferences on the three examined outcomes (i.e., subjective wellbeing, job satisfaction and withdrawal). A series of hierarchical regressions were conducted with centered main effect terms in the first step, and an interaction term created by multiplying the centered main effect terms in the second step, to evaluate these hypotheses. Looking to the interaction terms in each of the model results (Tables 7, 8 and 9), we find little support for the hypothesis, as none of the terms are significant at the traditional p < .05 criterion. Therefore, H5 remains largely unsupported. Though the proposed interactions described in H5 were largely not found here, some non- hypothesized main effects emerged as predictive of our respective outcomes. Across both subjective wellbeing (Table 7) and withdrawal (Table 9), direct effects of enacted boundary 39 management are observed across both levels of the model, such that greater enacted segmentation is associated with higher subjective wellbeing and lower withdrawal from work. Additionally, boundary management preferences emerged as a significant predictor within the first model step for both job satisfaction and withdrawal, such that higher segmentation preferences were associated with lower job satisfaction (Table 8) and greater withdrawal from work (Table 9). Hypothesis 6 suggest that perceptions of organizational diversity climate will relate to worker identity management, such that it will be (a) positively related to affirming identity management and (b) negatively related to distancing identity management. Accordingly, I conducted two linear regressions to test this hypothesis. In line with expectations, diversity climate was significantly and positively predictive of affirming identity management, and negatively predictive of distancing identity management (Table 10). Therefore, H6 was supported. Finally, Hypothesis 7 suggests that diversity climate perceptions should directly and positively predict (a) subjective wellbeing and (b) job satisfaction, and (c) negatively predict work withdrawal. Table 11 demonstrates the results of each of three linear regressions testing these notions, showing complete support. Therefore, H7 was supported here. Mediational Analyses Acknowledging the supported links between diversity climate perceptions, distancing identity management, and enacted boundary management, I tested whether distancing identity management mediated the relationship between diversity climate perceptions and enacted boundary management using Hayes’ (2013) PROCESS Macro model 4 (see Table 12). The overall mediation model was significant [R2 = .03, MSE = 2.04, F(1, 214) = 6.07, p = .014], as 40 was the confidence interval of the indirect effect calculated via bootstrapping of 5000 samples, b = -.08, SE = .04, 95%CI [-.18, -.01]. Overall, these results suggest that diversity climate is both directly and indirectly related to enacted boundary management, as partially mediated through distancing identity management. Further, building upon the links between distancing identity management, enacted boundary management, and subjective wellbeing and withdrawal, I additionally tested whether enacted boundary management mediated these relationships between distancing identity management and outcomes. Again using Hayes’ (2013) PROCESS Macro 4 (see Table 13), I find that the overall mediation model was significant when predicting subjective wellbeing [R2 = .05, MSE = 2.02, F(1, 213) = 10.81, p = .001], as was the confidence interval of the indirect effect calculated via bootstrapping of 5000 samples, b = .03, SE = .01, 95%CI [.003, .06], suggesting that indeed, distancing identity management is related to subjective wellbeing through enacted boundary management. Similar results were found when considering withdrawal as the outcome (Table 14); both the overall mediation model [R2 = .05, MSE = 2.02, F(1, 213) = 10.81, p = .001] as well as the indirect effect (b = -.01, SE = .01, 95%CI [-.04, -.001]. Therefore, enacted boundary management appears to mediate the effects of distancing identity management on both subjective wellbeing and withdrawal. Exploratory Analysis Though H5 received limited support when examined via hierarchical regression, I examined whether congruence between boundary management preferences and enacted boundary management operationalized via polynomial regression yielded similar results (see page 54 for greater description of this analytic technique). Upon estimating congruence terms for each of the three outcome variables (i.e. subjective wellbeing, job satisfaction and withdrawal), I 41 ran three linear regressions with the congruence term predicting its respective outcome (Table 15). Contrary to the results presented via hierarchical regression, results using polynomial regression demonstrate that boundary management congruence significantly (positively in the case of subjective wellbeing and job satisfaction, and negatively in the case of withdrawal) predicts each of the examined outcomes. Discussion The purpose of Study 1 was to take a within-group approach toward the examination of boundary management in LGB workers, specifically examining the relevance of theoretically informed antecedents and outcomes of the enacted boundaries between work and home lives. While support for hypotheses here was mixed (particularly limited for the proposed moderation effects), there was indeed evidence found for factors that contribute to enacted boundary management (including the novel link demonstrated between identity management and enacted boundary management), as well as outcomes related to these managed boundaries for sexual minority workers. Concerning antecedents of enacted boundaries, both boundary management preferences and identity management emerged as predictors of the construct. Against expectations, work autonomy was not found to moderate the relationship between boundary management preferences and enacted boundary management, as would have been predicted via situational strength theory (Mischel, 1977). In interpreting this finding, it is perhaps possible that boundary management preferences are such a strong individual difference that they are predictive of enacted boundary management across most situations; alternatively, it is possible that the lack of effect occurred given issues related to the measurement of work autonomy. Finally, it is possible that the theory here was mis-specified, in that perhaps the target of autonomy needs to be of 42 consideration; that is, perhaps autonomy over time is what is most important toward boundary management in line with preferences, rather than autonomy over work tasks, which was not differentiated here. The lack of relationship between the two assessed operationalizations of work autonomy (O*NET scores and self-reported autonomy) was unexpected. Though the two assessed O*NET indicators (freedom and structure) were highly correlated with one another, warranting their combination into a two-item composite score, it is possible that the O*NET operationalization is construct deficient in some way as compared to the construct as defined by the self-report scale. It is further worth acknowledging that while O*NET scores are derived across workers holding the same occupation, using a single score to represent a job cannot definitionally differentiate within-occupation variance, as one individual in a given job may have greater autonomy than another. Alternatively, noting that O*NET scores are based on incumbent data likely sourced prior to the COVID-19 pandemic, perhaps self-reported scores of work autonomy are more reflective of their work context in the unprecedented times in which this data was collected, as compared to O*NET which may be more representative of the work context in traditional circumstances. Perhaps the most significant contribution of the presented results is the demonstrated association between distancing identity management and enacted boundary management. Counter to expectations, this identity management strategy predicted enacted boundaries across both boundary directions, work-to-life and life-to-work, though expected to most strongly apply to life-to-work boundary management. This finding suggests that perhaps one’s desire to conceal their identity at work may lead to complete segmentation across spheres; rather than keeping track of boundaries as differentially permeable and flexible depending on boundary direction, it 43 may be psychologically beneficially for those managing workplace stigma to enact complete segmentation when seeking identity concealment for the sake of boundary simplicity and clarity. Of note, however, were the potential measurement issues associated with the identity management scale used in the present study- specifically, though the two-factor model of identity management fit the data better than did the single-factor model, neither displayed fit up to traditional standards of acceptability. While worthwhile to acknowledge the greater evidence provided for the treatment of identity management strategies as separate rather than ends of the same construct, questions still remain regarding the overall measure’s quality. At the item level, removing any item from either scale did not demonstrate improved measure reliability, suggesting no particular item can be identified as the driving issue for weakened fit. Further examinations of scale skewness demonstrated normal distributions across both measures (affirming: skewness = -.11, S.E. = .16; distancing: skewness = .09, S.E. = .16). Though these cursory examinations did not identify the reasons for low factor fit, it is possible that this measure, originally published in 2004, would benefit from re-examination as it is likely that sexual orientation identity management at work may have evolved given increased legislative and general public acceptance of sexual minority individuals since the early 2000’s. Finally, the present results underscore the important role diversity climate perceptions play within examined processes. Specifically, diversity climate perceptions were found to both directly predict identity management strategies, as well as work attitudinal outcomes of job satisfaction, subjective wellbeing, and work withdrawal. Though perhaps not entirely surprising given previous meta-analytic evidence noting the importance of diversity climate perceptions in predicting work attitudes (Mor Barak et al., 2016), the present study contributes via the highlighted emphasis of importance for LGB workers specifically, as well as the more novel 44 finding of the direct and indirect relationships of the variable on enacted boundary management. Indeed, the finding that diversity climate is indirectly related to boundary management through identity management is a contribution to both identity and work-life literatures, underscoring the interrelatedness of perceived identity acceptance on enacted work-life boundaries. While the purpose of Study 1 was to examine LGB workers who vary in boundary management preferences, identity management strategy enactment, and work context, what remains unanswered is how the management of potential stigma via identity management strategies may impact boundary management as compared to those who do not face that particular type of stigmatization. Accordingly, I conclude with the presentation of a second study taking a between-group approach, comparing heterosexual and LGB workers’ respective congruence between boundary management preferences and enacted boundary management, testing the notion that LGB workers may experience less congruence as a result of their added identity management constraint. 45 CHAPTER 7: STUDY 2 METHOD Procedure & Measures Study 2 was also a multi-wave survey, assessing participants at two time points. At Time 1, participants were surveyed regarding their boundary management preferences, enacted boundary management, and demographics using the same measures as used in Study 1 (α = .88- .94). All individuals with quality Time 1 responses were invited to participate in the Time 2 survey one week later, which asked participants to complete the same measures of subjective wellbeing (α = .92), withdrawal (α = .92), and job satisfaction (α = .92) as used in Study 1. The item used to assess the degree to which the COVID-19 pandemic impacted participants’ work did not impact the results of hypothesis testing when analyzed as a covariate, so results are presented here without that variable as a focus. Participants All Study 2 participants were recruited through Amazon Mechanical Turk (MTurk), with inclusion criteria of current employment at least part-time in non-self-employed circumstances, and residence in the United States. A total of 636 Time 1 responses were collected, with 412 passing a qualitative check screening out open-ended responses that were gibberish, irrelevant, or pasted directly from external Internet sources (N = 224). This screening standard is in line with current recommendations toward data quality when using such platforms as MTurk (Chmielewski & Kucker, 2020). Quotas were placed within the survey to ensure a relatively equal breakdown of heterosexual (N = 209) and LGB (N = 203) participants. Of the 203 LGB participants at Time 1, 65.5% reported identifying as bisexual, 17.7% as gay, and 16.7% as gay. 46 Of the 412 Time 1 participants eligible for Time 2, 306 participated in the second survey (74.3%) approximately one week later. Those reporting sexual orientation at Time 2 differing from that reported at Time 1 were removed from analyses (N = 14), leaving a final two-time- point sample of 292 participants. The sexual orientation group breakdown for the final sample was 51.7% heterosexual (N = 151), 48.2% LGB (N = 141), with the LGB sub-sample broken down into 65.2% bisexual, 19.9%% lesbian and 14.9% gay. Participant gender breakdown was 49.3% female, 48.3% male, and 1.7% non-binary. Participants were on average 36.40 years old (SD = 9.94), working on average 40.49 hours per week (SD = 7.73), employed at their current organization for an average of 6.63 years (SD = 5.71), and a slight majority of participants were working outside of the home (51.6%; note data was collected in September/October 2020 during the COVID-19 pandemic). Example job titles included IT analyst, financial advisor, and clinical social worker. Racial/ethnic breakdown for this sample was the following (participants could self-identify as multiple options): 80.1% White, 9.2% Black/African-American, 7.2% Latinx, 3.8% East Asian, 1.4% South Asian, 1.0% American Indian/Alaska Native, 0.3% Native Hawaiian/Pacific Islander, and 1.7% identifying as an “other” racial identity. Participants were compensated $1.00 for completing the Time 1 survey (lasting on average 8 minutes), and $1.00 for completing the Time 2 survey (lasting on average 4 minutes). This pay rate exceeds federal minimum wage rates (i.e. $7.25/hour), in line with equitable compensation recommendations for MTurk and similar platforms (Brawley & Pury, 2016). Analytic Technique Congruence between boundary management preferences and enacted boundary management was operationalized in this study using polynomial regression. Polynomial 47 regression was first outlined by Edwards and Parry (1993), and has been argued to provide a clearer interpretation of both individual and environment profile effects while reducing many of the constraints proposed by traditional congruence indices (Nye, Prasad, Bradburn & Elizondo, 2018). Using each outcome as the dependent variable (i.e., subjective wellbeing, withdrawal, job satisfaction), I estimated polynomial regression models including main effects for boundary management preferences and enacted boundary management, quadratic (i.e. squared) terms, and interactions between preferred and enacted boundary management. Due to the possibility of multicollinearity in the model, all terms were centered prior to analysis; despite this effort, two terms (i.e., work to non-work enactment and the non-work to work enactment quadratic term) were excluded from each of the models due to the variables exceeding the software’s tolerance limit (analyses adjusting SPSS’ default tolerance limit toward greater inclusion still failed to include these two terms). Tables 16, 17, and 18 display the results of these polynomial regressions for subjective wellbeing, withdrawal, and job satisfaction respectively, with the predicted values from these regressions saved and utilized as congruence terms (interpreted as higher values meaning greater congruence) for hypothesis testing. 48 CHAPTER 8: STUDY 2 RESULTS Table 19 displays means, standard deviations, and intercorrelations between Study 2 variables. Note that higher scores on both boundary management variables indicate greater preferred/enacted segmentation, with lower scores indicating greater boundary integration. Viewing Table 19, it is of note that sexual orientation appears related to each of the assessed outcomes, such that sexual minority workers appear lower in subjective wellbeing and job satisfaction, and higher in reported withdrawal, than heterosexual workers. Additionally, sexual orientation appears related to enacted boundary management, such that sexual minority employees (coded as the higher value) were reporting less enacted segmentation than were heterosexual employees, against expectations. Participant demographics were again largely uncorrelated with study focal variables, though as was found in Study 1, age was significantly associated with all study outcomes, suggesting that older workers reported higher subjective wellbeing, higher job satisfaction, and lower withdrawal than did younger workers. Hypothesis Testing To examine the predictions laid out in Hypotheses 8 and 9, a series of mediation models were conducted using Hayes’ (2013) PROCESS Macro model 4. Separate mediation models were conducted for each of the three outcomes (i.e. subjective wellbeing, withdrawal, and job satisfaction), as the boundary management congruence scores calculated via polynomial regression are specific to each outcome. Concerning subjective wellbeing, the mediation model conducted with sexual orientation as the predictor, boundary management congruence as the mediator, and subjective wellbeing as the outcome was significant, R2 = .02, MSE = .98, F(1, 278) = 5.57, p = .019. Considering H8’s 49 prediction that sexual orientation would relate to congruence between preferred and enacted boundary management, this notion was supported in this instance, as sexual orientation significantly and negatively predicts boundary management congruence (b = -.28, p = .02; Table 20), with sexual minority workers appearing to experience lower congruence than heterosexual workers. Looking again to Table 20, H9a appears supported, as boundary management congruence is significantly and positively predictive of subjective wellbeing (b = .32, p < .001). Of additional note, sexual orientation appears to be directly predictive of subjective wellbeing in addition to boundary management congruence, suggesting that sexual minority workers experience lower subjective wellbeing as compared to heterosexual workers (b = -.50, p < .001; Table 20). Finally, evidence for mediation is shown via the indirect effect, as the 95% confidence interval of the value does not include zero, b = -.09, SE = .04, 95%CI [-.17, -.02]. A mediation model with sexual orientation as the predictor, boundary management congruence as the mediator, and withdrawal as the outcome was conducted to examine how Hypotheses 8 and 9 faired with respect to withdrawal. The overall mediation model was not significant [R2 = .01, MSE = 0.99, F(1, 290) = 1.69, p = .19], and neither was the indirect effect speaking to the presence of mediation, as the confidence interval of the indirect effect contained zero, b = .04, SE = .03, 95%CI [-.02, .10]. However, Table 21 allows us to examine whether the individual paths within the mediation model may speak to our hypothesized links. H8’s predicted link between sexual orientation and boundary management congruence was not supported here (b = -.15, p = .19); however, H9b was supported, as boundary management congruence was indeed predictive of withdrawal (b = -.24, p < .001), as was sexual orientation (b = .16, p = .05; Table 21), suggesting that sexual minority workers experience greater withdrawal as compared to heterosexual workers. 50 A final mediation model was conducted with sexual orientation as the predictor, boundary management congruence as the mediator, and job satisfaction as the outcome. Similar to the withdrawal model, the overall mediation model for job satisfaction was not significant [R2 = .01, MSE = 1.00, F(1, 281) = 2.27, p = .13], and neither was the indirect effect speaking to the potential presence of mediation as the confidence interval of the indirect effect contained zero, b = -.10, SE = .06, 95%CI [-.69, .03]. However, we can examine Table 22 to determine whether individual paths within the mediation model may support any hypothesized links. H8 was not supported here, as the relationship between sexual orientation and boundary management congruence was not significant, b = -.18, p = .13 (Table 22). However, H9 was supported, as boundary management congruence was significantly and positively predictive of job satisfaction, b = .54, p < .001. Therefore, across all mediation models, H8 received some (but limited) support, while H9 was fully supported. Combined Path Model Finally, I conducted a combined path model in MPlus to see whether the PROCESS results replicated when all variables were included in a single model (i.e. sexual orientation predicting each of the three boundary management congruence indices, with each congruence index predicting its respective outcome). Boundary management congruence indices were further correlated given their conceptual similarity, showing strong associations (Table 19). The model showed good fit, with all indices in line with traditional cutoffs (RMSEA = .076, CFI = .981, TLI = .956, SRMR = .050), though the Chi-square test of model fit was significant (χ2 (9) = 24.10, p = .0042). Table 23 demonstrates results for the standardized structural model, which largely replicates findings calculated within the PROCESS models. All paths linking boundary 51 management congruence and each of the three outcomes was significant; however, as previously demonstrated via the PROCESS models, neither link between sexual orientation and boundary management congruence operationalized for job satisfaction nor withdrawal was significant, though the link between sexual orientation and boundary management congruence operationalized via subjective wellbeing did display significance. Similarly, the only significant indirect effect was found between sexual orientation and subjective wellbeing (b = -.05, S.E. = .02, p = .025), whereas the indirect effects on job satisfaction (b = -.03, S.E. = .02, p = .126) and withdrawal (b = -.02, S.E. = .02, p = .216) were not significant. 52 CHAPTER 9: DISCUSSION Recognizing that those facing identity-based stigma for invisible identities may need to present themselves in strategic ways so as to minimized experienced prejudice, the purpose of the present dissertation was to examine whether such identity management may predict enacted work and life boundaries, acting as a potential constraint leading to eventual worsened work- relevant attitudinal outcomes. Though not all hypothesized associations were supported here, the present study combination does advance our understanding of both identity management and work-life boundary management for employees facing workplace stigma. Beginning with the proposed association between employee identity management and subsequent enacted work-life boundaries, Study 1 did provide support that those engaged in distancing identity management strategies were more likely to segment their work and non-work spheres. This relationship was the underlying logic driving the proposed H8 difference in LGB and heterosexual employees’ respective boundary management congruence examined in Study 2- though, that hypothesis received less support, as the only significant difference emerged in the case of boundary management congruence operationalized using subjective wellbeing. Taken in total, the set of Study 2 findings provided evidence of sexual orientation and boundary management congruence both as direct predictors of work relevant outcomes, though the proposed indirect effect was not strongly supported. Acknowledging that disparities did indeed emerge when comparing LGB and heterosexual employees’ subjective wellbeing, job satisfaction, and withdrawal from work, one may wonder what mechanisms are contributing to these differences if not differential boundary management congruence. Indeed, while boundary management congruence was found to mediate 53 the relationship between sexual orientation and subjective wellbeing, it was not supported as a mediator for either job satisfaction or work withdrawal. Some potential alternative mechanisms include experienced workplace incivility (Zurbrügg & Miner, 2016), discrimination as a result of identity disclosure (Ragins et al., 2007) and lack of protective legislation (Ragins & Cornwell, 2001), and lessened inclusion (Shore et al., 2011), suggesting a number of areas for continued research on the development of interventions to lessen noted disparities. An additional notion examined across both studies was the importance of boundary management congruence on work-relevant attitudinal outcomes. Examined via differing analytic strategies across studies (i.e., hierarchical regression in Study 1 and polynomial regression in Study 2), it was operationalization through polynomial regression that demonstrated the greatest support for the role of congruence in predicting subjective wellbeing, job satisfaction, and work withdrawal. Indeed, boundary management congruence was found to be a significant predictor across each of the three outcomes. As the corresponding hierarchical interactions in Study 1 received no support, one may wonder whether the resulting cross-study difference lies in a method effect benefitting polynomial regression- as the congruence terms are definitionally created with respect to the outcome, it may be that the likelihood of association between congruence term and outcome is resultingly quite high. However, a recent application of the procedure provides a contrary example, such that congruence between preferred and actual working time arrangement operationalized via polynomial regression did not predict work-life balance as expected (Brauner, Wöhrmann & Michel, 2020). This example highlights that the method can indeed discriminate between situations in which congruence is and is not relevant to a particular outcome, and given the method’s proposed benefits above and beyond the limitations 54 plaguing difference-score approaches (Edwards & Parry, 1993), this approach has been put forth as a strong operationalization of congruence (Nye et al., 2018). Though support for the interaction between preferred and enacted boundary management was more limited in Study 1, Study 1 demonstrated multiple direct effects of boundary management preferences and enactment on outcomes, finding that those with greater enacted segmentation reported higher subjective wellbeing and lessened work withdrawal than did those with lower segmentation. At the same time, boundary management preferences for segmentation were negatively associated with job satisfaction and positively linked to withdrawal in the first steps of their respective models. Seemingly contradictory in nature, perhaps the difference in the association of segmentation preferences/enactment on outcomes lies in the differential stability of the respective predictors, as well as the timing of study administration. As data collection for both studies occurred during the COVID-19 pandemic (with roughly half of each study sample working from home at the time of data collection), it may be that experienced segmentation may have been particularly valued in this time, as the lines between work and home became increasingly blurred; moreover, while segmentation preferences are traditionally conceptualized as more stable in nature (Ashforth et al., 2000; Clark, 2000), the unprecedented nature of the pandemic may have been strong enough to shift individual preferences. That is, individuals who may have previously desired boundary integration pre- pandemic may now find themselves longing for segmentation, as their work and home lives now overlapped in unprecedented ways (indeed, both study distributions for boundary preferences were negatively skewed suggesting overall preference for segmentation, though without pre-test preference levels it is hard to say definitively whether any shift occurred). Therefore, for those who consistently desire separation of their work and home lives, it is worthwhile to consider the 55 potential motivational reasoning behind those preferences- if one prefers to segment one’s spheres given a desire to disconnect from an unsupportive, hostile, and non-acceptive work environment, then that notion may be in support of the relationships proposed in the present dissertation. Future research attempting to replicate these relationships post-pandemic would help elucidate these musings. Theoretical and Practical Implications The presented set of studies has both theoretical and practical implications. Theoretically, this dissertation seeks to integrate the diversity and work-life literatures, allowing a widened theoretical lens with which to examine the nuance of the work-life interface. By drawing upon identity management and boundary management literatures, the current work advances existing theory toward better understanding of important work-life processes for a currently underexamined population, addressing broader critiques of the literature as most often focused on more privileged samples (Aycan, 2008). By demonstrating the role that identity management plays in workers’ enacted boundary management, the current findings underscore the pervasiveness of stigma, and how employees’ attempts to mitigate experienced stigma may affect their lives in previously unexamined ways; consequently, future research should continue to incorporate models of stigma within their proposed understanding of organizational functioning for both marginalized and non-stigmatized employees. Even beyond the role of stigma, the present research fills a gap toward increased evidence of boundary management antecedents, one of which being boundary management preferences. In its current state, the boundary management literature has focused almost entirely on the consequences of managed boundaries (Gardner et al., 2021), rather than understanding what factors or personal characteristics may contribute to employee boundaries between work 56 and life spheres. Though the difference between preferred and enacted boundaries was implicit in early boundary management theorizing (Ashforth et al., 2000; Clark, 2000), this dissertation takes a needed step forward in not only demonstrating this linear relationship, but also examining the relationship with greater nuance in congruence via polynomial regression. Future research can continue this effort by not only identifying alternative predictors of managed boundaries (perhaps locus of control, or family circumstances including coupled or parenting status), but to further investigate additional outcomes of boundary management congruence beyond those examined in the present studies, potentially including organizational commitment and job involvement. Practically, this dissertation seeks to highlight the consequences of barriers faced by employees with invisible stigmatized identities, examined through the lens of minority sexual orientation. Indeed, Study 2 results highlight the disparities in workplace-relevant attitudes experienced by sexual minority workers as compared to heterosexual workers, while Study 1 demonstrated the importance of perceived diversity climate in predicting these same work- relevant outcomes. It is the hope that the present findings may speak more broadly to individuals of other marginalized invisible identities, including religious minorities and individuals with invisible disabilities, who all may face stigma within their workplace. Even beyond invisible identities, the present work may have implications for those of marginalized visible identities, including race and gender. Though the conceptualization of identity management here is most relevant to the dynamics of disclosure, the salience with which one underscores their visible identity at work (as related conceptually to both codeswitching and authenticity) may still relate to one’s enacted boundaries between work and life; for example, perhaps a woman who feels her 57 gender is stigmatized at work may seek to minimize discussion of her family role while in the office so as to avoid highlighting her gender further. In considering how organizations may practically consider the present results, Study 2 findings highlight the importance of congruence between preferred and enacted boundary management for employees. Perhaps then, organizations could seek to create a culture in which employees can manage their boundaries as close to their preferences as possible. While this notion may be more difficult for some roles than others, organizations at least encouraging employees to disconnect and segment when desired and possible, as well as cultivating a climate in which employees can bring their whole selves to work (perhaps through improved diversity climates) may allow employees’ a greater range of possible enacted boundary management strategies. Further, managerial steps toward adjusting and implementing unique and personalized schedules for employee when able, encouraging employee participation in optimizing synchronous versus asynchronous outside-of-work-hours communication, and role-modeling healthy work/non-work behaviors in line with preferences may be actionable steps toward increased employee boundary management congruence (Perrigino & Raveendhran, 2020). Limitations and Future Directions As with all research, the current set of studies feature limitations worthy of acknowledgement. First, though Study 1 outcomes were separated in time from other assessed variables, identity management, boundary management, and diversity climate were assessed within the same time point, suggesting common method variance may be a concern for those relationships, as well as difficulty in determining causality. Indeed, one potential alternative that cannot be disentangled given present methodology is the extent that identity management relates to enacted boundary management through boundary management preferences, as Table 1 shows 58 distancing identity management as demonstrating actually a stronger relationship with preferred than with enacted boundary management. Another possibility is that segmentation preferences are informed by current employee job satisfaction and withdrawal, such that those unhappy with and currently withdrawing from their current organization prefer greater segmentation. Future research further separating the measurement of variables across survey administrations can help to disentangle these relationships and address the present issues of causality. However, by following the recommendations of Podsakoff and colleagues (2003) (including the use of varied scale anchors and the maintenance of participant anonymity), I hope to have mitigated the concerns of common method variance within the present research. Of note, the final representation of LGB participants within both studies varied, with the Study 1 sample featuring primarily older gay men, as compared to Study 2 which primarily featured bisexuals within the LGB subgroup. Acknowledging these sample differences had me wonder if either sampling was more or less representative of “true” sexual orientation identity breakdowns, finding that recent estimates by both the Pew Research Center (2013) and the Williams Institute (Gates, 2011) demonstrate bisexuals as constituting the largest subgroup within the LGBT classification within the US. Accordingly, this notion would suggest that the Study 2 sample may not be far off from actual identity representation, or at least came out as more representative as compared to Study 1. Regardless, the current set of studies still lumps together these three sexual identities into a single group, potentially missing important nuance between the individual identity categories (Arena & Jones, 2017; Corrington et al., 2019). Though subgroup differences were difficult to examine here given unequal and small sample sizes of each subgroup within each study, future research could benefit from purposefully examining the hypothesized relationships within each 59 of the examined identities individually so as to more explicitly delineate the potential differences in experiences. Indeed, it is not only possible but quite likely that differences in identity management strategies may exist across LGB subgroups, potentially grounded in both gender differences as well as identity visibility (e.g., a bisexual employee in an opposite-sex relationship may be able to more easily conceal their sexual orientation than either a gay or lesbian employee), so therefore understanding those potential differences is worthy of further examination. Beyond sexual orientation, the two studies further differed with respect to age representation, with Study 1 featuring older workers than those in Study 2- given the associations between age and outcome variables found across both studies, an important future step may be to consider the intersectional implications of age and sexual orientation on boundary management and work-relevant attitudes. An additional opportunity for future research lies in the consideration of participant relationship status. Though relationship status was not assessed here, it is possible that LGB employees not currently in a relationship may have greater ease in concealing their sexual orientation at work (if desired) than do LGB employees currently in a relationship. Though there are recent examples of research investigating the cross-over effects of sexual orientation disclosure at work on couples at home (e.g. Williamson et al., 2017; Holman, 2018), the potential moderating effects of relationship status itself does not appear to be sufficiently explored, allowing ample opportunity for future work. Conclusion For employees facing identity-based stigma, bringing one’s whole self to work may expose oneself to experienced prejudice; accordingly, the present dissertation sought to unpack the notion that those seeking to conceal a stigmatized identity at work may subsequently separate 60 their work and life spheres, regardless of their ideal boundaries between their work and home lives. Across two studies, the present dissertation demonstrates a novel relationship between stigmatized identity management and enacted boundary management (Study 1), while further underscoring the importance of congruence between preferred and enacted boundary management in predicting a number of work-relevant attitudes (Study 2). Accordingly, the current research contributes both theoretically and practically, via the integration of diversity and work-life literatures, as well as the spotlight on currently understudied populations as relevant to potentially all employees facing marginalization at work. 61 APPENDICES 62 APPENDIX A Tables and Figures Table 1. Study 1 means, standard deviations, scale reliabilities and intercorrelations. M SD 1 2 3 4 5 6 7 8 9 10 11 12 1 BM Preferences 5.55 1.04 (.88) 2 BM Enactment 4.52 1.45 .53 (.93) 3 Affirming IM 4.21 1.38 .03 .08 (.90) 4 Distancing IM 3.82 1.45 .30 .22 -.38 (.87) Autonomy 5 78.5 11.07 -.13 -.12 .14 -.17 (.90) (O*NET) 6 Autonomy (Self) 3.66 0.88 -.06 -.01 .23 -.13 .14 (.81) Diversity 7 4.15 0.73 -.07 .15 .24 -.17 .07 .28 (.89) Climate 8 Job Satisfaction 5.23 1.46 -.13 .03 .12 -.17 .12 .41 .46 (.92) 9 SWB 3.26 0.97 -.03 .14 .17 -.09 .16 .29 .29 .60 (.92) 10 Withdrawal 1.93 0.67 .07 -.11 .07 .14 .05 -.12 -.20 -.49 -.38 (.91) 11 Gender 1.31 0.46 .13 -.10 .05 -.02 .05 -.04 -.01 -.09 -.17 .15 // 12 Race 0.88 0.33 -.08 -.02 -.02 -.17 -.20 .09 .10 .12 .13 -.09 -.08 // 13 Age 50.01 14.08 -.12 .12 -.10 -.06 -.09 .08 .01 .19 .24 -.39 -.41 .22 NOTE. Cronbach’s alpha listed on diagonal, “//” indicated for variables that do not have reliabilities. Bolded values are significant at p > .05. BM = Boundary management. IM = Identity management. SWB = Subjective wellbeing. Gender coded as 1 = “Male,” 2 = “Female.” Race coded as 0 = “Non-White,” 1 = “White.” 63 Table 2. Hierarchical regression results of boundary management preferences and autonomy operationalized via self-report as predictive of enacted boundary management. Model Variable B SE B β P-Value R2 1 Intercept 4.52 0.08 <.001 BM Preferences 0.75 0.08 0.53 <.001 Autonomy (Self) 0.03 0.10 0.02 0.727 0.29 2 Intercept 4.52 0.08 <.001 BM Preferences 0.74 0.08 0.53 <.001 Autonomy (Self) 0.04 0.10 0.02 0.716 BM Preferences X Autonomy 0.08 0.08 0.05 0.355 0.29 NOTE. BM = Boundary management. 64 Table 3. Hierarchical regression results of boundary management preferences and autonomy operationalized via O*NET scores as predictive of enacted boundary management. Model Variable B SE B β P-Value R2 1 Intercept 4.58 0.13 <.001 BM Preferences 0.77 0.13 0.52 <.001 Autonomy (O*NET) -0.01 0.01 -0.06 0.510 0.28 2 Intercept 4.56 0.13 <.001 BM Preferences 0.79 0.13 0.53 <.001 Autonomy (O*NET) -0.01 0.01 -0.03 0.720 BM Preferences X Autonomy -0.02 0.01 -0.10 0.274 0.29 NOTE. BM = Boundary management. 65 Table 4. Linear regression results of distancing and affirming identity management predicting enacted boundary management. Outcome Predictor B SE B β P-Value R2 Enacted BM Intercept 3.66 0.27 <.001 Distancing IM 0.22 0.07 0.22 <.001 0.05 Enacted BM Intercept 4.16 0.32 <.001 Affirming IM 0.09 0.07 0.08 0.234 0.01 NOTE. BM = Boundary management. IM = Identity management. 66 Table 5. Multiple regression results of affirming and distancing identity management strategies as predicting enacted boundary management. Variable B SE B β P-Value R2 Intercept 2.53 0.5 <.001 Affirming IM 0.20 0.08 0.19 0.007 Distancing IM 0.30 0.07 0.30 <.001 0.08 NOTE. IM = Identity management. 67 Table 6. Linear regression results of distancing identity management as predictive of enacted life-to-work and work-to-life segmentation. Outcome Predictor B SE B β P-Value R2 Life-to-Work Segmentation Intercept 3.66 0.29 <.001 Distancing IM 0.24 0.07 0.23 <.001 0.05 Work-to-Life Segmentation Intercept 3.67 0.32 <.001 Distancing IM 0.21 0.08 0.18 <.001 0.03 NOTE. IM = Identity management. 68 Table 7. Hierarchical regression results of boundary management preferences and enacted boundary management as predictive of subjective wellbeing. Model Variable B SE B β P-Value R2 1 Intercept 3.26 0.07 <.001 BM Preferences -0.13 0.07 -0.14 0.074 Enacted BM 0.15 0.05 0.22 0.007 0.04 2 Intercept 3.24 0.07 <.001 BM Preferences -0.11 0.08 -0.12 0.171 Enacted BM 0.14 0.06 0.20 0.014 BM Preferences X Enacted BM 0.04 0.04 0.07 0.36 0.04 NOTE. BM = Boundary management. 69 Table 8. Hierarchical regression results of boundary management preferences and enacted boundary management as predictive of job satisfaction. Model Variable B SE B β P-Value R2 1 Intercept 5.23 0.1 <.001 BM Preferences -0.28 0.11 -0.20 0.014 Enacted BM 0.14 0.08 0.14 0.089 0.03 2 Intercept 5.14 0.11 <.001 BM Preferences -0.2 0.12 -0.14 0.091 Enacted BM 0.11 0.08 0.10 0.199 BM Preferences X Enacted BM 0.11 0.06 0.13 0.062 0.05 NOTE. BM = Boundary management. 70 Table 9. Hierarchical regression results of boundary management preferences and enacted boundary management as predictive of withdrawal. Model Variable B SE B β P-Value R2 1 Intercept 1.93 0.05 <.001 BM Preferences 0.12 0.05 0.18 0.023 Enacted BM -0.10 0.04 -0.21 0.01 0.04 2 Intercept 3.24 0.07 <.001 BM Preferences 0.09 0.06 0.15 0.087 Enacted BM -0.09 0.04 -0.18 0.026 BM Preferences X Enacted BM -0.03 0.03 -0.09 0.196 0.04 NOTE. BM = Boundary management. 71 Table 10. Linear regression results of diversity climate perceptions predicting affirming and distancing identity management. Outcome Predictor B SE B β P-Value R2 Affirming IM Intercept 2.35 0.53 <.001 Diversity Climate 0.45 0.13 0.24 <.001 0.06 Distancing IM Intercept 5.19 0.57 <.001 Diversity Climate -0.33 0.13 -0.17 0.015 0.03 NOTE. IM = Identity management. 72 Table 11. Linear regression results of diversity climate perceptions predicting subjective wellbeing, job satisfaction and work withdrawal. Outcome Predictor B SE B β P-Value R2 Subjective Wellbeing Intercept 1.63 0.37 <.001 Diversity Climate 0.39 0.09 0.29 <.001 0.09 Job Satisfaction Intercept 1.37 0.52 <.001 Diversity Climate 0.93 0.12 0.46 <.001 0.21 Withdrawal Intercept 2.68 0.26 <.001 Diversity Climate -0.18 0.06 -0.2 0.004 0.04 73 Table 12. PROCESS Model 4 results of diversity climate perceptions predicting enacted boundary management through distancing identity management. Outcome Distancing IM b SE t p LLCI ULCI Intercept 5.19 0.57 9.16 <.001 4.07 6.31 Diversity Climate -0.33 0.13 -2.46 .014 -0.60 -0.07 Enacted BM b SE t p LLCI ULCI Intercept 1.98 0.65 3.03 .002 0.69 3.27 Diversity Climate 0.38 0.13 2.82 .005 0.11 0.64 Distancing IM 0.26 0.07 3.82 <.001 0.12 0.39 NOTE. BM = Boundary management. IM = Identity management. 74 Table 13. PROCESS Model 4 results of distancing identity management predicting subjective wellbeing through enacted boundary management. Outcome Enacted BM b SE t p LLCI ULCI Intercept 3.58 .27 13.43 <.001 3.14 4.22 Distancing IM .22 .07 3.29 .001 0.09 0.35 Subjective Wellbeing b SE t p LLCI ULCI Intercept 3.07 0.25 12.23 <.001 2.58 3.57 Distancing IM -0.08 0.05 -1.82 .069 -0.18 0.01 Enacted BM 0.11 0.05 2.45 .015 0.02 0.20 NOTE. BM = Boundary management. IM = Identity management. 75 Table 14. PROCESS Model 4 results of distancing identity management predicting withdrawal through enacted boundary management. Outcome Enacted BM b SE t p LLCI ULCI Intercept 3.68 0.27 13.43 <.001 3.14 4.22 Distancing IM .22 .07 3.29 .001 0.09 0.35 Withdrawal b SE t p LLCI ULCI Intercept 1.93 0.17 11.11 <.001 1.59 2.27 Distancing IM 0.08 0.03 2.46 .015 0.02 0.14 Enacted BM -.07 0.03 -2.10 .037 -0.13 -0.004 NOTE. BM = Boundary management. IM = Identity management. 76 Table 15. Linear regression results of boundary management congruence operationalized via polynomial regression predicting subjective wellbeing, withdrawal, and job satisfaction. Outcome Predictor B SE B β P-Value R2 BM Congruence Intercept 3.26 0.06 <.001 Subjective Wellbeing 0.24 0.07 0.24 <.001 0.06 BM Congruence Intercept 1.93 0.04 <.001 Withdrawal -0.21 0.04 -0.32 <.001 0.10 BM Congruence Intercept 5.23 0.09 <.001 Job Satisfaction 0.47 0.1 0.32 <.001 0.10 NOTE. BM = Boundary management. 77 Table 16. Study 2 polynomial regression results predicting subjective wellbeing. Variable B Std. Error Beta t p-value Intercept 3.23 0.10 31.693 <.001 WNW Preferences -0.30 0.09 -0.34 -3.55 <.001 NWW Preferences 0.16 0.06 0.20 2.68 .008 NWW Enactment 0.14 0.04 0.25 3.51 .001 WNW Preferences X WNW Preferences -0.02 0.04 -0.05 -0.39 .697 NWW Preferences X NWW Preferences 0.08 0.03 0.24 3.13 .002 WNW Enactment X WNW Enactment 0.01 0.02 0.02 0.34 .734 WNW Preferences X WNW Enactment -0.01 0.04 -0.02 -0.18 .855 NWW Preferences X NWW Enactment -0.04 0.03 -0.10 -1.60 .11 NOTE: WNW = Work to Non-Work; NWW = Non-Work to Work; Preferences and enactment refer to desired/enacted level of segmentation. Work to Non-Work Enactment and the Non-Work to Work Enactment quadratic term were excluded variables due to exceeding tolerance limits. 78 Table 17. Study 2 polynomial regression results predicting withdrawal. Variable B Std. Error Beta t p-value Intercept 3.86 0.08 50.269 <.001 WNW Preferences -0.01 0.06 -0.01 -0.12 .908 NWW Preferences 0.18 0.04 0.32 4.19 <.001 NWW Enactment 0.02 0.03 0.05 0.68 .500 WNW Preferences X WNW Preferences -0.001 0.03 -0.002 -0.02 .987 NWW Preferences X NWW Preferences 0.06 0.02 0.23 2.95 .003 WNW Enactment X WNW Enactment -0.02 0.02 -0.06 -0.88 .379 WNW Preferences X WNW Enactment 0.05 0.03 0.15 1.51 .133 NWW Preferences X NWW Enactment 0.001 0.02 0.003 0.04 .966 NOTE: WNW = Work to Non-Work; NWW = Non-Work to Work; Preferences and enactment refer to desired/enacted level of segmentation. Work to Non-Work Enactment and the Non-Work to Work Enactment quadratic term were excluded variables due to exceeding tolerance limits. 79 Table 18. Study 2 polynomial regressions results predicting job satisfaction. Variable B Std. Error Beta t p-value Intercept 4.88 0.16 30.24 <.001 WNW Preferences -0.46 0.13 -0.33 -3.46 .001 NWW Preferences 0.37 0.09 0.30 4.00 <.001 NWW Enactment 0.05 0.06 0.05 0.74 .460 WNW Preferences X WNW Preferences -0.12 0.07 -0.20 -1.68 .093 NWW Preferences X NWW Preferences 0.12 0.04 0.22 2.89 .004 WNW Enactment X WNW Enactment 0.003 0.04 0.01 0.08 .940 WNW Preferences X WNW Enactment 0.14 0.07 0.21 2.18 .030 NWW Preferences X NWW Enactment -0.07 0.04 -0.10 -1.63 .105 NOTE: WNW = Work to Non-Work; NWW = Non-Work to Work; Preferences and enactment refer to desired/enacted level of segmentation. Work to Non-Work Enactment and the Non-Work to Work Enactment quadratic term were excluded variables due to exceeding tolerance limits. 80 Table 19. Means, standard deviations, and intercorrelations between Study 2 variables. Variable M SD 1 2 3 4 5 6 7 8 9 10 11 1 Sexual Orientation 0.48 0.5 // 2 BM Preferences 5.8 1.02 -.05 (.88) 3 BM Enactment 4.4 1.51 -.13 .30 (.94) 4 BM Congruence (JS) 0 1 -.09 -.07 .33 // 5 BM Congruence (SWB) 0 1 -.13 -.25 .55 .81 // 6 BM Congruence (W) -0.01 1 -.07 .41 .43 .73 .52 // 7 Subjective Wellbeing 3.38 1.01 -.30 -.08 .22 .28 .35 .18 (.92) 8 Job Satisfaction 4.98 1.62 -.13 -.03 .15 .34 .27 .25 .51 (.92) 9 Withdrawal 2.06 0.76 .14 -.13 -.17 -.23 -.16 -.31 -.33 -.56 (.92) 10 Gender 1.51 0.5 .19 .09 -.14 -.06 -.13 -.001 -.15 -.12 .12 // 11 Race 0.8 0.4 -.05 .05 -.03 -.07 -.07 -.04 .17 .08 -.15 -.004 // 12 Age 36.4 9.94 -.14 .06 .08 .05 .01 .05 .15 .16 -.23 -.02 .17 NOTE: Bolded values significant at p < .05. Sexual orientation coded as 1 = “Heterosexual,” 2 = “Lesbian, Gay or Bisexual.” BM = “Boundary Management.” SWB = “Subjective wellbeing.” JS = “Job satisfaction.” W = “Withdrawal.” Gender coded as 1 = “Male,” 2 = “Female.” Race coded as 0 = “Non-White,” 1 = “White.” Scale reliabilities are listed on the diagonal. 81 Table 20. PROCESS Model 4 results of sexual orientation predicting subjective wellbeing through boundary management congruence. Outcome BM Congruence b SE t p LLCI ULCI Intercept 0.42 0.19 2.24 .03 0.05 0.79 Sexual Orientation -0.28 0.12 -2.36 .02 -0.51 -0.05 Subjective Wellbeing b SE t p LLCI ULCI Intercept 4.12 0.18 23.43 <.001 3.77 4.46 Sexual Orientation -0.50 0.11 -4.46 <.001 -0.72 -0.28 BM Congruence 0.32 0.06 5.71 <.001 0.21 0.43 NOTE: BM = “Boundary Management.” 82 Table 21. PROCESS Model 4 results of sexual orientation predicting withdrawal through boundary management congruence. Outcome BM Congruence b SE t p LLCI ULCI Intercept 0.08 0.08 .96 0.34 -0.08 0.24 Sexual Orientation -.15 0.12 -1.30 0.19 -0.38 0.08 Withdrawal b SE t p LLCI ULCI Intercept 1.97 0.06 33.83 <.001 1.86 2.09 Sexual Orientation 0.16 0.08 1.95 0.05 0.00 0.33 BM Congruence -.24 0.04 -5.67 <.001 -0.32 -0.16 NOTE: BM = “Boundary Management.” 83 Table 22. PROCESS Model 4 results of sexual orientation predicting job satisfaction through boundary management congruence. Outcome BM Congruence b SE t p LLCI ULCI Intercept 0.27 0.19 1.42 0.15 -0.10 0.63 Sexual Orientation -0.18 0.12 -1.50 0.13 -0.41 0.05 Job Satisfaction b SE t p LLCI ULCI Intercept 5.47 0.29 19.19 <.001 4.91 6.03 Sexual Orientation -0.33 0.18 -1.81 0.07 -0.69 0.03 BM Congruence 0.54 0.09 5.89 <.001 0.36 0.72 NOTE: BM = “Boundary Management.” 84 Table 23. Standardized structural model results for relationships between sexual orientation and Study 2 outcomes through boundary management congruence indices. Outcome Predictor Estimate S.E. P-Value Job Satisfaction BM Congruence 0.34 0.06 <.001 Subjective Wellbeing BM Congruence 0.35 0.05 <.001 Withdrawal BM Congruence -0.31 0.05 <.001 BM Congruence (JS) Sexual Orientation -0.09 0.06 0.116 BM Congruence (SWB) Sexual Orientation -0.14 0.06 0.017 BM Congruence (W) Sexual Orientation -0.08 0.06 0.206 NOTE. BM = Boundary management, JS = Job satisfaction, SWB = Subjective wellbeing, W = Withdrawal. Sexual Orientation coded as 0 = “Heterosexual,” 1 = “Lesbian, Gay or Bisexual.” 85 Figure 1. Mapping of reviewed identity management techniques across a general spectrum of identity management. 86 Figure 2. Proposed model predicting enacted boundary management via identity management, boundary management preferences, and job characteristics, as it relates to individual outcomes depending on ideal segmentation preferences. 87 Figure 3. Proposed model linking sexual orientation to level of congruence between preferred and enacted boundary management as predictive of attitudinal outcomes. NOTE. “LGB” = lesbian, gay and bisexual. “BM” = boundary management. 88 APPENDIX B Study Measures Identity Management (Button, 2004) INSTRUCTIONS: Please rate the extent to which you agree with each of the following statements using the provided scale. Scale: 1 = “Strongly Disagree,” 2 = “Disagree,” 3 = “Slightly Disagree,” 4 = “Neither Agree nor Disagree,” 5 = “Slightly Agree,” 6 = “Agree,” 7 = “Strongly Agree” Avoiding/Distancing 1. I avoid situations where heterosexual coworkers are likely to ask me personal questions. 2. I let people know that I find personal questions to be inappropriate so that I am not faced with them. 3. I avoid personal questions by never asking others about their personal lives. 4. In order to keep my personal life private, I refrain from “mixing business with pleasure.” 5. I withdraw from conversations when the topic turns to things like dating or interpersonal relationships. 6. I let people think I am a “loner,” so that they won’t question my apparent lack of a relationship. Integrating/Affirming 1. In my daily activities, I am open about my homosexuality whenever it comes up. 2. Most of my coworkers know that I am gay. 3. Whenever I’m asked about being gay, I always answer in an honest and matter-of-fact way. 4. It’s okay for my gay and lesbian friends to call me at work. 5. My coworkers know of my interest in gay and lesbian issues. 6. I look for opportunities to tell my coworkers that I am gay/lesbian. 7. When a policy or law is discriminatory against gay men and lesbians, I tell people what I think. 8. I let my coworkers know that I’m proud to be lesbian/gay. 9. I openly confront others when I hear a homophobic remark or joke. 10. I display objects (e.g. photographs, magazines, symbols) which suggest that I am gay. 89 Boundary Management (Adapted from Kreiner, 2006) Boundary Management Preferences INSTRUCTIONS: Please rate the extent to which you agree that each of the following statements reflect your ideal preferences. Scale: 1 = “Strongly Disagree,” 2 = “Disagree,” 3 = “Slightly Disagree,” 4 = “Neither Agree nor Disagree,” 5 = “Slightly Agree,” 6 = “Agree,” 7 = “Strongly Agree” Original Items: 1. I don’t like to have to think about work while I’m at home. 2. I prefer to keep work life at work. 3. I don’t like work issues creeping into my home life. 4. I like to be able to leave work behind when I go home. Adapted Additional Items: 5. I don’t like to have to think about my non-work life when I’m at work. 6. I prefer to keep my non-work life outside of work. 7. I don’t like issues from my life outside of work creeping into my work life. 8. I like to be able to leave my non-work life behind when I go to work. Enacted Boundary Management INSTRUCTIONS: Please rate the extent to which you agree that each of the following statements reflect your true behaviors. Scale: 1 = “Strongly Disagree,” 2 = “Disagree,” 3 = “Slightly Disagree,” 4 = “Neither Agree nor Disagree,” 5 = “Slightly Agree,” 6 = “Agree,” 7 = “Strongly Agree” Original Items: 1. I don’t think about work while I’m at home. 2. I keep work life at work. 3. I don’t let work issues creep into my home life. 4. I leave work behind when I go home. Adapted Additional Items: 5. I don’t think about my non-work life when I’m at work. 6. I keep my non-work life outside of work. 7. I don’t let issues from my life outside of work creep into my work life. 8. I leave my non-work life behind when I go to work. 90 Perceptions of Organizational Diversity Climate (McKay, Avery & Morris, 2008) INSTRUCTIONS: Please rate the extent to which you agree with each of the following statements using the provided scale. Scale: 1 = “Strongly Disagree,” 2 = “Disagree,” 3 = “Neither Agree nor Disagree,” 4 = “Agree,” 5 = “Strongly Agree” 1. I trust my company to treat me fairly. 2. My company maintains a diversity-friendly work environment. 3. My company respects the views of people like me. 4. Top leaders demonstrate a visible commitment to diversity. 91 Job Autonomy (Thomas & Prottas, 2005) INSTRUCTIONS: Please rate the extent to which you agree with each of the following statements using the provided scale. Scale: 1 = “Strongly Disagree,” 2 = “Disagree,” 3 = “Neither Agree nor Disagree,” 4 = “Agree,” 5 = “Strongly Agree” 1. I have the freedom to decide what I do on my job. 2. I have a lot to say about what happens on my job. 3. I decide when I take breaks. 4. It is basically my own responsibility to decide how my job gets done. 92 Satisfaction with Life Scale (Deiner, Emmons, Larsen & Griffin, 1985) INSTRUCTIONS: Please rate the extent to which you agree with each of the following statements using the provided scale. 1 = “Strongly Disagree,” 2 = “Disagree,” 3 = “Slightly Disagree,” 4 = “Neither Agree nor Disagree,” 5 = “Slightly Agree,” 6 = “Agree,” 7 = “Strongly Agree” 1. In most ways, my life is close to my ideal. 2. The conditions of my life are excellent. 3. I am satisfied with my life. 4. So far, I have gotten the important things I want in life. 5. If I could live my life over, I would change almost nothing 93 Job Satisfaction (Cammann, Fichman, Jenkins & Klesh, 1983) INSTRUCTIONS: Please rate the extent to which you agree with each of the following statements using the provided scale. 1 = “Strongly Disagree,” 2 = “Disagree,” 3 = “Slightly Disagree,” 4 = “Neither Agree nor Disagree,” 5 = “Slightly Agree,” 6 = “Agree,” 7 = “Strongly Agree” 1. All in all, I am satisfied with my job. 2. In general, I don’t like my job. 3. In general, I like working at my job. 94 Withdrawal (Hanisch & Hulin, 1990) INSTRUCTIONS: Please indicate how often you engage in each behavior on the following scale: 1 = "Never,” 2 = “Rarely,” 3 = “Sometimes,” 4 = “Frequently,” 5 = “Always” 1. Made excuses to miss meetings. 2. Drank alcohol after work because of things that happened at work. 3. Stayed home from work when you had even a minor illness. 4. Took frequent or long breaks. 5. Made excuses to go somewhere to avoid the workplace. 6. Went to work late. 7. Did not work to the best of your ability. 8. Wanted to leave work early. 9. Spent time on non-work activities (e.g. talking, emailing, web browsing) while at work. 10. Ignored non-essential tasks. 11. Thought about leaving the organization. 12. Tried to find another job. 13. Made plans to leave the organization. 95 REFERENCES 96 REFERENCES Ammons, S. K. (2013). Work-family boundary strategies: Stability and alignment between preferred and enacted boundaries. Journal of Vocational Behavior, 82(1). 49-58. Ashforth, B. E., & Kreiner, G. E. (1999). “How can you do it?”: Dirty work and the challenge of constructing a positive identity. Academy of Management Review, 24, 413-434. Ashforth, B. E., Kreiner, G. E., & Fugate, M. (2000). All in a day’s work: Boundaries and micro role transitions. Academy of Management Review, 25(3), 472-491. Anderson, M. Z., Croteau, J. M., Chung, Y. B., & DiStefano, T. M. (2001). Developing an assessment of sexual identity management for lesbian and gay workers. Journal of Career Assessment, 9(3), 243-260. Arena Jr., D. F., & Jones, K. P. (2017). To “B” or not to “B”: Assessing the disclosure dilemma of bisexual individuals at work. Journal of Vocational Behavior, 103, 86-98. Aycan, Z. (2008). Cross-cultural perspectives to work-family conflict. In K. Korabik, D., Lero, & D. Whitehead (Eds.), Handbook of work-family integration. Cambridge: Cambridge University Press. Boswell, W. R., & Olson-Buchanan, J. B. (2004). Experiencing mistreatment at work: The role of grievance filing, nature of mistreatment, and employee withdrawal. Academy of Management Journal, 47(1), 129-139. Brauner, C., Wöhrmann, A. M., & Michel, A. (2020). Congruence is not everything: A response surface analysis on the role of fit between actual and preferred working time arrangements for work-life balance. Chronobiology International, 37(9-10), 1287-1298. Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazon's Mechanical Turk a new source of inexpensive, yet high-quality, data? Perspectives on Psychological Science, 6(1), 3-5. Button, S. B. (2001). Organizational efforts to affirm sexual diversity: A cross-level examination. Journal of Applied Psychology, 86, 17-28. Button, S. B. (2004). Identity management strategies utilized by lesbian and gay employees: A quantitative investigation. Group & Organization Management, 29(4), 470-494. Cable, D. M., & Edwards, J. R. (2004). Complementary and supplementary fit: A theoretical and empirical integration. Journal of Applied Psychology, 89(5), 822-834. Cammann, C., Fichman, M., Jenkins, G. D., & Klesh, J. R. (1983). Assessing the attitudes and perceptions of organizational members. In S. E. Seashore, E. E. Lawler, P. H., Mirvis, & 97 C. Cammann (Eds.), Assessing organizational change: A guide to methods, measures and practices. New York: Wiley. Carlson, D. S., Ferguson, M., & Kacmar, K. M. (2016). Boundary management tactics: An examination of the alignment with preferences in the work and family domains. Journal of Behavioral and Applied Management, 16(2), 51-70. Chrobot-Mason, D., Button, S. B., & DiClementi, J. D. (2001). Sexual identity management strategies: An exploration of antecedents and consequences. Sex Roles, 45, 321-336. Clair, J. A., Beatty, J. E., & Maclean, T. L. (2005). Out of sight but not out of mind: Managing invisible social identities in the workplace. Academy of Management Review, 30(1), 78- 95. Clark, S. C. (2000). Work/family border theory: A new theory of work/family balance. Human Relations, 53(6), 747-770. Corrington, A., Nittrouer, C. L., Trump-Steele, R. C., & Hebl, M. (2019). Letting him B: A study on the intersection of gender and sexual orientation in the workplace. Journal of Vocational Behavior, 113, 129-142. Crandall, C. S., & Coleman, R. (1992). AIDS-related stigmatization and the disruption of social relationships. Journal of Social and Personal Relationships, 9(2), 163-177. Cumming, G. (2009). Inference by eye: Reading the overlap of independent confidence intervals. Statistics in Medicine, 28(2), 205-220. Desrochers, S., Sargent, L. D., & Hostetler, A. J. (2012). Boundary-spanning demands, personal mastery, and family satisfaction: Individual and crossover effects among dual-earner parents. Marriage & Family Review, 98(5), 443-464. Diener, E. D., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The satisfaction with life scale. Journal of Personality Assessment, 49(1), 71-75. Driscoll, J. M., Kelley, F. A., & Fassinger, R. E. (1996). Lesbian identity and disclosure in the workplace: Relation to occupational stress and satisfaction. Journal of Vocational Behavior, 48(2), 229-242. Edwards, J. R., & Parry, M. E. (1993). On the use of polynomial regression equations as an alternative to differences scores in organizational research. Academy of Management Journal, 36(6), 1577-1613. Faul, F., Erdfelder, E., Lang, A. & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39, 175-191. 98 Feldman, D. B., & Crandall, C. S. (2007). Dimensions of mental illness stigma: What about mental illness causes social rejection? Journal of Social and Clinical Psychology, 26(2), 137-154. Gardner, D. M., Lauricella, T., Ryan, A. M., Wadlington, P. L., & Elizondo, F. (2021). Managing boundaries between work and non-work domains: Personality, job characteristics and adopted style. Journal of Occupational and Organizational Psychology, 94(1), 132-159. Gates, G. J. (2011). How many people are lesbian, gay, bisexual and transgender? Retrieved from: https://williamsinstitute.law.ucla.edu/publications/how-many-people-lgbt/ Goffman, E. (1963). Stigma: Notes on the management of spoiled identity. Englewood Cliffs, NJ: Prentice Hall. Hackman, J. R., & Oldham, G. R. (1975). Development of the job diagnostic survey. Journal of Applied Psychology, 60(2), 159-170. Hackman, J. R., & Oldham, G. R. (1976). Motivation through the design of work: Test of a theory. Organizational Behavior and Human Performance, 16, 250-276. Hanisch, K. A., & Hulin, C. L. (1990). Job attitudes and organizational withdrawal: An examination of retirement and other voluntary withdrawal behaviors. Journal of Vocational Behavior, 37(1), 60-78. Harari, M. B., Thompson, A. H., & Viswesvaran, C. (2018). Extraversion and job satisfaction: The role of trait bandwith and the moderating effect of status goal attainment. Personality and Individual Differences, 123, 14-16. Hauser, D. J., & Schwarz, N. (2016). Attentive turkers: MTurk participants perform better on online attention checks than do subject pool participants. Behaviors Research Methods, 48(1), 400-407. Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis. New York, NY: Guilford Press. Hebl, M. R., & Kleck, R. E. (2002). Acknowledging one’s stigma in the interview setting: Effective strategy or liability? Journal of Applied Social Psychology, 32(2), 223-249. Hebl, M., Tickle, J., & Heatherton, T. F. (2000). Awkward moments. In T. F. Heatherton, R. E. Kleck, M. Hebl & J.G. Hull (Eds). Stigma: Social psychological processes New York: Guilford. Hegarty, P. (2002). ‘It’s not a choice, it’s the way we’re built’: Symbolic beliefs about sexual orientation in the US and Britain. Journal of Community & Applied Social Psychology, 12(3), 153-166. 99 Hegarty, P., & Golden, A. M. (2008). Attributional beliefs about the controllability of stigmatized traits: Antecedents or justifications of prejudice? Journal of Applied Social Psychology, 38(4), 1023-1044. Holman, E. G. (2018). Theoretical extensions of minority stress theory for sexual minority individuals in the workplace: A cross-contextual understanding of minority stress processes. Journal of Family Theory & Review, 10(1), 165-180. Jones, E., Farina, A., Hastorf, A., Markus, H., Miller, D., & Scott, R. (1984). Social stigma: The psychology of marked relationships. New York, NY: Freeman. Judge, T. A., & Zapata, C. P. (2015). The person-situation debate revisited: Effect of situation strength and trait activation on the validity of the Big Five personality traits in predicting job performance. Academy of Management Journal, 58(4), 1149-1179. King, E. B., Reilly, C., & Hebl, M. (2008). The best of times, the worst of times: Exploring dual perspectives of “coming out” in the workplace. Group & Organization Management, 33(5), 566-601. Kossek, E., E., Lautsch, B. A., & Eaton, S. C. (2006). Telecommuting, control, and boundary management: Correlates of policy use and practice, job control, and work-family effectiveness. Journal of Vocational Behavior, 68(2), 347-367. Kreiner, G. E. (2006). Consequences of work-home segmentation or integration: A person- environment fit perspective. Journal of Organizational Behavior, 27, 485-507. Kristof-Brown, A., Jansen, K., & Colbert, A. (2002). A policy-capturing study of the simultaneous effects of fit with jobs, groups, and organizations. Journal of Organizational Behavior, 14, 61-81. Lyons, B. J., Pek, S., & Wessel, J. L. (2017). Toward a “Sunlit Path:” Stigma identity management as a source of localized social change through interaction. Academy of Management Review, 42(4), 618-636. Lyons, B. J., Volpone, S. D., Wessel, J. L., & Alonso, N. M. (2017). Disclosing a disability: Do strategy type and onset controllability make a difference? Journal of Applied Psychology, 102(9), 1375-1383. Lyons, B., Wessel, J., Ghumman, S., Ryan, A. M., & Kim, S. (2014). Applying models of employee identity management across cultures: Christianity in the USA and South Korea. Journal of Organizational Behavior, 35(5), 678-704. Madera, J. M., Dawson, M., & Neal, J. A. (2013). Hotel managers’ perceive diversity climate and job satisfaction: The mediating effects of role ambiguity and conflict. International Journal of Hospitality Management, 35, 28-34. 100 McKay, P. F., Avery, D. R., & Morris, M. A. (2008). Mean racial-ethnic differences in employee sales performance. Personnel Psychology, 61(2), 349-374. McKay, P. F., Avery, D. R., Tonidandel, S., Morris, M. A., Hernandez, M., & Hebl, M. R. (2007). Racial differences in employee retention: Are diversity climate perceptions the key? Personnel Psychology, 60(1), 35-62. Mellner, C., Aronsson, G., & Kecklund, G. (2014). Boundary management preferences, boundary control, and work-life balance among full-time employed professionals in knowledge-intensive, flexible work. Nordic Journal of Working Life Studies, 4(4), 7-23. Mischel, W. (1977). On the future of personality measurement. American Psychologist, 32(4), 246-254. Mor Barak, M. E., Lizano, E. L., Kim, A., Duan, L., Rhee, M. K., Hsiao, H. Y., & Brimhall, K. C. (2016). The promise of diversity management for climate of inclusion: A state-of-the- art review and meta-analysis. Human Service Organizations: Management, Leadership & Governance, 40(4), 305-333. Mustanski, B. S., Chivers, M. L., & Bailey, J. M. (2002). A critical review of recent biological research on human sexual orientation. Annual Review of Sex Research, 13(1), 89-140. Ng, T. W., & Feldman, D. C. (2010). The relationships of age with job attitudes: A meta- analysis. Personnel Psychology, 63(3), 677-718. Nippert-Eng, C. E. (1996). Home and work. Chicago, IL: The University of Chicago Press. Nye, C. D., Prasad, J., Bradburn, J., & Elizondo, F. (2018). Improving the operationalization of interest congruence using polynomial regression. Journal of Vocational Behavior, 104, 154-169. Pavot, W., Diener, E. D., Colvin, C. R., & Sandvik, E. (1991). Further validation of the Satisfaction with Life Scale: Evidence for the cross-method convergence of well-being measures. Journal of Personality Assessment, 57(1), 149-161. Perry, B. L. (2014). Symptoms, stigma, or secondary social disruption: Three mechanisms of network dynamics in severe mental illness. Journal of Social and Personal Relationships, 31(1), 32-53. Pew Research Center (2013). A survey of LGBT Americans. Retrieved from: https://www.pewsocialtrends.org/2013/06/13/a-survey-of-lgbt-americans/ Podsakoff, P. M., MacKenzie, S. C., Lee, J., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879-903. 101 Powell, G. N., & Greenhaus, J. H. (2010). Sex, gender, and the work-to-family interface: Exploring negative and positive interdependencies. Academy of Management Journal, 53(3), 513-534. Ragins, B. R. (2008). Disclosure disconnects: Antecedents and consequences of disclosing invisible stigmas across life domains. Academy of Management Review, 33(1), 194-215. Ragins, B. R., & Cornwell, J. M. (2001). Pink triangles: Antecedents and consequences of perceived workplace discrimination against gay and lesbian employees. Journal of Applied Psychology, 86(6), 1244-1261. Ragins, B. R., Singh, R., & Cornwell, J. M. (2007). Making the invisible visible: Fear and disclosure of sexual orientation at work. Journal of Applied Psychology, 92(4), 1103- 1118. Riek, B. M., Mania, E. W., & Gaertner, S. L. (2006). Intergroup threat and outgroup attitudes: A meta-analytic review. Personality and Social Psychology Review, 10(4), 336-353. Roberson, Q. M. (2012). Managing diversity. In S. W. J. Kozlowski (Ed.), Oxford handbook of industrial and organizational psychology. New York: Oxford University Press. Robinson, S. L., & Rousseau, D. M. (1994). Violating the psychological contract: Not the exception but the norm. Journal of Organizational Behavior, 15(3), 245-259. Rothbard, N. P., Phillips, K. W., & Dumas, T. L. (2005). Managing multiple roles: Work-family policies and individuals’ desires for segmentation. Organization Science, 16(3), 243-258. Ruybal, A. L., & Siegel, J. T. (2019). Attribution theory and reducing stigma toward women with postpartum depression: Examining the role of perceptions of stability. Stigma and Health, 4(3), 320-329. Sawyer, K. B., Thoroughgood, C., & Ladge, J. (2017). Invisible families, invisible conflicts: Examining the added layer of work-family conflict for employees with LGB families. Journal of Vocational Behavior, 103, 23-39. Shaw, J. D. (1999). Job satisfaction and turnover intentions: The moderating role of positive affect. The Journal of Social Psychology, 139(2), 242-244. Shore, L. M., Randel, A. E., Chung, B. G., Dean, M. A., Holcombe Ehrhart, K., & Singh, G. (2011). Inclusion and diversity in work groups: A review and model for future research. Journal of Management, 37(4), 1262-1289. Smith, N. A., Sabat, I. E., Martinez, L. R., Weaver, K., & Xu, S. (2015). A convenient solution: Using MTurk to sample from hard-to-reach populations. Industrial and Organizational Psychology: Perspectives on Science and Practice, 8(2), 220-228. 102 Stewart, R., Volpone, S. D., Avery, D. R., & McKay, P. (2011). You support diversity, but are you ethical? Examining the interactive effects of diversity and ethical climate perceptions on turnover intentions. Journal of Business Ethics, 100(4), 581-593. Thompson, C. A., & Prottas, D. J. (2006). Relationships among organizational family support, job autonomy, perceived control, and employee well-being. Journal of Occupational Health Psychology, 11(1), 100-118. Tiggeman, M., & Anesbury, T. (2000). Negative stereotyping of obesity in children: The role of controllability beliefs. Journal of Applied Social Psychology, 30(9), 1977-1993. Tilcsik, A., Anteby, M., & Knight, C. R. (2015). Concealable stigma and occupational segregation: Toward a theory of gay and lesbian occupations. Administrative Science Quarterly, 60(3), 446-481. Tzelgov, J., & Henik, A. (1991). Suppression situations in psychological research: Definitions, implications, and applications. Psychological Bulletin, 109(3), 524-536. Verquer, M. L., Beehr, T. A., & Wagner, S. H. (2003). A meta-analysis of relations between person-organization fit and work attitudes. Journal of Vocational Behavior, 63(3), 473- 489. Weiner, B., Perry, R. P., & Magnusson, J. (1988). An attributional analysis of reactions to stigmas. Journal of Personality and Social Psychology, 55(5), 738-748. Williamson, R. L., Beiler-May, A., Locklear, L. R., & Clark, M. A. (2017). Bringing home what I’m hiding at work: The impact of sexual orientation disclosure at work for same-sex couples. Journal of Vocational Behavior, 103, 7-22. Zurbrügg, L., & Miner, K. N. (2016). Gender, sexual orientation, and workplace incivility: Who is most targeted and who is most harmed? Frontiers in Psychology, 7, 1-12. 103