By Deepshikha Chatterjee A DISSERTATION IDENTITY MANAGEMENT AS A GOAL-ORIENTED RESPONSE TO STIGMA COMMUNICATION AT WORK: A SELF-REGULATION PERSPECTIVE Submitted to(cid:1) Michigan State University(cid:1) in partial fulfillment of the requirements for the degree of Psychology–Doctor of Philosophy 2018 ABSTRACT IDENTITY MANAGEMENT AS A GOAL-ORIENTED RESPONSE TO STIGMA COMMUNICATION AT WORK: A SELF-REGULATION PERSPECTIVE By Deepshikha Chatterjee In responding to stigma communication at work, people are often under the influence of environmental constraints that may significantly curtail the availability of different identity management strategies. This dissertation highlights that identity management in workplace settings should be seen as a goal-driven and norm-guided response. I present a conceptual model that opens up the black box of why people choose to accept or challenge stigma communication at work by integrating identity management and stigma literatures with self-regulation literature. Three multi-method studies were crafted. It was found that work environments indeed constrain possible identity management strategies in response to stigma communication. Under the press of self-enhancement goals women were more likely to accept stigma communication, and accepting stigma communication was implicated in higher levels of depletion and strain. Further, women were (a) more likely to choose accepting identity management strategies when stigma was communicated by their supervisors and clients, (b) less likely to challenge when clients communicated stigma, and (c) more likely to use challenge identity management strategies when they had high levels of coping self-efficacy, and also when they were high on promotion focus. Future research questions and practical recommendations are presented to help practitioners harness the value of diversity in organizations. This dissertation is dedicated to you, Prith. May we always help each other answer our callings. iii ACKNOWLEDGEMENTS I would like to thank my advisor, Dr. Ann Marie Ryan, for her advice and guidance that made this dissertation possible. Ann Marie, I owe you a huge debt of gratitude. I had thought that by the end of the five years, I would have found just the right words to express how thankful I am for all your contributions to my personal and professional growth. Instead I still find myself unable to formulate those sentences. I am utterly overwhelmed by how central you have been to my journey so far, and how foundational you will be as I navigate life after graduate school. How might I, in a few words, thank you for showing me the ropes of research by encouraging me to play with ideas, for putting in place systems that could support my growth, for being incisive in your feedback, or for your prompt turnaround times? How might I ever thank you for your faith in my capabilities when even I doubted myself? With your every word and action, you sowed the seeds of confidence and self-efficacy in me. When I needed a helping hand, you were always there for me as my advisor, guide, mentor, champion, and role model. When I ran into walls, you ensured that I was never denied opportunities to make my dreams possible. So how do I thank you for having defined my doctoral experience, and for becoming the compass that guides how I conduct myself in the world? Perhaps all I can say, Ann Marie, is that knowing you and working alongside you has been a privilege, and for that I am eternally grateful. I would like to thank Dr. J. Kevin Ford for serving on my dissertation committee. Kevin, your authentic leadership and guidance have been the mainstays of my experience at MSU. You have shared the riches of wisdom and learning with every single conversation – informal and formal – and you have generously shared your time to have hundreds of such conversations. In you, I have found the prototype of the leader I’d like to be. Your way of seeing through complexity and distilling elegant insights is an art form and I will continue my efforts to iv try and replicate that. In you, I have found the teacher I dream to be someday. If I could inspire a class the way you do, I know I will have amounted to a force for good in the world. You have been my second advisor all along. In you, I have found a mentor who instinctively responded to my need to grow independently. You trusted me enough to honor that independent streak even when the consequences of error were high. Thank you for placing such faith in me, it has mattered more than you know! I would also like to extend my deep gratitude to Dr. Frederick T.L. Leong for serving on my dissertation committee. Fred, I have benefitted from your deep knowledge in diverse domains, but perhaps more importantly, I have benefitted from your insightful thoughts on my own career reflections. As a seasoned mentor, you never once told me what to do, but were instrumental in helping me define my career path nonetheless. I am deeply thankful for how you have helped me connect the dots of my own needs, values, and interests and the resultant gift of clarity. Your name will always shine bright as a prominent influence in my journey. Finally, I would like to thank Dr. Christopher Nye for serving on my dissertation committee. Chris, I have knocked on your door for advice on multiple projects, and you have always provided direction. Many times, I walked in with problems that I could barely define clearly. What I appreciated most in those moments was your ability to slice through the hazy problem statement to arrive at a clear definition and an elegant solution. I would like to thank you for your insights that helped progress my work. But most of all, I would like to thank you for how you converted these moments into learning opportunities for me. This dissertation would not be possible without the continual support and encouragement of my best friend, Prith (and the guy I had the wisdom to marry). Prith, I distinctly remember the day when I said, “Thy shalt not check out” in response to your tardiness in reviewing my v statement of purpose for graduate school. Over the past five years you have outdone yourself in staying engaged with every aspect of my journey. You have spent countless Skype dates hearing me talk about research. When you found something boring or too technical, you patiently let me explain the constructs until you too could enjoy the ideas as much as I did. You were ready to offer help where you could, and where you could not, you helped by being present. You were present on nights when you were tired, on days when you were swamped with your own work commitments, and on weekends when you could have been out exploring the world instead. You have walked this road with me not because of perceived obligations, but with genuine interest and excitement. Your strong desire to see me succeed meant that you supported even the most eccentric of my requests such as honoring my wish to study for the comprehensives at the arboretum out of our parked car! You never once questioned why I had to work on weekends or got upset about my need to carve out time for work even on our vacations. Instead, you made it your mission to find cafes around all our vacation stops so I could stay productive. It did not matter if we were sitting by river streams, watching crackling camp fires, or hiking up mountains – you championed every discussion, providing astute and discerning criticisms of my ideas. In all those moments, I was deeply aware of my privilege. I knew I was in the presence of a love that sought to give unconditionally. Seeing you treat my goals as yours, and watching you create pathways for my success has bolstered my belief that we are a partnership to be reckoned with. Thank you for ensuring that the thousands of miles between us are fashioned into a unique force that continues to strengthen us as a team. You truly do deserve that t-shirt on surviving my PhD and if I could, I would share one-fourth of my degree with you for you absolutely deserve it. vi There are some who will never read this acknowledgement. Yet, I must thank them for they are my great loves, without whom this work could never be completed. I have found solace in their daily conversations, constant companionship, and enduring love. Seeing their faces lights up my world and I am never quite at home, or quite as happy, until I have them by my side. I have thoroughly enjoyed exploring new vistas and horizons with them. To my little Ginger and Bonnie: I have relished coming back home to your shenanigans. Thank you for your paw prints that are all over this dissertation. You have both bounded around in the pages of my initial ideas, and together you have once conspired to kill the entire literature review. Your faces became a constant background to the words herein as you peered at me from behind the laptop begging for some attention. I thank you for all those times for you always know how to make me smile. Gin, you are the golden cloud of sunshine that I wake up to, and Bon you are the glistening moonbeam that makes for a restful slumber. Between the two of you, my days are made perfect. To know who I am, exactly as I am, I have always counted on my sister, Chitralekha Yadav. Chin, you are the little sister that every elder sister can only dream to have. I watch you in awe as you strive to support and build those around you, making sure that what you say with your words you deliver in deeds. I cannot help but take pride in how you have grown up to be such a fiercely independent and capable individual. Each time I look at how far we have come, I marvel at our journey that was stacked with challenges from the get go. It was not that long ago when these challenges had seemed insurmountable, and yet, here we are today: I credit your intellect, generosity, wit, and wise counsel in helping me ride out all the difficult times gracefully. I am thankful that in you I had a general who fought on the frontlines with me, for Chin, that has made all the difference. vii In the friendships that shelter me, I count your name as a blessing, Jessica Webb. You have played many roles in my life so far. You have been a fabulous peer, a loving friend, and a trusted confidant. Your generous spirit, infectious can-do attitude, and your incisive mind are just some of the attributes that I cherish. You have also been the person I reach out to for the comforts of a reliable transactive memory system. But above all, I’d like to thank you Jess for treating me like a sister. Thank you for the gift of your ready rebukes when I was being too hard on myself, and for your ready hugs when life was being too hard on me. Now that we have tied a bow on this chapter of our lives, I am acutely aware that we will spend more time away than together. That said, I also know that the roots of our friendship go deep, that we will never need calendars to schedule conversations, and that we will always have a safe harbor in each other. To the three pillars who have supported me steadfastly over the years – Santi, Deepak, and Priya – we started out as friends, but along the way we became a family. You three have known my dream in its most nascent state. You have contributed to shaping it with your constant gifts of friendship, love, and laughs. These five years, you have always understood what I was working towards, never made demands on my time, and never lumped any guilt on me for having to be so internally focused. Instead, when things seemed particularly uphill for me, you reminded me to believe in myself, and you chose to celebrate every success that came my way like it was your own. Thank you for being such an incredible support system: at the end of long, hard days, yours are the voices I want to hear, and yours are the faces I want to see. I would like to thank my parents, Gaya Prasad Yadav and Pushpa Yadav, my aunt, Saraswati Deswal, and my grandparents, Ved Kaur and Sumer Singh Deswal, who in their infinite wisdom taught me the values of intellectual curiosity, persistence, and self- determination. To my grandma: Nani, I miss your voice skipping over the airwaves, and viii especially miss hearing your “I labloo [sic]”. I always joked about how you could never master the correct pronunciation of “I love you”, but you were a master practitioner of unconditional love. You taught me that true love does not seek to possess, control, or find flaws, instead it is an intentional act of generosity, kindness, and building one another. Nani, with this singular lesson, you ensured that my life could be filled with genuine love and happiness for that which I share with others, I get back ten-fold. To my grandpa: Nana, thank you for being a patient teacher and a fabulous storyteller. I remember tracing the veins on your hands as a child and pretending that they were an ancient roadmap to far off lands, with treasures that were waiting to be explored. I now know that the treasures were right there under my fingertips in the form of your engaged parenting until the very end. To my aunt: Bhappi, you were an aunt in name, but you executed all the duties of a loving mother. You have played a critical role in the formative years of my life and any success I find today is predicated on your efforts. To mom and dad: thank you for the opportunities to learn the life skills of dogged determination and individual striving. You prepared me to thrive no matter what the external conditions. I would like to thank two of my biggest champions in Prith’s family: Ma, Purnima Chattopadhyay, thank you for all your prayers on my behalf and constant supply of love. Your blessings and support have meant the world to me. To Bhaiya, Pinaki Chattopadhyay, thank you for your constant encouragement over the years, and for always taking such unbridled joy and pride in my accomplishments. If I had all the space, I would have described the many contributions of my dearest friends included here. Unfortunately, this acknowledgement is already too long, so suffice it to say that I have been enriched by the wisdom and love that each one of you have so generously shared with me: Sarena Bhatia, Abdifatah Ali, Samantha Perry, Brent Lyons, Gargi Sawhney, Daniel Pickhardt, and Anita Keller thanks for being the inspiring figures who showed me the ix path forward each time I came to you with questions on life. Thanks also to the members of the Diversity Lab, and a special shout out to all the research assistants who have worked tirelessly on my many projects: Qi, Janani, Kate, Meg, Abby, Casey, Justin, and Megan. I would also like to thank the mentors who prepared me for MSU. At Harold Washington College, Dr. Les White, Ms. LaRhue Finney, Dr. Art Divito, Mr. Kamran Swanson, and Dr. Athan Vouzinias; and at Illinois Institute of Technology, Dr. Kemp Ellington, Dr. Roya Ayman, Dr. Ron Landis, and Mr. Steve Beck – you each have contributed in many ways to ensuring that I was MSU-ready. Some of you taught me how to write effectively and how to think critically, some of you shared the gifts of math, calculus, and statistics with me, others taught me the basics of organizational psychology research, and yet others who shared their life lessons and wisdom. I could not have made it this far without your enthusiastic investment in me. Finally, I would be remiss if I did not acknowledge the huge role that the United States of America has played in my journey, for it is here that I encountered mentors and teachers who identified my potential and laid the foundation for my growth. Like a loving family, the US accepted me in its fold, offered me safety, treated me with respect, and supported my pursuit of Happiness. For me, this is home, this is family, and this indeed is the shining city upon a hill where dreams can and do come true. It is my firm belief that my story would not have been possible anywhere but here. I am ever so grateful to have received such love and generosity, and ardently hope that I am able to reciprocate that as I move into the next phase of my life. x TABLE OF CONTENTS LIST OF TABLES................................................................................................................... xiv LIST OF FIGURES ................................................................................................................. xvi INTRODUCTION ...................................................................................................................... 1 CHAPTER 1: PROPOSING A NEW IDENTITY MANAGEMENT CONCEPTUALIZATION 5 Key Self-regulatory Identity Management Models .......................................................... 8 Self-regulation and Goals: A Brief Overview ................................................................ 10 Why Consider Self-Regulation in Responses to Stigma Communication? ..................... 14 Identity Centrality and Accepting and Challenging Stigma Communication .................. 15 The Role of Self-regulatory Foci in Accepting or Challenging Stigma........................... 16 Identity Management Strategies: Norm-bound and Expectancy Driven ......................... 18 Accepting or Challenging Stigma Communication: Differential Depletion of Self- Regulatory Resources .................................................................................................... 20 Navigating Multiple Goals: Goal Choice Under Stigma Threat ..................................... 22 Goals and Differential Identity Management Strategies ................................................. 26 CHAPTER 2: DEFINING FOCUS OF THE PROPOSED RESEARCH ................................... 31 Bounding the Overall Theoretical Model ....................................................................... 31 An Overview of the Multi Study Approach ................................................................... 34 An Overview of Justification of Operationalization Choices in the Proposed Research .. 35 Conclusion .................................................................................................................... 44 CHAPTER 3: PILOT STUDY .................................................................................................. 45 Participants ................................................................................................................... 46 Pilot A: Piloting the Vivid Recall Task .......................................................................... 47 Measures ....................................................................................................................... 49 Results for Pilot A ......................................................................................................... 51 Preliminary Analyses ......................................................................................... 51 Discussion of Pilot A .................................................................................................... 61 Pilot B: Performance Discussion Task ........................................................................... 63 Results of Pilot B .......................................................................................................... 66 Discussion ..................................................................................................................... 69 CHAPTER 4: STUDY ONE ..................................................................................................... 70 Power Analysis ............................................................................................................. 72 Participants ................................................................................................................... 73 Design........................................................................................................................... 74 Procedure ...................................................................................................................... 74 Manipulation and Tasks ................................................................................................ 75 Measures ....................................................................................................................... 75 xi Results .......................................................................................................................... 77 Preliminary Analyses ......................................................................................... 77 Descriptive Statistics ......................................................................................... 79 Empirical Model: Main Effects and Mediation .................................................. 86 Hypothesis Testing ............................................................................................ 88 Discussion ..................................................................................................................... 92 Identity management strategies and self-regulatory depletion under stigma communication .................................................................................................. 93 Role of individual differences in choice of identity management strategies ........ 96 Role of context in choice of identity management strategies ............................ 101 Limitations ...................................................................................................... 102 CHAPTER 5: STUDY TWO .................................................................................................. 104 Power Analysis ........................................................................................................... 106 Participants ................................................................................................................. 107 Design......................................................................................................................... 107 Procedure .................................................................................................................... 108 Manipulation and Tasks .............................................................................................. 109 Measures ..................................................................................................................... 110 Empirical Model, Main Effects, and Mediation ........................................................... 111 Results ........................................................................................................................ 112 Preliminary Analyses ....................................................................................... 115 Descriptive Statistics ....................................................................................... 115 Hypothesis Testing .......................................................................................... 117 Proximal goals and participants’ choice of specific accepting or challenging IM strategies ......................................................................................................... 120 Supplementary analyses: Treating ordinal data as continuous data ................... 133 Discussion ................................................................................................................... 138 Identity management strategies and self-regulatory depletion .......................... 139 Role of individual differences in navigating stigma communication ................. 139 Role of Goals in choice of self-regulatory depletion and identity management strategies ......................................................................................................... 144 Limitations ...................................................................................................... 150 CHAPTER 6: STUDY THREE .............................................................................................. 154 Power analysis ............................................................................................................ 156 Screening Criteria ....................................................................................................... 158 Compensation Strategy ............................................................................................... 158 Participants ................................................................................................................. 160 Method ....................................................................................................................... 160 Event-level measures (Within-person; Level 1 variables). ........................................... 162 Analytic approach ....................................................................................................... 166 Results ........................................................................................................................ 169 Preliminary Analyses ....................................................................................... 169 Main Effects and Mediation ............................................................................. 172 Supplementary Analysis .................................................................................. 174 xii Discussion ................................................................................................................... 175 Stigma communication, role hierarchy, and outcomes...................................... 175 Stigma communication, daily goals, and identity management strategies ......... 176 Limitations ...................................................................................................... 177 CHAPTER 7: GENERAL DISCUSSION ............................................................................... 179 Identity management under stigma communication in work settings............................ 181 Identity management under stigma communication at work: Role of goals .................. 185 Identity management under stigma communication at work: Influence of role hierarchy ................................................................................................................................... 190 Identity management under stigma communication at work: Key outcomes ................ 193 Limitations and Future Research Directions ................................................................ 194 Practical Implications .................................................................................................. 204 Conclusion .................................................................................................................. 208 APPENDICES ........................................................................................................................ 210 APPENDIX A: RATIONALE ..................................................................................... 211 APPENDIX B: MAPPING .......................................................................................... 217 APPENDIX C: EXPERIMENTAL PROCESS FLOWS .............................................. 222 APPENDIX D: RATIONALE FOR SCALE ............................................................... 229 APPENDIX E: MEASURES USED ............................................................................ 233 APPENDIX F: VIVID RECALLS EXAMPLES ......................................................... 255 APPENDIX G: MPLUS SYNTAX ............................................................................. 256 APPENDIX H: TRAINING PROTOCOL ................................................................... 257 REFERENCES ....................................................................................................................... 266 xiii LIST OF TABLES Table 1: Exemplars of Identity Management Strategies When Self-enhancement, Self-verification, or Belonging Goal is More Valued and Proximal………………………………………………...28 Table 2: Means, standard deviations, reliabilities, and correlations for study variables (N=70). ……………………………………………………………………………………………………58 Table 3: Hypotheses for Study One.……………………………………………………………...71 Table 4: Means, standard deviations, skew, and kurtosis on variables in Study One……………..78 Table 5: Correlations between Study One variables.…………………………………………......81 Table 6: Examples of vivid recalls of sexist events at work for Study One……………………….82 Table 7: Sexist incident vivid recall qualitative data coded for Study One.……………………….84 Table 8: Mediation results of Study One (Indirect paths were specified in MPlus).……………...90 Table 9: Study Two hypotheses.……………………………………………………………...…105 Table 10: Means, standard deviations, skew, and kurtosis of Study Two variables. …………..115 Table 11: Study Two correlations. ……………………………………………………………..116 Table 12: Comparison of accepting and challenging identity management by goals…………..120 Table 13: Sub-strategies of accepting identity management strategies from exploratory model……………………………………………………………………………………………121 Table 14: Accepting strategies based on goals.………………………………………………...123 Table 15: Challenging identity management strategies from exploratory model. ……………..124 Table 16: Challenging strategies based on goals.………………………………………………126 Table 17: Accepting identity management strategies.………………………………………….128 Table 18: Challenging identity management strategies.………………………………………..129 Table 19: Coding open-ended data for new identity management strategies. …………………130 Table 20: Accepting identity management strategies treated as continuous variables.………...134 xiv Table 21: Challenging identity management strategies treated as continuous variables……….136 Table 22: Hypotheses for Study Three.………………………………………………………...156 Table 23: Means, Standard deviations, Skew, and Kurtosis of Study Three variables. ……….169 Table 24: Frequency of endorsement of accepting identity management strategies. ………….170 Table 25: Frequency of endorsement of challenging identity management strategies. ……….170 Table 26: Correlations of study variables. ……………………………………………………..171 Table 27: Mediation hypotheses for Study Three. ……………………………….…………….174 Table 28: Meta-analytic findings to inform task choice in the current research…..……………..212 Table 29: Literature review on how women may manage their identities.………………………217 Table 30: Identity Management Strategies items used in the current study.……………………..221 Table 31: Reported reliability and validity information for the unpublished SSCCS.…………..230 Table 32: Comparing the SRF-18 and the SSCCS scales.……………………………………….231 Table 33: Experimental condition responses Pilot A.………………………………………...…255 Table 34: Control condition responses Pilot A………………..………………………………...255 Table 35: Day to day sexist situations to train participants to recognize sexist events. ………….260 xv LIST OF FIGURES Figure 1: A conceptual model of identity management as self-regulated, goal-oriented, choice behaviors in organizational settings.……………...…….…………………………………………8 Figure 2: Heuristic flowchart of how people may navigate and choose from multiple goals ……...…………………………………………………………………………………………….25 Figure 3: A conceptual model……………………………..……………………………………...33 Figure 4: Boxplot for time spent on anagram tasks with outliers for experimental group ………... .…………………………………………………………………………..……………………….54 Figure 5: Q-Q plot for time spent on anagram tasks with outliers for experimental group …………………..………………………………………………………………………………..54 Figure 6: Boxplot for time spent on anagram tasks with outliers for control group ………………………..…………………………………………………………………………..55 Figure 7: Q-Q plot for time spent on anagram tasks with outliers for control group ………..…………………………………………………………………………………………..55 Figure 8: Scree plot using Varimax rotation for Identity Management Strategies Scale…… ………………………………………………………………………………………………..…..59 Figure 9: Conceptual model for Study One …………………………………………………..…..71 Figure 10: Emotions coded from qualitative data for Study One……………………………..…..85 Figure 11: Empirical paths with estimates for Study One……………………………………..….86 Figure 12: Conceptual model for Study Two.…………………………………………………...105 Figure 13: Empirical paths with estimates for Study Two……………………………………....112 Figure 14: Conceptual model for Study Three. ………………………………………................155 Figure 15: Empirical paths with estimates for Study Three……………………………………..168 xvi INTRODUCTION “At the end of the day, I chose to make a career in the United States. I chose to leave a country where everybody looked like me, and if somebody is uncomfortable dealing with me initially, I have to live with it…and overtime if people get more comfortable, that’s great, I have to help them get comfortable” – Indira Nooyi, CEO, Pepsico. Indira Nooyi’s statement about her goal-oriented, choice behaviors in navigating her stigmatized social identity in work settings highlights the central crux of this paper. A stigmatized social identity is one where an individual’s visible or hidden identity is devalued based on his/her membership in a group that is considered inferior to other groups in society. Stigma is defined as “a social construction that involves at least two fundamental components: (a) the recognition of difference based on some distinguishing characteristic, or 'mark'; and (b) a consequent devaluation of the person" (Dovidio, Major, and Crocker, 2000: 3). When confronted with stigmatizing events, people attempt to restore their sense of self and choose from a wide array of behavioral strategies ranging from ignoring stigma, avoiding stigma situations, using humor to diffuse the situation etcetera, which are collectively referred to as identity management strategies (Goffman, 1963). People choose identity management strategies, not as knee-jerk, passive responses, but as active responses that are in line with their goals. Indira prioritized her professional goals over self-verification or belonging goals; the resultant identity management strategy of “I have to live with it” was in service to her central goal of creating a place for herself in the C-suite where women and persons of color are underrepresented (Bureau of Labor Statistics, 2016). While scholars describe the different types of identity management strategies an individual may choose, they often do not describe why someone may choose to accept or challenge the applicability of stigma attributes to themselves in any given communication 1 episode. In the identity management literature, the general assumption is that at any time an individual faces stigma, a diverse array of strategies may be available to the individual. While this may be true in some social settings, work settings may offer an individual a restricted array of identity management strategies to enact, and among these options, the norms of the work place may make some strategies unrealistic options. For example, if a team-member makes a stigmatizing comment to a woman while working on a high-stakes deliverable at work, she may not see challenging the stigma communication as an option because doing so would create potential for conflict and the team’s performance on the deliverable could suffer. Further, individuals may have different goals that affect their choices. For example, three women in STEM fields who are faced with a sexist comment may choose different strategies for dealing with it depending on whether they are most concerned about being seen as competent, being valued for who they are as individuals, or fitting in. Specifically, the current work posits that constrained environments call for enhanced self-regulatory efforts as individuals make active choices on how they would like to manage their identities given the interaction of situational press with their own goals. Thus, the purpose here is to integrate identity management research with self-regulation perspectives to showcase that identity management strategies are in fact goal-oriented, choice behaviors and should be treated as such. Identity management, with some notable exceptions (e.g., Chaudoir & Fisher, 2010; Clair, Beatty, & Maclean, 2005; Jones & King, 2014; Meisenbach, 2010; Ragins, 2008; Shih, Young, & Bucher, 2013), has not been studied from a process perspective. When it has, scholars have largely tended to assume that reactions to stigma and the resultant identity management strategies unfold in the same way regardless of situational attributes. People choose identity management strategies depending on both the situational context in which stigma was 2 communicated and the relation of the stigmatized individual to the perpetrator (Goffman, 1963, 1974). For example, encountering stigmatizing communication from one’s supervisor where a person does not have much power to challenge the stigma is inherently different from encountering stigma from an acquaintance where norms of communication are relaxed and the individual has more degrees of freedom to choose from an array of identity management strategies (Ashforth & Humphrey, 1993). Given that organizations are becoming increasingly diverse, and the likelihood of stigmatization of minorities is a pressing concern (Harrison, Price, & Bell, 1998; Pelled, Eisenhadt, & Xin, 1999), we need to think about how the unique demands of work settings may impact how identity management strategies unfold at work. For example, in the current conceptualizations of identity management, key variables such as self-regulatory focus, goals and motives, organizational norms and hierarchies, and efficacy expectations are either not well-integrated or are underplayed. Thus, I seek to highlight these influences in the discussion of identity management, intend to shed light on the underlying assumptions, and explicitly challenge those assumptions to situate identity management in work settings. This integration offers three main contributions to the identity management and stigma literatures. First, by including a self-regulation view, one can explicate why decisions to accept or challenge stigma communication are made. The focus is on showing that identity management strategies are multiplicatively determined by an individual’s responsiveness to the situational demands, personal attributes, and his/her own goals. Thus, effective identity management requires an individual to use self-regulatory resources to navigate the complexities of stigma communication events that are often negatively charged, devaluing, and identity threatening. Second, such a self-regulated view allows us to be more specific about when and why individuals may choose some identity management strategies over others – for example, self- 3 regulatory foci of approach and avoidance can play a key role in guiding individuals’ choice of immediate action after a stigmatizing event occurs. Integrating a self-regulatory, goals perspective with identity management will help situate the study of stigma and identity management in organizational settings by explicitly requiring scholars to pay attention to the unique demands that individuals confront in work settings; particularly, by including a discussion of multiple goals and goal hierarchies, some of the assumptions embedded in stigma and identity management literature can be highlighted and challenged. Third, a clearer self- regulatory view in discussing identity management not only brings to fore several novel research questions, but can also help with more practical matters such as advising individuals with stigmatized identities to select identity management strategies that are effective and useful, and guiding leaders and human resource management professionals to better understand the different behavioral strategies people may choose in navigating stigmatizing situations and how these might be tied to critical organizational outcomes. These ideas are elaborated in the chapter that follows. 4 CHAPTER 1: PROPOSING A NEW IDENTITY MANAGEMENT CONCEPTUALIZATION Early identity management studies focused on empirically testing the strategies proposed by Goffman (1963), later work focused on adding to the list of identity management strategies (e.g., Blanz, Mummendey, Mielke, & Klink, 1998; Button, 2004), and recently there has been an attempt to organize these strategies into typologies (e.g., Meisenbach, 2010). Stigma communication, defined as the recognition of and reactions to people who embody stigma (Smith, 2007), has been consistently treated as an important antecedent to identity management (e.g., Meisenbach, 2010; Petriglieri, 2011; Richman & Leary, 2009). Where scholars have studied identity management in organizational settings, the focus is often on identity management as prototypical patterns of behavior in response to general environmental press (cf., Button, 2001; Lyons, Wessel, Tai, & Ryan, 2014; Richman & Leary, 2009; Shih, Young, & Bucher, 2013); I posit that the literature can also benefit from taking an event-based view of identity management whereby stigma communication in different situations is likely to guide differential choice of identity management strategies. For example, King, Mohr, Peddie, Jones, and Kendra (2017) in their within-person, experience sampling study found that lesbian, gay, and bisexual (LGB) workers used less concealing strategies when interacting with other LGB individuals in the organization and also when people offered cues that showcased their acceptance of LGB people. In contrast, when interacting with external clients LGB workers used more concealing strategies. Broadly, both stigma and identity management scholars have tended to frame stigma communication and identity management by taking a ‘stigmatized person view’ of those to whom stigma may have been communicated instead of viewing them as ‘people with stigma’. 5 The inherent assumptions underlying such frameworks are problematic because they impose identity centrality and self-verification motives on individuals and obfuscate our ability to understand why people may choose to accept or challenge stigma communication.1 Meisenbach (2010) states that an individual who accepts stigma communication “determines that the stigmatized aspect is part of their identity” and engages in strategies that are more defensive (e.g., passive acceptance, apologizing, using humor) (p. 278). In contrast, an individual who challenges stigma communication engages in “primarily proactive” strategies such as “social activism…and public education techniques” (p. 284). Identity centrality is defined as “the importance or psychological attachment that individuals place on their identities” (Settles, 2004, p. 487). People vary in the extent to which they choose to define themselves by facets of their identities. The assumption that people will necessarily do something about the stigma threat based on their attitudes about the applicability of stigma to self is unwarranted. People may choose to do nothing at all, not because they are accepting the stigma communication, but simply because the stigmatized identity is not their central identity and they are just not as distressed by the stigma communication as someone with a high identity centrality may be (Quinn & Chaudoir, 2009). Especially in work settings, it is highly likely that people low on identity centrality focus on other more salient goals in the stigma situation. 1 Notable exceptions that explicitly include identity centrality and/or self-verification in conceptualizing identity management exist, but these are limited to either discussing how individuals construct their professional identities (cf., Roberts, 2005), or how people manage their hidden identities (e.g., Lyons, Wessel, Ghumman, Ryan, & Kim, 2014; Quinn & Chaudoir, 2009; Ragins, 2008), or how racial and professional identity centrality relates to impression management efforts around one’s racial identity (Roberts, Cha, & Kim, 2014). 6 This problem is compounded by the fact that scholars have tended to ascribe self- verification goals as the central concerns of an individual in a stigmatizing event. An individual is said to be self-verifying when s/he strives to ensure a consistent and coherent sense of self by sharing, asserting, and communicating his/her identity to others (Swann, 1983; 1987). However in achievement settings other concerns may override the self-verification goals. Consider the following example: A woman working in a STEM research setting has to work on an all-male team and her supervisor makes sexist comments. Here, one cannot assume that her response to stigmatizing events will be to self-verify and assert her gender. She may choose to focus on self- enhancement goals, and protect her self-esteem by delivering high quality work outputs. The discussion so far posits two critical ideas: (a) the process of identity management may in fact look different in constrained work settings than it does in social settings, and (b) identity centrality and self-verification motives should not be implicitly assumed. The model in Figure 1 summarizes the crux of the argument that identity management should ideally be viewed from a lens of self-regulation, goal-oriented, choice behaviors. In the remainder of the paper, detailed rationale for building various components of the model is presented. 7 Figure 1: A conceptual model of identity management as self-regulated, goal-oriented, choice behaviors in organizational settings. (Note the dotted lines in the model show that ‘Should I do something’ and ‘Could I do something’ factors will operate under a context of stigma communication). Key Self-regulatory Identity Management Models With the exception of several models, a self-regulation or goal perspective that considers responses to specific stigma events or episodes is not prevalent in the literature. Before detailing the current model, we provide a brief explanation as to how the current work extends beyond other models with a self-regulatory or communication event frame. The Disclosure Process Model (Chaudoir & Fisher, 2010) focuses on self-regulation in context of concealed stigmatized identities (e.g., sexual orientation), contains pathways that are not relevant when stigma characteristics are not hidden, and is grounded in an interpersonal approach rather than a within-episode lens. In contrast, the current work can be applied to all forms of stigma and focuses on the intrapersonal decision of accepting or challenging a given 8 stigma communication. In addition to self-regulatory promotion and prevention approaches considered in the Disclosure Process Model, the current works includes consideration of navigating multiple goals, an important element of self-regulation not fully explored to date. The model proposed by Richman and Leary (2009) suggests that stigma events precipitate a negative affective response, and are immediately followed by different construals of the stigma event (e.g., high perceived cost of rejection), which are then related to an individual’s motivation of wanting to engage in prosocial, withdrawal/acceptance, and antisocial responses. While the model accounts for motivations in stigma responses, it does so without accounting for the role of constrained environments, multiple goals and/or goal hierarchies, and does not explicitly include a discussion of identity management strategies. Similarly, while the literature on coping with stigma can be thought of as self-regulatory in nature, it too misses this focus and instead frames stigma responses within the stressor and appraisal of stressors view (cf., Miller & Major, 2000). While Swim and Thomas (2006) have taken a self-regulated, goal-oriented view of peoples’ coping behaviors in response to discrimination that has some parallels to our conceptualization, their discussion does not consider the impact of constrained settings such as work domains and also does not explicitly include identity management behaviors as a research focus. To summarize, I extend beyond the identity management literature that does not incorporate a self-regulatory perspective (e.g., Clair, Beatty, & Maclean, 2005; Jones & King, 2014; Meisenbach, 2010; Ragins, 2008; Shih, Young, & Bucher, 2013) by explicitly adopting that perspective. I further extend beyond the work that has self-regulatory elements (e.g., Chaudoir & Fisher, 2010; Swim & Thomas, 2006) in four ways: (a) focusing on a within episode or event level of analysis (i.e., an intrapersonal lens), (b) explicitly incorporating the 9 notion of environmental constraint on individual response options, (c) adding consideration of the multiple goals likely operating concurrently in considering responses to stigma communication (i.e., self-verification, self-enhancement, belonging), and (d) acknowledging the importance of identity centrality, which is treated inconsistently in models of identity management. The current work takes as a starting point the identity management taxonomy of the Stigma Management Communication Model (Meisenbach, 2010), which focuses on stigma communication events, and is more comprehensive in detailing the specific ways people choose to either accept or challenge stigma communication episodes. The current conceptualization builds on this by explicitly incorporating self-regulation and goals perspectives, and incorporating contextual and individual differences variables in the discussion of identity management behaviors. Self-regulation and Goals: A Brief Overview Karoly (1993) defined self-regulation as: Those processes, internal and/or transactional, that enable an individual to guide his/her goal-directed activities over time and across changing circumstances (contexts). Regulation implies modulation of thought, affect, behavior, or attention via deliberate or automated use of specific mechanisms and supportive metaskills [sic]. The processes of self-regulation are initiated when routinized activity is impeded or when goal- directedness is otherwise made salient (e.g., the challenge, the failure of habitual action patterns, etc.) Self-regulation may be said to encompass up to five interrelated and iterative component phases: (1) goal selection, (2) goal cognition, (3) directional maintenance, (4) directional change or reprioritization and (5) goal termination (p. 25). Control Theory (Carver & Scheier, 1981) is widely used in explicating how self- regulation works and asserts that an individual aspires to reduce the sensed deviations between his/her current-state and his/her desired goal-state. For example, an individual seeking to produce fifty widgets in three days will assess how far he/she is along toward that goal and modify behavior to achieve the goal. In the case of identity management, a stigma communication 10 episode can enter into the calculus of considering the current state versus an end-goal (e.g., how one wishes to be viewed by others). However, self-regulation is a finite resource and is susceptible to depletion. Exerting cognitive and/or emotional effort at managing negative emotions as in the case of encountering stigma (Inzlicht, McKay, & Aronson, 2006) can lead to depleted states where individuals make maladaptive decisions ranging from poor health decisions, behavioral and affective outbursts, poor interpersonal functioning, drug abuse, and even criminality (Hagger, Wood, Stiff, & Chatzisarantis, 2010). Goals are central to the self-regulation process and are defined as mental schemas or “symbolic structures” that guide human behavior towards valued outcomes (Karoly, 1993). Goals help direct one’s efforts in the desired direction, can be broken down into structures that describe the goal content, and are organized into goal hierarchies. That is, at any given moment in time individuals have multiple goals that range from macro goals to more granular goals (cf., Ford 1992; Schmidt, Dolis, & Tolli, 2009). Goal hierarchies specify different goals at increasing levels of complexity (i.e., goal content) and abstractness and range from the lowest level goals called the action plan goals, achievement goals, principle goals, to the highest level as described by self-goals (cf., DeShon & Gillespie, 2005). For the most part, behavioral scripts are enacted at the action goals level, and higher-order goals are achieved by working on lower-order goals (Carver & Scheier, 1981; DeShon & Gillespie, 2005). People function under a web of highly differentiated and inter-related multiple goals. To navigate this complex web, individuals have to (a) make self-regulated choices about selecting context-appropriate goals, (b) rely on goal representations to elicit critical information about the self as it relates to the environment, and (c) select goals depending upon the requirements of the situation, and the requirements of the self. That is, while people take into account their needs and 11 motives, they also factor in various situational influences such as norms that provide information on which goals are best pursued in the current situation, and whether or not goal-pursuit will lead to desired results. Expectancy beliefs such as self-efficacy (e.g., Ford 1992) further help people to select between different goals and requisite course of action. Thus, self-regulation, together with goals, motives, needs, and efficacy expectations form the multiplicative complex of psychological processes by which people persist on their tasks. People often have to choose from different goals in the goal hierarchy and draw upon their self-regulatory and motivational resources in goal pursuit (Suarez Riveiro, Cabanach, & Arias, 2001). A staggering number of goals can be identified at different levels of the goal hierarchies, and it is virtually impossible for any one conceptualization to discuss them all at the same time (DeShon & Gillespie, 2005; Ford, 1992). Thus, to bound the current discussion, this dissertation focuses on three key higher-order self goals: self-enhancement, self-verification, and belonging as these are not only relevant for the work context, but are also important under conditions of stigma threat (cf., Swim and Thomas (2006) for a discussion of two additional self- goals). Self-enhancement goals are those where individuals strive to protect their self-esteem, seek to promote positive self-perceptions, and are motivated to build their self-image favorably in the eyes of others (Baumeister, 1982). One way in which employees pursue self-enhancement goals is by effectively pursuing and accomplishing performance goals. At least in individualistic cultures, self-enhancement is critical for how performance is perceived and evaluated by ones’ supervisors and peers, is a deeply embedded socio-organizational schema that aligns with the norm of productivity in organizations, and is therefore tied to perceived competence of employees and deemed critical for generating tangible and intangible value for one’s team(s) and organization. Self-verification goals are those where an individual is concerned about asserting 12 his/her identity (Swann & Read, 1981). Organizations with diverse workforces are ripe environments for cueing self-verification goals – when people high on identity centrality are made aware of their different identities, it may create a push for them to protect their identities. Finally, an individual is said to be pursuing belonging goals when s/he is concerned about getting along with people around them (Baumeister & Leary, 1995). Belonging goals are critical to consider because these are at the heart of human interactions and relationships; organizations depend on employees’ ability to get along. The remainder of the discussion will be bound by limiting the discussion to these three higher-order goals. In organizational settings, people who can better self-regulate themselves, set appropriate goals, and strive in goal pursuit showcase positive organizational outcomes such as task performance (Locke & Latham, 1990). Organizations require people who are able to monitor the environment for feedback, who can self-evaluate where they stand with respect to the discrepancy introduced by the feedback and their original goals, who can self-reflect on the actions needed to address the discrepancy, and who can self-regulate to initiate the said actions (Kanfer, 1990). Stigma communication can be seen as a threatening feedback from one’s environment (Carver & Scheier, 1981; Schmidt, Dolis, & Tolli, 2009) and has the potential for being a strong contagion for other goals that an individual may have held at a more proximal level. The productivity norm in most organizations is likely to create a press on people such that when faced with stigma communication, in addition to choosing appropriate identity management strategies, they also have to get their work done. Thus, while in social settings an individual may choose from a vast array of identity management strategies, in an organizational setting, people face norms and situational press that may limit their expression significantly. Therefore, under stigma threat, self-regulation takes on 13 an even more important role in organizational settings as compared to social settings. The next section further expounds upon why we should take a self-regulatory view of stigma and identity management. Why Consider Self-Regulation in Responses to Stigma Communication? Research shows that stigma threat depletes self-regulation reserves (e.g., Bair & Steele, 2010; Inzlicht, McKay, & Aronson, 2006). As noted, in organizations, people pursue multiple goals that may sometimes even conflict with each other (Karoly, 1993; Schmidt, Dolis, & Tolli, 2009). People find pursuing multiple goals taxing enough, but if stigma threat also occurs, it is likely to compound the psychological burden due to managing multiple, conflicting goals, and also safeguarding ones’ identity all at the same time. An individual has to make decisions on what to do with a stigma communication: Should one accept or challenge? What kind of identity management strategies should be used? However, these decisions are made in relation to the proximal and valued goals at hand. The question then is: (a) do people persist with proximal goals and ignore the stigma threat, or (b) does addressing stigma threat become the proximal goal and, if so, under what conditions is this most likely to be the case, or (c) do the norms of the environment, peoples’ expectancy beliefs, and self-regulation ability allow people to engage in both goals concurrently (i.e., reduce the stigma threat and continue goal pursuit at hand)? These questions have rarely been investigated in the literature and are presented here as core to the proposed framework. The current conceptualization posits that instead of viewing people as passive recipients of stigma, they should be viewed as active agents who monitor the environment closely for the kind of responses that would be considered appropriate, who self-evaluate on what is more important in that situation (e.g., belonging versus self-enhancement goals), who self-reflect on 14 how they can best reduce the discrepancies by choosing different identity management strategies, and most importantly, who self-regulate their responses. Considering self-regulation and goals as foundational elements in stigma responses can help us explicitly ground identity management strategies as goal-bound and motivated, norm-bound, and expectancy driven responses. I build on each of these characteristics in the following sections. Further, people at work have to manage multiple goals at different hierarchical levels (DeShon & Gillispie, 2005; Lord et al., 2010) that may even conflict with each other (Ford, 1992). Scholars have rarely looked at how peoples’ core, activated goals influence their choice of identity management strategies: How do stigmatizing events impact people’s ability to self-regulate? How does identity centrality affect responses? Do identity management strategies depend on the proximal and valued goals? Do people choose different identity management strategies under self-regulatory depletion? These questions set the stage for discussing identity management as goal-bound strategies; after noting the role of identity centrality in deciding how to respond to a stigma communication, I focus on multiple goals and the role of self-regulatory foci in identity management. Identity Centrality and Accepting and Challenging Stigma Communication Identity is a complex, multi-layered phenomenon and people may be differentially attached to different parts of their identity. That is, just because someone has a visible minority identity that can be stigmatized does not imply that the individual considers that particular feature of their identity as their most central characteristic. For example, a Caucasian woman may not hold her gender as her central identity-defining feature and may instead use her majority status as the central identity. Alternatively, an Indian woman may not consider her racial identity as her central identity and may instead choose her occupational identity as the central defining feature for the self. People who have high identity centrality are more attached to that identity, 15 and therefore, threats to that identity will likely be perceived as a more salient and distressing stimuli (Thoits, 1991). Hence, a stigma communication event may require greater self-regulation resources from an individual high in identity centrality than for an individual for whom the particular identity and any stigma communication is seen as less goal relevant. High identity centrality individuals are also more likely to have their sense of self-esteem tied up in their group affiliation (e.g., Crocker, 2002) and are more likely to engage in challenging stigma communication to restore their self-esteem (Sedikides, Gaertner, & Toguchi, 2003). In contrast, someone with low identity centrality is more likely to acquiesce when faced with stigma communication. Accepting the stigma communication may help them navigate the situation better by safeguarding their self-regulatory reserves 2 for other more salient goals and needs in the situation (e.g., focusing on a work product). Thus, under stigma communication: Proposition 1a: Identity centrality is associated with self-regulatory reserves such that the greater an individual’s identity centrality, the greater the self-regulatory reserves needed to manage stigma communication Proposition 1b: Identity centrality relates to how people respond to stigma such that those high in identity centrality will tend to challenge stigma communication and those low will tend to accept stigma communication The Role of Self-regulatory Foci in Accepting or Challenging Stigma Carver and Scheier’s (1981, 1990) Control Theory posits that an individual self-regulates behavior once s/he detects a discrepancy between the current state and their desired-end-state. Self-regulatory systems with positive values are “discrepancy reducing” because an individual is 2 Self-regulatory depletion (the outcome variable in the model) refers to lowering of self- regulatory reserves. 16 motivated to minimize the discrepancy between the current and the desired-end state; self- regulatory systems with negative values are “discrepancy amplifying” because an individual here is motivated to maximize the discrepancy between current and the undesired-end state (Higgins, 1998, p. 2). Consider a performance context where a Malay woman wants to manage a competitive, high-visibility marketing account in her organization. When she talks about her goals with her manager (a Caucasian male), she is told that “you are aiming too high – even men have had a hard time earning that account”. She processes the stigmatizing comment, but promptly reverts to her central goal of wanting to make a mark in the organization by showcasing her competence. She chooses to view the race for the prized account as an opportunity to shine a light on her abilities and doubles her efforts on creating a solid presentation for the board of directors. In this case, by doubling her efforts she can be seen as reducing the perceived discrepancy and moving towards her goals. In contrast, consider an Indian woman who too would like to make a bid for the same high-visibility marketing account. When she speaks to her manager (Caucasian male), he states that “The account is staffed with an all male team…are you sure you want to duke it out with those competitive guys?” She processes the stigmatizing comment, and chooses to drop out of the race for the account. She rationalizes that the opportunity was too risky to attempt when the team would most likely not accept her as their leader anyway. In this example, the woman can be seen as amplifying the discrepancy and moving away from her goal. Higgins (1998) distinguished these systems asserting that, “promotion focus is concerned with accomplishments, hopes and aspirations. It regulates the presence and absence of positive outcomes” whereas “prevention focus is concerned with safety, responsibilities, and obligations. It regulates the absence and presence of negative outcomes” (cf., Higgins, 1998, p. 16). In the 17 stigma communication context, an individual with a promotion self-regulatory focus is likely to pursue the ideals of educating others against the deleterious effects of stigma, hoping for people to be kind and respectful to each other, and aspiring for a more just-world. Such a person is more likely to challenge than accept stigma communications. If a person has a prevention self- regulatory focus, s/he is likely to avoid getting into conflict with others by pointing out the stigma communication and avoid incurring interpersonal costs to self. Such a person is likely to accept rather than challenge stigma communication. Thus, under stigma communication: Proposition 2: Individuals with promotion self-regulatory focus are more likely to challenge stigma communication while individuals with prevention self-regulatory focus are more likely to accept stigma communication Identity Management Strategies: Norm-bound and Expectancy Driven Rarely have identity management researchers questioned if peoples’ role, power status, and perceptions of psychological safety impact the decision to accept or challenge stigma and the choice of identity management strategies. This oversight is troubling because it is important to consider the relationship of the perpetrator and the person to whom stigma is communicated (cf., Meisenbach, 2010). For example, in supervisor-follower relationships in highly hierarchical organizations, there may be a norm of not challenging ones’ superiors. Research shows that people with stigmatized identities tend to consider the role of norms in deciding whether to reveal or hide their identities (Clair, Beatty, & MacLean, 2005), especially when there are strong norms for conformity (DeJordy, 2008). As another example, in the absence of a positive diversity climate, people might lack the psychological safety to challenge stigma communication (Swim & Thomas, 2006) or to reveal the invisible aspects of their identity (e.g., revealing one’s sexual orientation; Hitchcock & Wilson, 1991). Similarly, King, Reilly, and Hebl (2008) content 18 analyzed the coming-out experiences of LGB individuals and found that supportive climates predicted disclosures. Finally, Phillips, Rothbard, and Dumas (2009) delineate that people consider the status distance between themselves and co-workers/supervisors before deciding whether or not to disclose personal information at work. While a reductionist view of identity management has its benefits, it also obscures our ability to understand the phenomenon holistically, so it is important to consider several contextual variables that can direct individuals’ decisions surrounding stigma communication. Thus, under stigma communication: Proposition 3a: Individuals who perceive a non-supportive diversity climate will be less likely to challenge and more likely to accept stigma communication while those who perceive a supportive diversity climate will be more likely to challenge and less likely to accept stigma communication Proposition 3b: Individuals experiencing low psychological safety will be less likely to challenge and more likely to accept stigma communication while those experiencing high psychological safety will be more likely to challenge and less likely to accept stigma communication Proposition 3c: Individuals who receive stigma communication from those lower in the role hierarchy are more likely to challenge stigma communication and less likely to accept stigma communication, while those who receive stigma communication from those higher in the role hierarchy are more likely to accept stigma communication and less likely to challenge stigma communication Identity management strategies should also be seen as expectancy-bound responses (Eccles & Wigfield, 2002; Carver & Scheier, 1981). An individual’s self-efficacy guides if she 19 will goal-set, choose activity, expend effort, and persist (Bandura, 1977). For example, if an individual does not have the self-efficacy to establish a working relationship with a colleague after challenging her on stigmatizing behaviors, then that individual is likely to not challenge the colleague. Specifically, under stigma communication: Proposition 3d: Individuals with low self-efficacy are less likely to challenge and more likely to accept stigma communication than those with high self-efficacy who are more likely to challenge and less likely to accept stigma communication Accepting or Challenging Stigma Communication: Differential Depletion of Self- Regulatory Resources Given that self-regulation is a finite resource (Hagger, Wood, Stiff, & Chatzisarantis, 2010), is it possible that it is differentially depleted in accepting or challenging stigma communication? Studies show that accepting a stigma communication can have a self-protective effect to an individual’s self-esteem (Crocker & Major, 1989) because people with stigma often discount the feedback from majority groups, and downplay the stigmatized attributes of their identity while overplaying the positive attributes (Crocker & Major, 1989). Thus, while it would seem that groups that are stigmatized by society should have lower self-esteem, empirical work does not support this view. Similar findings are seen in dirty work literature where people who work in stigmatized professions (e.g., morticians, garbage collectors) use their knowledge of stigmatization as a self-protective mechanism (Ashforth & Kreiner, 1999). They build stronger communities, and engage in collective identity management strategies such as condemning the condemners, transcending above their jobs etc. Accepting stigma communication may even be an empowering approach (Shih, 2004) such as when the LGBTQ group co-opted stigmatizing terms such as “queer”. 20 Similarly, we do not know if individuals expend more or less self-regulatory effort in challenging stigma communication, Klar and Kasser (2009) using both college and national samples found that individuals who reported actual or intended activist behavior scored higher on measures of well- being, life satisfaction and positive affect, personal growth, and social integration. Thus, challenging the status quo might facilitate building of self-regulatory reserves. However, it is apparent that stigmatizing events can create large goal-performance discrepancies. Under these conditions, the individual has to decide whether or not the action towards discrepancy reduction or an avoidance strategy would be better. Strategies such as ‘why even try – nothing I do will change the perceptions about my identity anyway’, or alternatively tabling the pursuit of the undesired-end-state and thereby just accepting stigma communication in the moment (Austin & Vancouver, 1996) can safeguard regulatory resources for that stigmatizing event. The few studies on self-regulatory foci and threatening conditions provide mixed findings3 (cf., Ståhl, Laar, & Ellemers, 2012; Trawalter and Richeson, 2006). Furthermore, it is unclear if prevention or promotion focused self-regulation could safeguard or deplete regulatory resources over time, or if the benefits or costs of these approaches are more short-termed and should be tested empirically. 3 Trawalter and Richeson (2006) find that during interracial interactions, when participants were told to have a positive interracial exchange (operationalized as inducing a state of promotion focus) versus when they were told to avoid having a negative interracial exchange (operationalized as inducing a state of prevention focus), individuals in the prevention focus condition performed worse on an executive control task (assessed via Stroop test). In contrast, Ståhl, Laar, and Ellemers (2012) find that stereotype threat facilitated initial cognitive performance, and replicated this effect in two different studies. The authors find that stereotype threat not only facilitated cognitive performance, but also did so when prevention focus was experimentally induced. In contrast, this facilitation effect was not seen when promotion focus was induced (note that cognitive performance utilizes the executive function). Over time, stereotype threat inhibited performance under prevention focus but not under promotion focus. 21 Research Question 1: Does prevention or promotion focused self-regulation in stigma communication events safeguard or deplete regulatory resources? Navigating Multiple Goals: Goal Choice Under Stigma Threat Consider the following example: If a woman is working on a presentation, but comes under stigma threat at work, does she abandon her self-enhancement goals, or does she continue with the proximal task at hand, but compromise on her self-verification and belonging goals? Stigma communication can impact all three goals deleteriously and simultaneously. So how does she navigate multiple goals under stigma threat? This is a particularly complex problem that the self- regulation literature is still grappling with. According to the Control Theory, people monitor their environment for discrepancies on key goals and in the face of any discrepancies they adjust their goals (Carver & Scheier, 1981). The adjustment process typically unfolds as follows: if the goal performance discrepancy is too large, and seems untenable to fix, then people generally revert to a lower-order goal in the goal hierarchy. So if someone’s self-verification goals are threatened, and if his/her identity centrality is high, it is likely to create a big discrepancy (e.g., Quinn & Chaudoir, 2009). In this case, instead of focusing on the higher order self-verification goal, an individual might prefer to focus on a lower order goal such as finishing up a particular work task. An important consideration that is not clearly stated in much identity management literature is that in organizations the norms of productivity may often restrict an individual from completely abandoning their self-enhancement and/or belonging goals in favor of self- verification goals. That is, at work, one needs to accomplish work tasks and get along with others in order to maintain employment. The issue of self-regulation as a finite resource is also important in this discussion. If stigma threat depletes self-regulation (e.g., Bair & Steele, 2010), then people should prefer to 22 revise their goals such that goal revision is not particularly taxing. Ideally, one could accomplish this by simplifying the goal and moving down the goal hierarchy. However, in a multiple goal context where goals tap into different motivational structures, another approach is also likely: people may abandon the goal revision with higher goal performance discrepancy in favor of another goal in a different motivational structural space that may have a smaller goal performance discrepancy. That is, it is likely that an individual may choose a self-enhancement goal (where the goal-performance discrepancy is low) instead of a self-verification goal (where the goal-performance discrepancy is high). This approach may have a protective impact on self- regulation resources. In a stigma communication event: Proposition 4: Whether a person decides to accept or challenge stigma communication will be related to goal performance discrepancy, such that an individual will choose to focus on the goal where the goal performance discrepancy is the lowest to safeguard self- regulatory reserves. Based on previous discussion, it follows that an individual’s identity centrality is likely to also affect the goal choices an individual makes. Specifically, under stigma communication: Proposition 5: Individual differences in identity centrality likely impact the goal choices of self-enhancement, self-verification, and/or belonging goals when faced with a discrepancy, such that people high on identity centrality are more likely to choose self- verification goals and less likely to choose self-enhancement goals or belonging goals. It is also likely, that an individual might take on the goal revision to pursue both self- verification and self-enhancement goals at the same time – this can be seen in identity management strategies such as ‘transcending above the stigma’ (i.e., when an individual “reduces a stigma’s offensiveness by calling attention to how the stigma attribute can be a means 23 that leads to a valuable end…Strippers have been found to focus attention on how their work provides food for their families”) (Meisenbach, 2010: 283). While there is preliminary work to offer some guidance on how people may navigate multiple goals, there is no established theory to explicate how these decisions are made (Lord et al., 2010; Schmidt, Dolis, & Tolli, 2009). Figure 2 adapts Schmidt, Dolis, and Tolli’s (2009) heuristic to illustrate how an individual might navigate multiple goals when faced with stigma communication. To keep the discussion from becoming overly complex, I use self-verification and self-enhancement goals in this example. I consider stigma communication in the context of environmental volatility. High environmental volatility here is defined as stigma threat to one’s most proximal and valued goal, low psychological safety, and norms that inhibit wide array of identity management choices (e.g., preclude challenging stigma communication). In contrast, low environmental volatility is defined as stigma threat to not the most proximal and/or valued goals, high psychological safety, and norms that facilitate a wide array of identity management choices. Self-regulatory promotion and prevention foci along with an individuals’ self-efficacy, and identity centrality are treated as key variables in navigating multiple goals and making appropriate choices in response to stigma threat. In line with Schmidt, Dolis, and Tolli (2009), the heuristic presents likely outcomes for two conditions: (a) where an individual is concerned with attempting to meet both goals – that is, where they have dual-goal expectancy to self- enhance and self-verify, and (b) where an individual is concerned with making a hard decision – that is choosing between the two goals. 24 Figure 2: Heuristic flowchart of how people may navigate and choose from multiple goals. Based on Schmidt, Dolis, and Tollis’ (2009) heuristic (Figure 2 is adapted based on Schmidt et al.’s heuristic; SV = Self-verification goals, SE = Self-enhancement goals), the following preliminary research questions come to mind: Are people likely to exhibit dual-goal expectancy with regard to higher-order self-goals? Or would people prefer to operate at lower order, action-plan levels of these higher order self-goals? How would people manage the incompatibility between these different goal structures concurrently? For example, self- verification requires people to assert their identity, while belonging goals push for an individual to compromise and go along with people: Research Question 2: Do people under high environmental volatility pursue dual-goal expectancy or do they restrict themselves to either-or choices? 25 Goals and Differential Identity Management Strategies Lord, Dieffendorf, Schmidt, and Hall (2010) suggest that in organizations people are often under the press of multiple goals at any given time; the goals may not always be aligned to each other. No attempt has been made to investigate how people may choose between multiple goals that are different from each other in terms of the underlying motivational structure. For example, how might one choose between self-verification, belonging, and/or self-enhancement goals? While determining the exact process will require much further conceptualization and research, as a first step, I posit that identity management choices could differ depending on which goals are more proximal for an individual. I describe three different goal contexts and show how different identity management strategies are likely to align differentially. Three Illustrative Examples. First, consider a performance context where a young woman works in the STEM field as a research analyst. She takes pride in her work and is primarily concerned about her self-enhancement goals at work. That is, her key goal is to build her self-esteem by showcasing effective performance in the achievement context. Assume that a colleague makes the following statement: “Say what you will, but women are just not as good at math as men are”. In this situation, what identity management strategies is she likely to choose? DeShon and Gillespie (2005) assert that in multiple goal settings, people focus on the proximal, more highly weighted goal even though other goals may also have been activated. Given her proximal and more valued goal is self-enhancement, she is going to be mindful about whether or not the identity management strategies serve to protect and/or advance the self-enhancement goal first. Thus, even though her self-verification goals may have been activated, she is more likely to choose a strategy that advances (or at least does not deter) her self-enhancement goals. 26 Second, consider a performance context where a young, black woman works in the STEM field as a research analyst. She is uniquely aware of her identity as a double minority, understands that people treat her as a token representative, and her self-verification needs are high. That is, her key motive is to assert her self-identity by taking pride in her racial and professional identity in the achievement context. In response to the sexist statement by a colleague, she is likely to choose identity management strategies that serve her proximal and more valued goal of self-verification first. Table 1 shows how differential identity management strategies are chosen in response to self-verification goals as compared to self-enhancement goals. Third, consider a performance context where a young, Indian woman works in an American STEM firm as a research analyst. She is uniquely aware of her identity as an immigrant, understands that people treat her as an outsider, but given her collectivistic upbringing, her belonging needs are ever pressing. That is, her cultural influences have taught her to build a strong network of relationships, and her key motive is to feel a sense of kinship and belonging with her colleagues. When faced with a stigmatizing comment, she is likely to choose identity management strategies that serve her proximal goal of belonging first. Again, note how under the press of belonging goals, she is likely to accept stigma communication in a passive way and is highly unlikely to challenge the stigmatization because it would go against her key goal to belong (Table 1). I rationally map the illustrative examples presented so far to the likely choice of strategies on Meisenbach’s (2010) framework (see Table 1). While other identity management taxonomies exist, Meisenbach’s taxonomy is most comprehensive in that it applies to a broad range of potentially stigmatizing conditions (i.e., social categories, behaviors, attributes) and focuses 27 specifically on responses within an episode or event as opposed to broad strategies (e.g., avoiding situations or people in general). The identity management strategies in the table are included as exemplars; research may indicate other strategies to be applicable as well. Table 1: Exemplars of Identity Management Strategies When Self-enhancement, Self- verification, or Belonging Goal is More Valued and Proximal An Individual Who Accepts the An Individual Who Challenges the Stigma Stigma Communication Communication • When Self- verification Goals More Valued bond with others who are stigmatized • recategorize (i.e., shifting focus from a stigmatized part of identity to a non- stigmatized part) When Self- enhancement Goals More Valued • recategorize (i.e., shifting focus from a stigmatized part of identity to a non- stigmatized part) • • make favorable social comparisons (i.e., compare self to other stigmatized groups, but finding a way to feel superior to them) logical denial about how the stigma attributes do not apply to self • educate (i.e., provide stigma countering information and evidence) • make favorable social comparisons (i.e., compare self to other stigmatized groups, but finding a way to feel superior to them) • simple denial strategy (i.e., stating that there is no stigma) • passive acceptance (i.e., say Unlikely to occur When Belonging Goals More Valued nothing) • self-deprecating humor • bond with others who are stigmatized Based on the rational mapping of goals and identity management strategies, a preliminary proposition is offered; specifically, under stigma communication: Proposition 6: An individual’s goals will be related to their response to stigma communication such that different identity management strategies will be chosen based on goals an individual considers proximal and more important. Acceptance identity management strategies will require less effort because these are about maintaining status-quo whereas challenge identity management strategies will require more effort because these inherently involve a “change” component. Further, in line with the 28 ‘law of least mental effort’, acceptance strategies can be reflective of an avoidance orientation which is less effortful, while challenge strategies may take more cognitive work (i.e., coming up with logical denials to educate the perpetrators while concurrently managing the burden of stigma communication) and these individuals will thus have to expend more effort at identity management (Kool, McGuire, Rosen, & Botvinich, 2010). Finally, the type of identity management strategies used can have important implications for outcomes such as job satisfaction, intentions to turnover, job strain and associated psychosomatic complaints, and self- regulatory depletion. In organizations where roles are organized in strict hierarchies, people may prefer to maintain the status quo and engage in acceptance strategies (Lach, Ingram, & Rayner, 2004). Employees who engage in the acceptance identity management strategies are less likely to face job strain (Bakker & Demerouti, 2007), and may in turn be less predisposed to psychosomatic complaints (Nelson, Cooper, & Jackson, 1995). However, on the flip side, acceptance strategies may not offer employees an avenue to exert their influence in managing their identities. According to the job-demands resources model, lack of autonomy might reduce such employees’ job satisfaction and also increase their intentions to quit (Bakker & Demerouti, 2007). Therefore, Proposition 7: Accepting identity management strategies will be related to (a) low job satisfaction, (b) high turnover intentions, (c) low strain, (d) low psychosomatic complaints, and (e) low self-regulatory depletion. In contrast, challenging stigmas can showcase an “activism” of sorts. Although in “process control” driven settings activist orientation or “taking charge” behaviors may be more strain inducing (Chiaburu & Baker, 2005: 625; Bakker & Demerouti, 2007), these same behaviors could also lead to higher job satisfaction and lowered intentions to quit (Burke & 29 Deszca, 1988). Klar and Kasser (2009) found that individuals who reported actual or intended activist behavior scored higher on measures of well-being, life satisfaction and positive affect, personal growth, and social integration. Similarly, police officers who ascribed to “social activist” career orientation reported lowest intent to turnover and highest job satisfaction when compared to police officers who ascribed to career orientations such as ‘self-investors’, ‘careerists’, and ‘artisans’ (cf., Burke & Deszca, 1988). Thus challenging the status quo strategies might offer employees an avenue to exert their influence in changing perceptions that they might disagree with or feel at odds with. This perceived autonomy might increase employees’ job satisfaction and also decrease their intentions to quit (Bakker & Demerouti, 2007). Thus, I propose the following: Proposition 8: Challenging identity management strategies will be related to (a) high job satisfaction, (b) low turnover intentions, (c) high strains, (d) high psychosomatic complaints, and (e) high self-regulatory depletion. 30 CHAPTER 2: DEFINING FOCUS OF THE PROPOSED RESEARCH This chapter is organized around three key subsections. In the first section, the overall theoretical model presented in chapter one is bounded based on key research questions and propositions that were investigated, and the rationale for including and/or excluding different elements of the overall model is also delineated. A portion of the overall theoretical model (i.e., the model that was empirically tested) is also presented graphically. In the second section, an overview of a multi study approach is presented and each study is briefly introduced. In the third section, stigma communication, self-regulatory depletion, goals, and identity management strategies are operationally defined. These operational definitions relate to the sampling strategy used and the methods used to conduct the empirical tests of the theory outlined in chapter one. Bounding the Overall Theoretical Model The scope of the broad conceptual model outlined in chapter one is narrowed (see Figure 3). While the theoretical model builds a case for taking into account contextual variables such as norms, power distance between stigma communicator(s) and receiver(s) due to role hierarchies, etcetera in understanding how people manage their identities under stigma communication, the model outlined here includes only a subset of these contextual variables. That is, while the model includes perceptions of diversity climate, and influence of role hierarchy differences between stigma communicator(s) and receiver(s), it does not include organizational norms or how hierarchical an organization itself is. The decision to include only a subset of variables was guided primarily by practical constraints such as onerous sampling and data collection requirements. The organizational contextual variables were added in the theoretical model to better couch the study of stigma and identity management in the workplace and to explain why looking at the workplace specific identity management strategies in context of goals is important. 31 While the role of contextual variables has been empirically tested in few studies – examples include, diversity climate (e.g., King, Reilly, & Hebl, 2008; McKay, Avery, Morris, Hernandez, and Hebl, 2007); psychological safety (e.g., Hitchcock & Wilson, 1992); role hierarchy (e.g., Pringle, 2008) – the current research effort could not undertake an examination of all the contextual variables discussed because that would have entailed data collection in multi- organizational samples. Thus, not all of the variables discussed in the theoretical model in chapter one were given a center stage in the empirical test of the ideas presented in the current work. The model presented below also does not include how people may choose differential identity management strategies based on the goal performance discrepancies created due to multiple goal selection discussed in chapter one, although the selection of identity management strategies based on different proximal goals is investigated. I did not test questions centering around how people may select from multiple goals under stigma communication because of the paucity of research on multiple goal selection. Thus the current work has been intentionally restricted to using the multiple goals perspective to build theoretical argument that goals and self-regulation perspective are needed in thinking about identity management in work situations. I posit that we need to first test the more basic, foundational links such as the role of goals and self-regulation at the between- and within-person level and how these are implicated in identity management before crafting the more complex arguments about multiple goal selection under stigma communication. Further, testing ideas on multiple goals under stigma communication in field studies where context level data can be gathered, while appealing, is not practical. Instead, I propose that these ideas be tested first in the laboratory under more controlled conditions. The overall conceptual model is shared below (see Figure 3). 32 Figure 3: A conceptual model. (Note: The dotted lines in the model show that the ‘Should I do something’ and ‘Could I do something’ factors will operate under a context of stigma communication) As such, the goal here was to empirically test the following foundational ideas: (a) Does stigma communication lead to self-regulatory depletion (i.e., Proposition 1a); (b) Is there differential self-regulatory depletion when people accept or challenge stigma communication (i.e., Propositions 7a and 7b); (c) Under different goal conditions do people select different identity management strategies (i.e., adapted from Proposition 6); and (d) Under stigma communication, do different identity management strategies impact key outcomes (e.g., psychosomatic health indices, strain, and self-regulatory depletion) (i.e., Propositions 7 and 8). Thus, the upcoming empirical research studies are the first step towards investigating the more foundational aspects of the overall research agenda that were outlined in Chapter One. Please refer to the model (see Figure 3) that encapsulates all the research questions and hypotheses that were empirically tested as described in subsequent chapters. 33 An Overview of the Multi Study Approach This project was a multi study approach toward understanding identity management behaviors at work. Specifically, the focus was to understand the outcomes, antecedents, and mediators of identity management behaviors when stigma is communicated at work. Three studies were designed to explore how people self-regulate and enact identity management strategies under stigma communication. The first two studies were cross-sectional online studies that used Qualtrics panels4 while the third was an experience sampling methodology study (also used Qualtrics panels). In addition, a Pilot Study was also conducted that tested the fundamental idea of whether or not stigma communication impacted self-regulatory depletion, in addition to testing the overall research protocols to be used in studies one, two, and three. Studies one, two, and three were all designed to add more nuance to the question of self-regulatory depletion under stigma communication by including different mediators such as identity management strategies and main effects such as proximal goals, identity centrality, and self-regulatory foci. In so doing these studies acted as building blocks towards testing a more comprehensive research agenda. As mentioned earlier, first a Pilot Study that focuses on evaluating experimental prompts, measures, and manipulations was conducted. In addition, I also assessed the efficacy of various experimental tasks. Tasks to be piloted included the vivid-recall task (used to operationalize stigma communication) and its association with self-regulatory depletion; and the performance discussion manipulation designed for Study Two. 4 To align the theoretical propositions with the research methods, participants who had work experience were recruited so they could speak to the effects of stigma and identity management experiences in a more ecologically valid manner. As such, Qualtrics panels’ services were retained for recruiting participants for all studies. It was determined that this was the best course of action to allow for better generalizability of results to the work contexts than using student samples. 34 Study One was a cross-sectional online Qualtrics investigation that evaluated if self- regulatory depletion varied as a function of accepting versus challenging stigma communication. Thus, this study tested claims that people may showcase differential self-regulatory depletion given their individual standing on identity centrality, self-regulatory foci, and self-efficacy. Study Two was a cross-sectional online Qualtrics investigation that built on Study One and investigated whether or not people show differential self-regulatory depletion as a function of their different proximal goals such as self-enhancement, self-verification, and belonging goals in addition to investigating the role of self-efficacy and identity centrality. Finally, Study Three built on the findings from studies one and two in a field-based experience sampling methodology study. The goal here was to evaluate if different identity management strategies were differentially related to key outcomes such as self-regulatory depletion, psychosomatic health indices, and strain. By sampling people based on stigma communication events as they occurred in their day-to-day life, this study avoided the disadvantages inherent in cross-sectional experiments where stigma communication is artificially induced or is assessed based on retrospective recalls. Also, this study was grounded in a within- person view of identity management behaviors and associated outcomes. Together, these three studies were intended to help illuminate the complex construct space of identity management behaviors as they relate to stigma communication. An Overview of Justification of Operationalization Choices in the Proposed Research In the discussion so far, key constructs were defined with an eye to developing the theory. Before presenting individual studies to test the theoretical relationships proposed earlier, it is important to present the key constructs from a methodological standpoint. As such, in this section four focal constructs of stigma communication, self-regulatory depletion, goals, and 35 identity management are described that are used in the upcoming studies (presented in subsequent chapters). Stigma communication, self-regulatory depletion, and goals have been studied using multiple rubrics; the upcoming sections ground the phenomena of interest by delineating the chosen perspective on each, and/or presenting a rationale for the various choices made in operationalizing. Stigma Communication. Stigma communication, so far, has been broadly defined as the recognition of, and reactions to, people who embody stigma (Smith, 2007). Because people can be stigmatized and can feel stigmatized for a multitude of reasons and identities, in the subsequent empirical work I examine stigma communication under the rubric of gender-identity. In line with NIH’s guidelines to include women and minorities in research samples (1994), and for practical reasons5 specific to research design, I specifically focus on women’s experiences to study how they self-regulate and enact identity management behaviors at work when encountering sexism (operationalization of stigma communication). Specifically, research streams on feminism, sexism, and anthropology show that not only are women stigmatized and derogated as the ‘weaker sex’ across most societies in the world (cf., Acker, 1973; Brown, 1970; Chatterjee, 1989; Kerber, 1988; Smith-Rosenberg, 1972), but that their contributions to the social, political, economic, and organizational spheres of life are also 5 The decision to rely on women as the focal minority group who may experience press of stigmatization was also driven by practical constraints. First, to test the stigma communication I needed to have stigma cues that could resonate with all the participants in the study. Thus, instead of defining stigma more broadly to include race, gender, sexual identity, religious identity etcetera, I decided to focus on one salient identity-defining attribute. Second, while other minority identities were equally viable options, choosing women as the focal minority group served the practical goal of minimizing5 onerous sample recruitment burdens. Finally, because women are the largest minority group, generalizability5 of any findings from the proposed study, while limited to women, will illuminate how a large part of the population construes and copes with stigma communication. 36 devalued when compared to those of men (cf., Ely & Padavic, 2007; Heilman & Haynes, 2005; Pateman, 1989; Reskin, 1988). The negative macro effects of cultures that impose traditional and highly sex-typed roles for women (cf., Block, 1973) also permeate organizational spheres and empirical research shows that women face a whole host of deleterious effects of stigmatization that range from sexual harassment at work, stereotype threat related performance deficits, biases in performance ratings, glass-ceiling effects etcetera (cf., Eagly & Karau, 2002; Fitzgerald, Drasgow, Hulin, Gelfand, & Magley, 1997; Hoobler, Wayne, & Lemmon, 2009; Wheeler & Petty, 2001). According to the tenets of feminist criticism these effects are to be expected because women are making a claim to their own rightful positions in the organizational hierarchies, thereby challenging the power structures that have traditionally been occupied by men (cf., Acker, 1973; Ng, 1993; Skrla, Reyes, & Scheurich, 2000; Walker, 2001). Self-regulatory depletion. In the subsequent research studies, I have adopted Baumeister, Bratslavsky, Muraven and Tice’s (1998) strength model of self-control. Baumeister et al. (1998) state that self-regulation draws from a finite resource, and liken the global resource to a muscle that if overused, could become fatigued. Authors posit that the more one self-regulates, the more self-regulatory depletion can be expected; that is, after exerting self-control on prior task(s), people are unable to regulate themselves and are prone to impulsivity6 (Baumeister et al, 1998). 6 There is some discussion in the literature on whether or not self-regulation and self-control are synonymous and could therefore be used interchangeably (cf., VanDellen, Hoyle, & Miller, 2013). While self-regulation is a more automatic process, often occurring outside our awareness, self-control is when people consciously exert an effort to override behavioral/affective/cognitive response(s). In their study, VanDellen et al, (2013) found that not all acts of self-regulatory behavior can be assumed to entail self-control, and therefore, we cannot expect self-regulatory depletion/ego-depletion in response to all tasks/activities that require self-regulatory reserves. In the proposed research studies, this thinking has already been built-in in the research conceptualization presented so far. For example, the current research conceptualization maintains that even though stigma communication is a negative affective stimulus for most people, those with low identity centrality may not have to engage in the same level of self-control in response 37 Several studies have found support for the ego-depletion hypothesis (cf., Debey, Verschuere, & Crombez, 2012; Ferrari & Pychyl, 2007; Thau, Aquino, & Poortyliet, 2007; Tice, Bratslavsky, & Baumeister, 2001), and preliminary evidence from a physiological perspective of self-regulatory depletion is also supportive (cf., Gailliot & Baumeister, 2007). There are, however, studies that have failed to support the theory (cf., Muraven & Slessareva, 2003); alternative conceptualizations7 have also been offered. However, recent meta-analytic findings corroborate that ego depletion does lead to self-regulatory failure (Hagger, Wood, Stiff, & Chatzisarantis, 2010). In line with how ego-depletion is typically studied, a dual-task paradigm was piloted to assess self-regulatory depletion (Hagger et al., 2010). Since the strength model posits that engaging in self-regulation depletes an individual’s resources from the global, finite pool of resources, it was expected that after having engaged in tasks that require self-regulatory effort, participants will showcase depletion when they are required to engage in a second task that also to stigma communication as compared to those with high identity centrality because they do not place the same level of valence on their identities as a source of self-esteem. Thus, these individuals may not see the same levels of self-regulatory depletion. 7 In addition to studying self-regulatory depletion from a cognitive perspective (cf., Ajzen, 1985), other socio-cognitive conceptualizations that seek to explain self-regulatory depletion exist – for example, the justification based account of self-regulatory depletion that states that self-regulatory failure does not always have to be about inability to impulse control and can instead also stem from an individual’s justifications for the discrepant behavior (De Witt Huberts, Evers, & De Ridder, 2014). Kool, Tops, Strubin, Bouw, Schneider and Jostmann (2014) use “ego fixation” to elucidate the pathways by which self-regulatory depletion occurs via an inability to shift from self-regulating. Authors posit that an ego-fixated individual is likely to be so occupied with exerting self-control that s/he continually, and involuntarily, persists with self- regulation to a point of engaging in “highly responsible, duty-driven behavior” in one sphere of control at the cost of other goals/activities (Kool et al., 2014; p. 7). Given the dearth of evidence to support the alternative socio-cognitive conceptualizations, and the meta-analytic evidence in favor of Baumeister et al’s (1998) strength model, the current work was grounded under the rubric of the strength model of self-control to examine stigma communication as it unfolds in organizational settings. 38 draws upon self-regulatory resources. Thus, according to the dual-task paradigm, participants in the experimental condition will be assigned two self-control tasks presented one after the other. In contrast, participants in the control group will first be presented with a task that does not require self-control followed by a task that requires self-control. When comparing the experimental and the control groups, one can expect for participants in the experimental condition to show more depletion than participants in the control condition on the second self- regulatory task. The phenomenon of self-regulatory depletion has been assessed using many different tasks. While the wide applicability of the self-control model inspires confidence in its potential utility to drive empirical work, the divergent tasks and methods used also make it difficult to select the best tasks that are likely to fit the current research needs. Thus, I relied on meta- analytic findings to inform the choice of tasks and methods to test the ideas presented in the current theorizing and decided to pilot the vivid recall task and the anagram task. In addition, a self-report measure of self-regulatory depletion (Ciarocco, Twenge, Muraven, & Tice, 2007) was also piloted as a possible dependent variable measure. A detailed discussion of the rationale for why these specific tasks were chosen is included (see Table 28 in Appendix A). Goals. In the current research agenda, I focus on three key goals of self-enhancement, self-verification, and belonging. A description of why these three were chosen was presented in Chapter One. Given the event-based view of stigma communication in the current research, it is important to consider whether or not the theoretical conceptualization of these higher-order goals is aligned with the operationalization – that is, when a stigma event occurs, is it likely for a person to think about self-enhancement, self-verification, and/or belonging as their proximal goals during the stigma event? I use DeShon and Gillespie’s (2005) motivating action theory to 39 answer this question. DeShon and Gillespie (2005) state that while the highest-level self-goals (e.g., agency, esteem, and affiliation) are often desired end states for most, these are fairly abstract and do not clarify a path forward to most people about how to achieve them. It is at the level of principle goals (e.g., growth, fairness, structure, social value etcetera) that people see a general way by which they can achieve the high level self-goals. At the next level are achievement goals (i.e., mastery-approach, perform-approach, and perform-avoid) that are general patterns of behavior that people follow in approaching the higher order goals. Finally, the action plan goals (e.g., seeking feedback, managing impressions, allocating resources, exploring problem etcetera) are the ones that people enact to achieve the principle goals, and thus to achieve their self-goals. The central point of this conceptualization is that even though goals are organized in a hierarchy, this hierarchy is “massively interconnected” such that there are multiple pathways between and across various goals (DeShon and Gillespie, 2005, p. 1107). Thus, even though the goals of self-enhancement, self-verification, and belonging are higher-level goals, they are highly interconnected with, influence, and are influenced by the lower order goals that may be more proximal during a stigma event. For example, in achievement contexts, when employees pursue performance goals and showcase their competence at work, they are pursuing their self- enhancement goals. Thus, in the current research operationalization, participants are instructed to think about self-enhancement goals8 in terms of highlighting their contributions to team members 8 Note this operationalization is an example of the broader conceptualization of self- enhancement goals where individuals strive to protect their self-esteem, seek to promote positive self-perceptions, and are motivated to build their self-image favorably in the eyes of others (Baumeister, 1982). One way in which employees pursue self-enhancement goals is by effectively pursuing and accomplishing performance goals and showcasing their competence at work (see Chapter One for a detailed explanation). 40 (in Study Two), and are trained to think of self-enhancement goals as putting their best foot forward in discussions and meetings with team-members, supervisors, and customers, shining a light on their good performance etcetera (in Study Three). Similarly, when people pursue their higher-order belonging goals in work contexts, it is likely that during the stigma event, they enact these goals by thinking about their principle goals of creating social value, or enact them at the action goals level by avoiding conflict. Thus, participants in the current research are instructed to think about belonging goals in terms of making every effort to get along with their team members (in Study Two), and are trained to think about belonging goals as fostering cordial relationships with others, or wanting to get along with others (in Study Three). Finally, when people pursue their higher-order self-verification goals, they may enact them by asserting and/or protecting their identities. Thus, in the current research, participants are instructed to think about their self-verification goals in terms of making every effort to assert their identity and be true to who they are in their communications. To conclude, even though the goals of self-enhancement, self-verification, and belonging are higher order goals, the manner in which these are operationalized is in fact aligned with the proximal goals. That is, the higher order goals are achieved via a conduit of the more proximal goals. Identity management strategies. Theoretical discussions of identity management strategies point us to a vast array of behavioral options that individuals may exercise when faced with identity threat9. Given the focus of the current research on women, it is imperative to define 9 Note identity threat is different from identity management. Identity management is a set of behaviors that people engage in when responding to perceived and/or actual identity threats whereby a person expects being marginalized, devalued, or discriminated against based on some characteristic (Steele, Spencer, and Aronson, 2002). In that sense, identity threat is an antecedent to identity management behaviors. 41 which of the identity management strategies women are more likely to use in performance contexts. While multiple measures exist to explore how people with hidden stigmatized identities manage their identities (e.g., LGBTQ individuals’ identity management strategies (Button, 1996; 2001; 2004), pregnant women (Jones, King, Gilrane, McCausland, Cortina, & Grimm, 2016), religious identities (Lyons, Wessel, Ryan, & Kim, 2012), and age (Lyons, Wessel, Tai, & Ryan, 2014), with the exception of Roberts, Settles, and Jellison’s (2008) work (discussed in detail later), I was unable to find measures to assess the identity management strategies of women10 (Raman, 2008). Thus, I adapted relevant items from measures that are most aligned with the current research agenda. First, I have adapted items from Moody-Ayers, Stewart, Covinsky, and Inouye’s (2005) adaptation of the perceived racism scale against African Americans in multiple domains including both employment and public domains (cf., McNeilly, Anderson, Armstead, Clark, Corbett, Robinson, Pieper, & Lepsito, 1996). The measure has been used in several studies (cf., Mendoza-Denton, Downey, Purdie, Davis, & Pietrzak, 2002), and shows good reliability ranging from α = .94 (Richman, Bennett, Pek, Siegler, and Williams, 2007) and α = .98 (Rucker, West, & Roemer, 2010). 10 The measures used for hidden identities do not lend themselves to being adapted for women; for example, strategies such as counterfeiting, or avoiding sharing information about one’s sexual orientation etcetera are not viable options for women. The literature on coping with stigma has some promising scales such as the ‘Ways of Coping scale’ (WOCS; Folkman & Lazarus, 1988b) but measures like these are based on coping strategies that are not the same as identity management strategies – for example, dimensions of the WOC include accepting responsibility, confrontive coping, escape-avoidance, planful problem solving, positive reappraisal, self-controlling, and seeking social support (cf., Rexrode, Peterson, & O’Toole, 2008). Similarly, there are measures of impression management behaviors (cf., Bolino, Turnley, 1999) that on the surface may seem related to identity management, but tap into dimensions that cannot be used to assess identity management strategies (e.g., self-promotion, ingratiation, exemplification, intimidation, and supplication). 42 Second, both recategorization and positive distinctiveness strategies as described in Roberts, Settles, and Jellison (2008)11 were in line with the identity management strategies presented as most likely to be used by women in work contexts, and as such one relevant item from each of these scales was used in the identity management measure for the current research agenda. Please note that I have ensured that the items align with the identity management strategies12 presented in chapter one along the two dimensions, accepting and challenging identity management strategies (see Tables 3 and 4 in Appendix B for mapping). The strategy to operationalize identity management strategies in the manner outlined so far is not without its problems. First, given that items from multiple measures were combined together to create a new measure, it no longer constitutes ‘minor adaptation’ of existing scales. Thus, the psychometric properties of the new measure were also evaluated in the Pilot Study (details presented in the next chapter). Second, creating a measure that is aimed at assessing the identity management strategies of women by using a measure of behavioral responses to racism (i.e., Moody-Ayers et al, 2005) has its problems. A reader may question this approach for concerns around deficient construct validity of the measure. To account for the possibility of 11 I had mentioned that Roberts, Settles, and Jellison’s (2008) study on how women of color strategically manage their identities was a promising lead for a relevant identity management measure. Unfortunately the authors limited their investigation to capturing just two identity management strategies: recategorization and positive distinctiveness. In addition, they created their measures by adapting Morgan’s (2002) items from the “social identity-based impression management scale” to capture identity management behaviors, but the reported reliabilities of their scales ranged from fair to moderate (Recategorization identity management scale α = .63; Positive distinctiveness identity management scale α = .78). 12 Previously, I had rationally mapped the illustrative examples of women’s identity management in performance contexts with respect to their goals and delineated the following strategies as most likely to be used (see Table 1): When accepting stigma communication, women were more likely to use bonding with others who are stigmatized, recategorizing, passive acceptance, and/or self-deprecating identity management strategies. In contrast, when challenging stigma communication, I had made the case that women will be more likely to make favorable social comparisons, use simple denials, use logical denials, and/or educate stigma communicators. 43 creating a measure that inadvertently excludes some identity management strategies that may particularly be useful for women, I also conducted a scan of the literature focusing specifically on women’s identity management strategies (see Table 29 and Table 30 in Appendix B). Some new strategies that were not included in the rational mapping presented in chapter one emerged as viable strategies women use to manage their identities: (a) bolstering/refocusing; (b) isolating self; and (c) avoiding stigma situations. Based on this review, items for these strategies from the Moody-Ayers et al. (2005) measure were retained in the identity management strategies measure for women. Conclusion So far this chapter delineated the key research questions that narrow the overall theoretical model. Based on the key research questions, four different research studies were introduced (i.e., including the Pilot Study). Finally, I delineated the key constructs from a more methodological standpoint. Next, in line with the multi-study approach, a Pilot Study and three multi-method studies have been detailed in the upcoming chapters. First the research question is delineated for each study. Since the detailed rationale for the key ideas and propositions has already been presented in chapter one, the discussion has been limited to summarizing why the specific study is needed, presenting the individual conceptual models, and delineating the research methods and results. A brief discussion is included, and limitations of each study are also discussed. 44 CHAPTER 3: PILOT STUDY There were six key reasons why a Pilot Study was deemed necessary before launching the three empirical studies. First, the vivid recall task (Inzchlit & Kang, 2010) and the performance tasks developed for studies one and two were piloted to gauge if these worked effectively in the proposed sample. The Pilot Study was also necessary to ensure whether or not the instructions provided to participants were understood and had the intended effect of stigma communication manipulation. Second, the Pilot Study was used to assess the relative efficacy of the two dependent variable measures: (a) the anagram task (Hagger et al., 2010) and (b) the state self-control capacity scale (SSCCS; self-regulatory depletion measure; Ciarocco, Twenge, Muraven, & Tice, 2007). The goal was to explore if both the anagram task and the SSCCS measure captured enough variability in responses, if the participants showed motivated engagement with the task and/or the measure. The goal was to assess which of the two measures of dependent variables should be used in studies one and two. Third, key measures to be used across the three studies were also piloted to establish their psychometric properties. This was specifically needed for measures that had been created for the purposes of the proposed research (e.g., the identity management strategies measure described in chapter two). Fourth, for Study Three (Chapter Six), it was necessary to explore the frequency with which women faced sexism at work. Thus, the Pilot Study also asked participants the number of times they faced sexism at work over a period of past four weeks. In addition, psychometric properties of a shortened measure of self-regulatory depletion were also tested for later use. 45 Fifth, the open-ended responses shared by the participants in their narratives (described later) were used as a sample to train coders to code for (a) sexist behaviors (i.e., gender harassment, unwanted sexual attention, and sexual coercion); (b) emotions; (c) various identity management strategies used (accepting and challenging strategies) based on Meisenbach’s (2010) definitions; and (d) perpetrators. . Finally, the Pilot Study also tested the foundational claim that runs through all the studies in this research agenda that stigma communication depletes self-regulatory reserves. Specifically, Hypothesis 1 proposed that stigma communication would be related to self-regulatory depletion such that under condition of stigma, individuals will show higher self-regulatory depletion. Participants In total, 159 women13 were recruited via Qualtrics’ panels. They were compensated based on the incentive structure that participants had previously negotiated with Qualtrics. The incentives typically come in various forms that participants agreed to such as e-rewards, discounts, sky miles, etcetera. To manage the participant burden (in terms of time and the number of measures administered), the sample was split into two cohorts of 72 and 87 participants each. The first 72 women were administered the vivid recall task and selected measures required to pilot this task (for Study One). Another 87 participants were administered the performance discussion task developed for Study Two and the associated measures. Two participants were removed from the analyses (described later) for a final sample size of 70 women. 13 The decision to limit the sample to women has already been described in Chapter Two. 46 Screening criteria. Participants who were United States’ citizens, were currently employed, and who worked at least 30 hours per week were recruited; all others were screened out. Additionally, for the sexist condition in piloting the vivid recall task (details presented later in Pilot A), respondents were screened out if they did not have a specific sexist incident to report. Pilot A: Piloting the Vivid Recall Task Design. A between-subjects, basic randomized design was used for this study. Participants were randomly assigned to either experimental condition, or to control condition to assess whether or not the self-regulatory depletion is due to the stigmatizing manipulation or if it is simply due to fatigue effects or other alternative explanations. Participants. 70 women participated in this Pilot Study. They ranged between ages 20 – 70 years (M= 37.10, SD = 11.60). Among the participants, 6.49% were African American, 77.92% were Caucasian, 11.69% were Hispanic, 1.3% were Pacific Asian, 1.3% Native American or American Indian, and 1.3% were from other racial/ethnic groups. Note, participants were allowed to check multiple racial/ethnic categories. Participants reported a total work experience ranging from less than a year to 47 years (M= 15.84, SD = 11.28), and reported being at their current job ranging from less than a year to 37 years (M= 15.84, SD =6.56). Diverse jobs were represented across the participants ranging from social work, sales associates in retail, office administrators, actresses, teachers, supervisors, housekeepers, line workers, servers etcetera. Manipulation. The vivid recall task was piloted based on Inzchlit and Kang’s (2010) manipulation14. Participants in the experimental condition were asked to vividly recall a time 14 Unlike Inzchlit and Kang (2010) who allowed participants to recall a threat to any of five identities (ethnicity, gender, religion, sexual orientation, and age), participants were asked to vividly recall an incident at work when they faced sexism. This was intended for three key 47 when they felt discriminated against because of their gender at work (i.e., when they faced sexism). To ensure that the participants reflected on the sexist event (i.e., the stigma communication event), they were asked to explain the event in detail by “thinking about all the thoughts, feelings, and sensations they had at the time of the experience” (Inzchlit and Kang, 2010; p. 475). To structure their story-telling, instructions were included that asked participants to also describe the following: (a) details on who communicated the stigma to them and their relationship with that person, (c) how they felt when confronted with sexism by describing physical, emotional, and psychological reactions, and (d) how they coped with the stigmatizing event. While Inzchlit and Kang (2010) gave participants just five minutes to write their narratives, participants in the current Pilot Study were given fifteen minutes to reflect about this experience and write a detailed narrative. In addition to embedding the performance conditions that elicited specific elements of stigma communication that I was interested in capturing, I also asked the participants to ensure that their responses were at least 600-800 words. Then they worked on the anagram task and were also administered the state self-control capacity scale. Participants in the control group were asked to write 600-800 words recalling a memory they would like to share and then they worked on the anagram task and were also administered the self-control capacity scale. The underlying rationale is that if stigma communication creates self-regulatory depletion, then participants in the experimental condition should show lower purposes. First, having participants write about one cue minimizes the likelihood that participants might reflect about the part of their identity that they consider more salient than gender (e.g., race) thereby creating conditions for a strong manipulation. Second, and in line with first, focusing on just one identity builds in enough time for participants to provide rich descriptions for one incident instead of having to write multiple small paragraphs about different forms. This was a limitation of the Inzchlit and Kang (2010) study where they were unable to content- analyze the responses for lack of richness. Third, this choice aligns with the current framing of the theory building effort within the sexism perspective. 48 willingness to persist on the dependent variable task (i.e., anagram task) and should also show higher depletion (i.e., lower scores) on the state self-control capacity scale than the control group participants. The overall experimental process is outlined (see Appendix C). Anagram Task. While Inzchlit and Kang (2010) used Thaler and Johnson’s (1990) lottery choice task that assesses risky decision-making, I used the anagram task given the meta- analytic support presented in Chapter Two. Self-regulatory effort was assessed by mean time spent on unsolvable anagrams – i.e., individuals’ demonstrated persistence in working on the unsolvable anagram task (cf., Baumeister, Gailliot, DeWall, & Oaten, 2006; Converse & DeShon, 2009; Muraven, Tice, & Baumeister, 1998). If the experimental manipulation is successful, and the vivid recall task depletes people’s self-regulatory reserves, then we would expect that the experimental group should show lower persistence on the anagram task than the control group. The anagram task (Smith, 2002) and manipulation check items were adapted from Converse and DeShon (2009). Measures Self-regulatory depletion. For the Pilot Study, two different measures were used to assess self-regulatory depletion. First, a behavioral measure of depletion was used. Participants were asked to work on unsolvable anagrams and the time spent working on these was recorded (explained above). Second, the 25-item state self-control capacity scale (SSCCS; Ciarocco, Twenge, Muraven, & Tice, 2007) was used to assess self-regulatory depletion dependent variable in this study. This is an unpublished measure, and rationale for why it was used in the current study is provided (see Table 31 and Table 32 in Appendix D). Sample items included: “I feel mentally exhausted”, “Right now it would take a lot of effort for me to focus on something”, “I cannot absorb any more information”, and “I feel sharp and focused”. The items were rated on 49 a Likert-type scale of not true (1) to very true (7). In line with the literature, in the current sample, the Cronbach’s alpha was .95. A short-version of the SSCCS scale was also piloted (Ciarocco et al., 2007) for Study Three. Ciarocco et al. (2007) recommended using items 3, 6, 9, 10, 11, 12, 13, 15, 16, and 18 (see Appendix E for items) and reported a reliability of .84 for the shortened scale. In the current sample, the Cronbach’s alpha was .87. Based on the reliability information, this shortened scale was retained for use in the daily surveys administered in the Study Three to minimize participant burden. Identity management strategies. Both accepting and challenging identity management strategies were assessed in this study. Based on the mapping presented earlier in chapter two, accepting (seven items) and challenging (seven items) identity management strategies were assessed (see items in Appendix E). Participants were asked to select the behavior or behaviors that describe how they dealt with sexism in the stigma communication event they had recalled on a scale from never (1) to very often (5). The Cronbach’s alpha for the accepting identity management strategies scale was .76. The Cronbach’s alpha for the challenging identity management strategies scale was .85. Coping self-efficacy. The 13-item Coping Self-efficacy Scale (CSE; Chesney, Neilands, Chambers, Taylor, and Folkman, 2006) was used to assess participants’ self-beliefs in coping with stigma communication (i.e., sexism). The measure has been designed to assess how people respond to life stressors by drawing upon their social network, engaging in problem focused coping, or engaging in emotion focused coping. Chesney et al., (2006) demonstrated predictive validity for well-being, and report sound psychometric properties (Cronbach’s α = .95). The measure has been used in exploring if coping self-efficacy buffers the impact of discrimination 50 on minorities such as transgender women of color (cf., Jefferson, Neilands, and Sevelius, 2013; α = .94). In the current study, instructions for the items stated, “When faced with sexist events such as the one you described how confident or certain are you that you can:…” followed by a presentation of the 13 items such as ‘break an upsetting problem into smaller parts’, ‘sort out what can be changed, and what cannot be changed’, and ‘make a plan of action and follow it when confronted with a problem’ (see Appendix E for full measure). Participants rated the items on a Likert-type scale from ‘cannot do at all’ (0), to ‘moderately certain can do’ (5), to ‘certain can do’ (10). In the current sample, Cronbach’s alpha for this scale was .94. Preliminary Analyses Results for Pilot A Prior to performing hypotheses testing analyses, all data were examined for accuracy, missing values, normality, and assumptions of multivariate analysis using SPSS 24.0. Analyses indicated that skew and kurtosis were within an acceptable range for most variables (i.e., absolute value of skew < 2, absolute value of kurtosis < 7; West, Finch, & Curran, 1995); therefore, variable transformations were not necessary. However, time spent on anagram task showcased significant skew and kurtosis due to outliers (detailed analysis is presented later). Following Hayes and Preacher’s (2014) recommendation, one dummy code was created for stigma (i.e., experimental condition where participants wrote the vivid recall of a sexism-related event) and no stigma conditions (i.e., control condition where participants wrote vivid recall of any memory they wanted to share). The stigma condition was used as the referent category. Vivid Recall Task. Overall, participants in the experimental condition provided 38 vivid recalls of a time when they had faced sexist events at work. A few examples of the vivid recalls 51 are highlighted (see Table 33 in Appendix F). On average, participants wrote an average of 96.76 words (i.e., roughly one paragraph each; SD = 61.73). Of the 34 participants in the control group condition (i.e., women who wrote about a memory they wanted to share), two participants reported incidents about sexism and/or sexual harassment. These two participants were excluded from further analysis thereby creating a control group of 32 women, and a final sample size of 70 participants. A few examples of the vivid recalls for the control condition are described (see Table 34 in Appendix F). On average, participants wrote an average of 106 words (i.e., roughly one paragraph each; SD = 68.94). While participants were instructed to write about 600-800 words for the vivid recall task, they tended to write much shorter recalls of the sexist incidents. Many participants referenced how having to recall these events was a painful experience, and it is likely that once they felt they had answered the key parts of the overall question, the women wanted to exit the recall of a painful/traumatic memory. It is also likely that the shorter recalls showcase a form of satisficing. Overall, the recommended length of the recalls was stipulated merely to ensure that participants felt the press to think about the questions asked in some detail. The success of the vivid recalls as a stigma manipulation did not depend on the number of words used, and self-regulatory depletion was cued even with the shorter recalls. Instead, what was more important for the current research design was that the participants answered who the perpetrator was, described the nature of the incident, described how they felt when confronted with sexist behavior, and described how they coped with the sexist situation. As the open-ended data shows participants responded to all these sub-parts, and the data yield interesting insights even with the shorter recalls. 52 Anagram Task. The mean time spent on the anagrams for the experimental condition was 239.54 seconds (N = 38, SD = 265.25) while the mean time for the control condition was 262.64 seconds (N = 34, SD = 360.70). Participants responded to three self-report items about the anagram task: ‘I exerted a lot of effort during the word puzzle task’, ‘I felt frustrated during the word puzzle task’, and ‘I had to exert self-control during the word puzzle task’. This three- item measure had a Cronbach’s alpha of .64. Participants in the experimental condition (M = 3.63, SD = .97) had higher mean and standard deviation when working on the anagram task than those in the control condition (M = 3.43, SD = .92); the mean differences were not significant. For the experimental group, analysis of time spent on anagrams indicated that skew and kurtosis were within an acceptable range (i.e., absolute value of skew < 2, absolute value of kurtosis < 7; West, Finch, & Curran, 1995). Specifically, data yielded a skew value of 1.87 and a kurtosis value of 2.62. To evaluate outliers, a visual scan was conducted using histograms, stem and leaf plots, normal Q-Q plots, and a boxplot all of which show significant departure from normality. The boxplot identified six observations as outliers (See Figure 4 and 5). 53 Figure 4: Boxplot for time spent on anagram tasks with outliers for experimental group. Figure 5: Q-Q plot for time spent on anagram tasks with outliers for experimental group. For the control group, analyses of time spent on anagrams indicated that skew and kurtosis were not within an acceptable range (i.e., absolute value of skew < 2, absolute value of kurtosis < 7; West, Finch, & Curran, 1995). Specifically, data yielded a skew value of 4.14 and a 54 kurtosis value of 20.17. To evaluate outliers, a visual scan was conducted using histograms, stem and leaf plots, normal Q-Q plots, and a boxplot all of which also showed significant departure from normality. The boxplot identified six observations as outliers (See Figure 6 and Figure 7). Figure 6: Boxplot for time spent on anagram tasks with outliers for control group. Figure 7: Q-Q plot for time spent on anagram tasks with outliers for control group. Given the non-normality in both the experimental and the control group condition data, and the outlier analysis, participants who were 2.5 standard deviations away from the mean were 55 treated as interesting outliers and removed from further analyses (Aguinis, Gottfredson, and Joo, 2013). This yielded an experimental group (N = 32) where participants spent an average of 133.66 seconds on anagrams (SD = 79.95). Skew value of .723 and kurtosis of -.14 were obtained within acceptable range values, and a visual check for normality of data was established. Using a similar strategy for the control group (N = 32) it was found that participants spent an average of 190.36 seconds on anagrams (SD = 137.46). Skew value of 1.02 and kurtosis of .08 were obtained within acceptable range values, and a visual check for normality of data was established. Vivid recall task and anagram persistence. In accordance with recommendations that state that after outliers are removed from a dataset, it is best to report both sets of results – that is, results with the outliers in the dataset and results without the outliers in the dataset, I conducted an independent samples t-test to test the relationship between vivid recall and persistence on the anagram task. When outliers are included in the dataset, results show that there are no statistical differences between the amount of time spent on anagrams in the experimental condition (M = 239.54, SD = 265.25) and control condition (M = 262.64, SD = 360.70); t (70) = -.31, p = .76. Specifically, results show that there is no difference between persistence on anagrams depending on whether or not participants were stigmatized. In other words, when outliers were not removed from the dataset, stigma communication due to vivid recall of sexist events seems not to negatively impact participants’ self-regulatory depletion. However, when the outliers are removed from the dataset, it was found that there are statistical differences in the expected direction between experimental (M = 133.66, SD = 79.95) and control groups (M = 190.36, SD = 137.46); t (49.82) = -2.02, p = .049 (note: Levene’s test for equality of variance was significant and the values reported here correspond to equal 56 variances not assumed). When outliers were removed from the dataset, participants who were exposed to stigma communication due to vivid recall of sexist events seemed to showcase statistically higher self-regulatory depletion, as evidenced by lower persistence on anagrams (i.e., less time spent on anagrams) than those in the control condition. Vivid recall task and self-regulatory depletion. The SSCCS measure was used as a second way to test participants’ self-regulatory depletion after stigma communication using the vivid recall task. Please note, the measure is scored such that lower scores are indicative of higher depletion. In Hypothesis one, I had posited that stigma communication would be related to self-regulatory depletion such that participants who recalled sexist events would show higher self-regulatory depletion (i.e., lower scores on the SSCCS) as compared to participants in the control condition (i.e., who shared a vivid recall of a memory they would like to share). To test this hypothesis, an independent samples t-test was conducted. Results show that there are statistical differences in the expected direction between experimental (M = 4.39, SD = 1.28) and control groups (M = 5.10, SD = 1.37); t (68) = - 2.24, p = .03. Participants who were exposed to stigma communication due to vivid recall of sexist events seemed to showcase statistically higher self-regulatory depletion, as evidenced by lower scores on the SSCCS than those in the control condition. Assessing the task order effects. Given that one key reason to do the Pilot Study was to evaluate the relative efficacy of the dependent variable tasks of unsolvable anagram and the self- regulatory depletion measure (i.e., the SSCCS scale), the Pilot Study was designed to also account for likely task order effects (due to the order in which these tasks were presented). The tasks were counterbalanced such that, in both the experimental and control conditions, 57 participants were randomly assigned either the anagram task first or the self-regulatory depletion scale first. When the self-regulatory depletion measure was presented first, it had no effect on participants’ depletion on the anagram tasks. However, when the unsolvable anagrams task was presented first, participants faced higher self-regulatory depletion as was evidenced by their lower SSCCS scores when they were administered anagrams first and the SSCCS measure later (see Table 2). Table 2: Means, standard deviations, reliabilities, and correlations for study variables (N= 70). Variables Condition (1) (Experimental group = 1; Control group = 2) Task order (2) (SSCCS presented first = 1; Anagram presented first = 2) SSCCS score (3) SD - -.38** 2 - 3 M - 4.71 1.36 .26* 1 - - - -.06 - (Note. *p <.05, **p <.01). The order effects were also analyzed using an independent sample t-test by selecting only the control condition data (N = 32) so that we could partial out the effects of the experimental manipulation. Results show that there are statistical differences when self-regulation depletion measure was administered first (MSSCCS = 5.72, SDSSCCS = .92) versus when the anagram task was administered first (MSSCCS = 4.39, SDSSCCS = 1.48); t (30) = 3.08, p = .00. Specifically, results show that even in the control condition where participants were sharing memories they wanted to share, there is a statistically significant difference in the positive direction between self-regulatory depletion depending on whether or not participants were administered the SSCCS first or if they were administered the anagrams first. 58 Identity Management Strategies Scale. Even though the reliabilities of the identity management scale were satisfactory, an exploratory factor analysis was conducted on the 14 items using Principle Axis extraction to evaluate the scale further. First, I relied on the K1 rule of looking for eigenvalues >1 and obtained a four-factor solution. However, the eigenvalues of the last two factors were barely above one, so in line with best practices in factor analysis and to avoid over-factoring (Russell, 2002), I discarded the K1 rule for the second EFA test. This time, an EFA using Principle Axis extraction was conducted using a varimax rotation to see the underlying factor structure. Instead of the K1 rule, I entered the fixed number of factors to extract since the measure was created based on theoretically derived items that mapped on to two factors of accepting and challenging identity management strategies. For both the tests described here, I also used a scree plot as the visual criterion to determine the number of factors, and the two scree plots looked very similar with two factors identified above the elbow (see Figure 8 for scree plot using Varimax rotation). Figure 8: Scree plot using Varimax rotation for Identity Management Strategies Scale. 59 The analysis yielded two factors and together they account for 54.24% of the total variance: Factor 1 accounted for 32.42% of the variance (eigenvalue = 4.538), Factor 2 accounted for 21.82% of the variance (eigenvalue = 3.055). The factor matrix shows the strength of association between the factors and the items, and items that referenced challenging identity management strategies have higher loadings with factor one, and accepting identity management strategies tended to have higher loadings with factor two. One item from accepting identity management strategies: “Trying to bond with others like me” had a higher association with the challenging identity management factor (.65) than it did with the accepting identity management factor (.23). It seems that participants are psychologically processing this as a challenge strategy perhaps because there is a component of proactively doing something about the stigma embedded in bonding with others. While literature on identity management considers this strategy to be an accepting identity management strategy, it is likely that participants’ subjective evaluation of this item could be due to the fact that they see bonding with others as a way to challenge the status quo. Alternatively, their subjective evaluation of this item could also be due to the #MeToo movement. Specifically, women are witnessing bonding with others like themselves as a viable strategy to counter sexual harassment, and this is often framed as a “challenging patriarchy” strategy in the national media and discourse. A reader may question if, in addition to the accepting and challenging behaviors, there may be other dimensions such as active or passive ways of engaging in these behaviors embedded in the current measure. Even though there are items that reflect such a dimensionality, given the EFA presented here, with the exception of one item, this does not seem to be a pervasive issue. 60 Vivid recall task and identity management. The identity management scores were checked for normality and variability for overall IM score, and also for the subscale scores of challenging IM strategies and accepting IM strategies. Results show that skew and kurtosis were within an acceptable range (i.e., absolute value of skew < 2, absolute value of kurtosis < 7; West, Finch, & Curran, 1995). In this Pilot Study, participants in the experimental condition (i.e., vivid recall of a sexist event) were instructed to fill out the identity management strategies scale “given the sexist event they had just described”. For the study to make sense, the participants in the control condition (i.e., those who had shared a vivid recall of any memory they chose to discuss) were told to imagine a hypothetical scenario in which “if they faced sexism” which of the items on the identity management scale would they endorse. Thus, in both the conditions, participants were cued to rate the identity management scale based on either actual reports of how they had in the past navigated a sexist event or how they might navigate a hypothetical sexist event. The primary purpose for the data on the identity management scale was to obtain the psychometric information presented earlier. An independent samples t-test was conducted, and it was found that participants in the sexist vivid recall condition (M = 2.92, SD = .71) versus those in control condition (M = 3.17, SD = .64) did not show statistically significant differences in endorsing identity management strategies; t(68) = -1.52, p = .13. This is in line with the notion that both the control and experimental group participants were cued to sexist events, and therefore there should not be statistically significant mean differences here. Discussion of Pilot A In piloting the vivid recall task it was established that the task serves its purpose as a vehicle for stigma communication. Specifically, retrospective recall of an event where 61 participants faced sexism made participants feel depleted as was evidenced in both their performance on the anagram task and on the self-regulatory depletion scale. The study also established that while the anagram task does serve well as a behavioral measure of depletion, it is also fraught with methodological complexity. Specifically, the data were not normally distributed and there were several outliers that pulled the sample means and obscured key relationships. Unfortunately, the nature of the task does not allow this researcher to confidently assert that these outliers were “error outliers”, and therefore should be deleted. It is likely that the participants truly did spend as long as they did on attempting to solve the unsolvable anagrams, and in fact the open-ended responses and their manipulation check items point to exactly that. Ideally, these outliers should be treated as “interesting outliers” because there are possible research questions embedded in this variance. Given the constraints of this research agenda however, such a stream of research was not deemed possible. That said, when the analysis was run after deleting the outliers, it changed the relationships to significant and in the expected direction. That is, individuals who were asked to think about sexist events and describe those in the vivid recall task spent significantly less time on the anagrams than did the participants in the control condition. In addition, a task order effect was detected in that those who were presented anagrams first were much more depleted when they answered the self-regulatory depletion scale than when the self-regulatory depletion was presented first, and this held even in the control condition. So while embedding two dependent variables can be helpful, it has the potential to reduce sample size for the final studies because to figure out the depletion due to vivid recalls using the self- regulatory depletion scale, the analysis will have to be run by selecting only those participants who were presented the self-regulatory depletion scale first. This, in conjunction with the 62 likelihood of obtaining non-normal data when using the anagram task, and having no theoretical reason to delete the outliers from the analysis, points to potential problems in using the unsolvable anagrams as the dependent variable going further. In contrast, the self-regulatory depletion scale of self-regulatory depletion performed quite well on metrics of normality and outliers, and was psychometrically sound with an excellent reliability alpha. In addition, the vivid recall task elicited enough variance in the self- regulatory depletion scores, and participants in the experimental condition showcased statistically higher depletion than did participants in the control condition. Given the differences between the two dependent variable measures (i.e., the unsolvable anagram task and the self- regulatory depletion scale), the study has provided strong evidence in favor of using the self- regulatory depletion scale over the unsolvable anagram task as the dependent variable of choice going further. A case could be made to include the anagram task by accounting for the order effect such that the self-regulatory depletion scale is always presented first, and then the unsolvable anagram task is presented. A benefit of using two tasks to assess depletion as a dependent variable would be that the research design could avoid the mono-method bias. However, when balanced against the practical concerns such as length of survey and the associated costs (Qualtrics’ cost structures for recruiting participants are contingent on the length of surveys) this strategy is less appealing given the potential non-normality issues outlined earlier about the unsolvable anagram task. Pilot B: Performance Discussion Task Participants. 87 women participated in the Pilot Study and ranged between ages 18 – 71 years (M= 38.98, SD = 12.09). Among the participants, 5.70% were African American, 87.40% were Caucasian, 3.45% were Hispanic, and 2.30% were Pacific Asians, and 1.15%were Native 63 American or American Indian, and 1 % were from other racial/ethnic groups. Please note participants were allowed to select more than one ethnicity. On average, participants reported being in the workforce for 19.3 years (ranged from 3 to 51 years, SD = 11.41). On average, participants reported being at their current job for 6.23 years (Min = less than a year to Max = 30 years, SD = 6.44). Diverse job titles were represented across the participants ranging from administration, analysts, managers, teachers, retailers, service, clerks, healthcare workers, designers, legal assistants, occupational therapists, technicians, and teachers etcetera. Participants were emailed an individual survey link via Qualtrics. The survey took about 12 minutes to complete and was designed as follows: after consenting to participate in the study, participants were randomly assigned to one of four conditions: ‘Belonging goals’, ‘Self- enhancement goals’, ‘Self-verification goals’, and ‘No goals’. They read a cover story about having to work with a virtual team (see Appendix E). Those in belonging goals condition were told that, “in order to effectively perform in your upcoming task with this virtual team, you should make every effort to get along with them. Trying to fit in with your team will help you create the best conditions for success on this task”. Those in self-enhancement goals condition were told that, “in order to effectively perform in your upcoming task with this virtual team, you should make every effort to highlight your work contributions to them. Ensuring that your work gets your team’s attention will help you create the best conditions for success on this task”. For the self-verification goals condition, participants were told that, “in order to effectively perform in your upcoming task with this virtual team, you should make every effort to assert your identity and be who you are in your communication with them. Trying to be assertive about your own uniqueness to your team members will help you create the best conditions for success on this task”. Those in no goals condition were simply told that, “you 64 will shortly be introduced to your virtual team”. To ensure that participants were attentive and engaged when thinking about these goals, they were instructed to write about the goals they had been assigned and how they might accomplish them in two-three sentences as a reflection exercise. In total 59 respondents (Ntotal = 61) filled out these goals: belonging goals (N = 20), self-enhancement (N= 18), and self-verification goals (N = 21). The control group was not asked to write anything. For stigma communication, participants were told that, “To help you meet your goal on this task, we have shared a part of your team’s internal discussion on a previous team member’s performance. We hope that this information will help you better understand the internal team dynamics; by getting a sense of the team’s acceptable work norms and standards, you should be able to achieve your goal and perform well on the upcoming team tasks.” This performance discussion was the stigma manipulation 15 where an all male team is presented as discussing a female colleague’s work performance in sexist manner. Two versions of this manipulation were piloted to assess the manipulation strength (see Appendix E). Manipulation check. Eight items were used to assess the effectiveness of the manipulation. Included were items such as (a) I think this team is sexist and (b) I am a bit 15 A claim could be made that since the participant is not facing the stigma communication as a ‘self-referencing’ event, she may not perceive stigma by just engaging with materials that are sexist towards someone else (i.e., Stephanie in the manipulation). However, literature on stigma consciousness shows how individuals from minority groups are concerned about the increased likelihood of being stereotyped in certain situations (cf., Pinel, 1999), and literature on stereotype threat showcases a concern with one’s behaviors inadvertently confirming the stereotypes about one’s ingroup (cf., Steele & Aronson, 1995). Based on these lines of research it is likely that watching someone from one’s ingroup face stigmatizing situations can create feelings of anticipated devaluation, rejection, and exclusion. In addition, if the respondents bought into the cover story, then they would expect to work with this all male virtual team on a word puzzle task. Given the team’s past explicit sexist behavior, it is likely that the participant may feel a heightened threat of self-referential stigmatization in anticipation of working with this team. 65 worried about having to work with this team (See Appendix E). The items were scored on a Likert-type scale rated from strongly disagree (1) to strongly agree (5). The Cronbach’s alpha for this scale was .86. Two items asked participants about the realism of the performance discussion: (a) Team members in the real world do not talk like this, and (b) I do not believe that team members actually said these things. The items were scored on a Likert-type scale rated from strongly disagree (1) to strongly agree (5). Cronbach’s alpha for the realism scale was .71. In addition, participants were asked to describe the goal condition they were in. Specifically, they were asked to report about the goal they were assigned for this task: (a) My goal was to highlight my work contributions to my team, (b) My goal was to assert my identity and to be true to who I am when interacting with my team, (c) My goal was to make every effort to get along with my team, and (d) I was not assigned a goal. If they failed to match the goal condition to which they were randomly assigned, then their data was not included in the analysis. Experiences with sexism. To establish base rates of frequency of sexism for Study Three, the pilot participants were also asked if they had experienced sexism at their job. Instructions clarified that, “we are asking about both the explicit, in your face type of sexist events, and the sexist events that may have seemed ambiguous, or subtle forms of sexist behaviors directed at you”. Participants also responded to the number of times they had experienced sexism on the job over the past four weeks. Results of Pilot B The study was conducted to evaluate which manipulation story was perceived as more (a) realistic, (b) communicated stigma to participants, and (c) did not result in making participants forget their assigned goals due to the strength of the manipulation. 66 Realism of the performance task manipulation. Participants found the manipulation to be realistic (M = 2.5, SD = .95; where lower means indicate higher realism). For the strong manipulation prompt (N = 39), participants rated the manipulation as realistic (M =2.35, SD = .86). The moderate manipulation prompt (N= 48), as compared to the strong manipulation prompt, was rated as slightly less realistic (M = 2.63, SD = .82). An independent samples t-test was conducted to test the mean difference across conditions and it was non-significant (t = -1.55, p = .13, df = 85). Communication of sexism. I also conducted a check on normality of data based on the pilot conditions (i.e., for sexism ratings). The strong pilot condition, means showed a kurtosis of -.61 and a skew of -.53, while in comparison, the moderate pilot condition means showed a slightly lower kurtosis of -.09 but a slightly higher skew of -.58. Negative skew indicates high values and negative kurtosis indicates a flat and lightly tailed distribution. Analyses indicated that skew and kurtosis were within an acceptable range (i.e., absolute value of skew < 2, absolute value of kurtosis < 7; West, Finch, & Curran, 1995). Q-Q plots show that the moderate condition means trend closer to the trend line than do plots for the strong condition thereby showing that means on sexism tended to be more normally distributed for the moderate condition. Overall participants (N = 87) perceived that the team was sexist (M = 4.06, SD = .67). Participants in the strong manipulation condition (N = 39) and in moderate condition (N = 48) both reported that the team was sexist. An independent samples t-test was conducted to test the mean difference across those in strong manipulation condition (M = 4.10, SD = .64) and those in the moderate manipulation condition (M = 4.02, SD = .70) and it was non-significant (t = .57, p = .57, df = 85). 67 Recall of goals. A total of 20 participants failed the attention checks on the assigned goals. In the moderate sexist condition, 20% of the participants failed the attention check (N = 12; N total = 60). In contrast, 22% of the participants (N = 11; N total = 50) failed the attention check in the strong sexist condition. The sample size was too small to establish a discernable pattern in failing the different goal conditions. Of the 20 participants who failed the check, seven were in the no goals condition and eight were in the belonging goals condition. Those in no goals condition ended up selecting randomly from the list of goals given. These participants were excluded from further analyses. Frequency of sexist events. Another goal of this Pilot Study was to evaluate the frequency of sexist events reported by women at work to help assess whether or not a four-week period would be enough for Study Three. Participants were asked to rate if they had experienced sexism on the job. 50.6% (44 of 87) stated that they had experienced sexism on the job (combining across those who reported facing explicit sexism and those who stated that it may not have been the perpetrator’s intention), 49.4% (43 of 87) stated that they had not faced sexism at work. Participants were also asked how many times over the past four weeks had they faced sexist events in an open-ended question. Of the 44 participants who had experienced sexist behavior at work, twenty-four participants had not experienced any sexist events over the past four weeks. Six had experienced sexist events once, eight had experienced sexist events twice, two had experienced it thrice, three had experienced it four times, and one person had experienced it over ten times. On average, there were 1.2 sexist events that occurred over a period of four-weeks for each participant who stated that they faced sexism at work. Of the 87 participants, there were .61 sexist events that occurred over a period of four-weeks for each participant. 68 Discussion Pilot B found that there was no difference between moderate and strong levels of performance discussion task. As such, a decision was made to use the moderate strength of manipulation because even though it was subtler than the strong manipulation, it was creating the intended effect on participants such that they reported feeling stigmatized. Another goal of this Pilot Study was to assess if a four-week window would be enough to gain about four data points per person for Study Three. Based on the fact that only 1.2 sexist events were observed over a period of four weeks in a general sample, a decision was made to sample women who faced higher incidents of sexism to have enough power for the ESM study. As such, the screening criteria for Study Three were set such that women who reported less than four sexist events over the past four weeks were screened out (for details see Chapter Six). 69 CHAPTER 4: STUDY ONE This study is an attempt to answer if there is differential self-regulatory depletion when people accept or challenge stigma communication. As discussed in Chapter One, if the decision to accept stigma communication is inherently different from the decision to challenge stigma communication in terms of cognitive and/or affective costs to a person, then there is likely differential self-regulatory depletion. Specifically, I posit that when people face stigma communication, those who challenge stigma will show higher self-regulatory depletion versus those who accept stigma. Furthermore, promotion and prevention self-regulatory foci, identity centrality, and self-efficacy are likely to impact the decisions to accept or challenge stigma communication. Specifically, an individual with high identity centrality is more likely to use challenge identity management strategies, and challenge identity management strategies in turn are likely to be related to higher self-regulatory depletion. Similarly, a promotion-focused individual is also likely to show higher self-regulatory depletion, while a prevention-focused individual is more likely to use acceptance identity management strategies and show lower self- regulatory depletion. Note that in this model, stigma communication is not a variable but an operating condition for all participants. The conceptual model to be tested here is presented below and each of the specific hypotheses is also described here (see Table 3). 70 Figure 9: Conceptual model for Study One. Table 3: Hypotheses for Study One. H# H2 Greater use of (a) accepting strategies will be associated with lower self-regulatory Hypothesis H3 H4 H5 H6 H7 H8 H9 H10 depletion and (b) challenging strategies will be associated with higher self-regulatory depletion Individuals’ self-regulatory focus will impact their choice of identity management strategies such that (a) greater promotion focus is associated with greater challenging behavior while (b) greater prevention focus is associated with greater accepting behavior People with higher identity centrality are more likely to show higher self-regulatory depletion People with higher identity centrality are more likely to challenge stigma while people with lower identity centrality are more likely to accept stigma People with higher self-efficacy are more likely to challenge stigma while people with lower self-efficacy are more likely to accept stigma People who perceive a supportive diversity climate are more likely to challenge stigma while people who perceive a non-supportive diversity climate are more likely to accept stigma People who receive stigma communication from those lower in the role hierarchy are more likely to challenge stigma while people who received stigma communication from those higher in the role hierarchy are more likely to accept stigma. (a) Accepting stigma will partially mediate the relationship between diversity climate and self-regulatory depletion and (b) challenging stigma will partially mediate the relationship between diversity climate and self-regulatory depletion (a) Accepting stigma will partially mediate the relationship between role hierarchy and self-regulatory depletion and (b) Challenging stigma will partially mediate the relationship between role hierarchy and self-regulatory depletion 71 (a) Accepting stigma will partially mediate the relationship between promotion focus and self-regulatory depletion and (b) Challenging stigma will partially mediate the relationship between promotion focus and self-regulatory depletion (a) Accepting stigma will partially mediate the relationship between prevention focus and self-regulatory depletion and (b) Challenging stigma will partially mediate the relationship between prevention focus and self-regulatory depletion (a) Accepting stigma will partially mediate the relationship between identity centrality and self-regulatory depletion and (b) Challenging stigma will partially mediate the relationship between identity centrality and self-regulatory depletion (a) Accepting stigma will partially mediate the relationship between self-efficacy and self-regulatory depletion and (b) Challenging stigma will partially mediate the relationship between self-efficacy and self-regulatory depletion H12 H13 H14 Table 3 (cont’d) H11 Power Analysis A power analysis for a priori structural equation models (SEM) was conducted (using http://www.danielsoper.com/statcalc3/calc.aspx?id=89) for an effect size of .15 and alpha of .05. The power analysis calculator recommended a minimum sample size of 500 participants. Since the issue of power in SEM is not straightforward (Quintana & Maxwell, 1999), a power analysis may not yield the best sample size estimates. Thus, two other strategies were used to address this issue. MacCallum, Browne, and Sugawara’s (1996) guidelines on calculating power based on Root Mean Square Error of Approximation (RMSEA) fit index would require at least 230 participants to achieve a power of 0.8 (with 55 degrees of freedom; cf., MacCallum et al., 1996; p. 145). However, according to the rule of thumb of 10 participants per estimate (Bentler & Chou, 1987) the current study should have had 310 participants (for 28 paths and three endogenous variables). This estimate qualified the minimum sample size criterion suggested for SEM (Weston & Gore, 2006; Quintana & Maxwell, 1999) and was also assessed as adequate in simulation studies (e.g., Chou & Bentler, 1995). Thus, 331 participants were sampled for this study (to account for possible attrition 21 additional participants were recruited). Of these, 20 participants failed to respond to key measures in the study. Specifically, 14 participants failed to 72 answer the vivid recall manipulation appropriately (e.g., gibberish responses) and six participants failed to provide any data on the dependent variable measure. These individuals were therefore excluded from further analyses, thereby leaving us with a total sample size of 311 participants. Participants 311 female16 Qualtrics workers were recruited via Qualtrics panels and were compensated based on the incentive structure that participants had previously negotiated with Qualtrics (see Pilot Study for details). Screening criteria. Women who had participated in the Pilot Study or who did not have a specific sexist incident to report were screened out for this study. Also, only respondents who were currently employed (full time or part time), and who were United States citizens were screened in for this study. Participants ranged between ages 18 – 74 years (M= 36.46, SD = 11.99). Among the participants, 11.90% were African American, 78.78% were Caucasian, 7.72% were Hispanic, 2.89% were Asian/Pacific Islander, 1.61% Native American or American Indian, and 1.93% were from other racial/ethnic groups. Note, participants were allowed to check multiple racial/ethnic categories. Participants reported a total work experience ranging from less than a year to 51 years (M= 15.92, SD = 11.07). Participants reported the tenure at their current job ranging from less than a year to 41 years (M= 6.33, SD =6.99). Diverse jobs were represented across the participants ranging from managerial staff, teaching, technology, healthcare, customer service, inventory specialists, supermarket associates, retailers, serving and wait staff in food industry, and manufacturing etcetera. 16 The decision to limit the sample to women has already been described in chapter two. 73 Design In this experiment, a correlational design was used. Participants wrote the stigma communication ‘vivid recall task’, and then the dependent measure was administered (i.e., the SSCCS). Procedure Participants were emailed an individual survey link via Qualtrics. The survey took about 24 minutes (median time) to complete and was designed as follows: after consenting to participate in the study, promotion and prevention self-regulatory foci, and identity centrality measures were administered. Next participants engaged in the Inzchlit and Kang’s (2010) vivid recall paradigm that was piloted earlier. The paradigm was slightly updated to align with the current study and instructed participants to recall sexist incidents that occurred at work. Briefly, participants wrote their vivid recall of a time when they were discriminated against based on their gender at work. The next set of instructions asked participants to select their identity management choices. Instructions stated, “select from the multiple behavioral options below. These behavioral strategies have been described as typical behaviors most people use to address stigmatizing situations such as the one you just narrated. If you are unable to find behaviors that you engaged in, you will then be given the opportunity to share how you felt, thought, or acted in the situation”. Next participants rated the Self-control capacity scale (SSCCS; Ciarocco et al., 2016) to provide a measure of their self-regulatory depletion. Finally, a measure on coping self-efficacy was administered, followed by control variables of trait Conscientiousness and trait anxiety were administered. 74 Manipulation and Tasks Please refer Pilot Study for the manipulation (i.e., vivid recall task; chapter three). Measures Self-regulatory depletion. Based on the Pilot Study results, a 25-item, self-report measure of self-regulatory depletion was administered as the dependent variable in the current study (see Pilot Study for a detailed rationale of why this unpublished measure was used; see SSCCS scale in appendix D). The Cronbach’s alpha for this scale was .95. Identity management strategies. Two seven-item scales were administered to assess accepting and challenging identity management strategies (see Pilot Study for detailed rationale on the items; see scales in Appendix E). In the current sample, the Cronbach’s alpha for the accepting identity management strategies was .66 (for seven items). The Cronbach’s alpha for the challenging identity management strategies was .78 (for seven items). Coping Self-efficacy scale. As described in the Pilot Study, a 13-item measure of coping self-efficacy was adapted to fit the sexist events referenced in the Pilot Study (see Appendix E for all items). As in the Pilot Study, the Cronbach’s alpha for the coping self-efficacy scale was .94. Self-regulatory foci. Promotion and prevention foci were measured by the Work Regulatory Focus scale which is an18-item measure (Neubert, Kacmar, Carlson, Chonko, & Roberts, 2008) that assesses prevention focus (i.e., security, oughts, and losses) and promotion focus (i.e., gains, achievement, and ideals) dimensions respectively. Sample items include: “At work, I am motivated by my hopes and aspirations”, “Fulfilling my work duties is very important to me”, “I take chances at work to maximize my goals for advancement”, and “I focus my attention on avoiding failure at work”. In the current sample, the Cronbach’s alpha for the 75 prevention focus scale was .87 (nine items). The Cronbach’s alpha for the promotion focus subscale was .88 (nine items). Identity centrality. Items of the Centrality dimension of the Multidimensional Inventory of Black Identity measure (MIBI; Sellers et al., 1997) were adapted to operationalize identity centrality in the proposed research. Settles (2004) adapted the MIBI items to operationalize ‘woman centrality’ by changing the racial reference to gender based reference, and the current research follows this model. Items included ‘Overall, being a woman has very little do with how I feel about myself’, and ‘My destiny is tied to the destiny of other females’ (See Appendix E) and were rated on a strongly disagree (1) to strongly agree (7) Likert-type response scale. Recently a dissertation on gender centrality has also used the Settles’ (2004) operationalization (cf., Behrendt-Mihalski, 2017) and reported good reliability index for the adapted Gender Centrality MIBI (α = .83). In the current sample, the Cronbach’s alpha for the identity centrality scale, with eight items, was .77. Perceptions of Diversity Climate. I adapted Button’s (1996; 2001) items to assess the extent to which an organization engages in discriminatory treatment toward sexual minorities. In the current study, I used eight of nine items. Participants responded on a 7-point Likert-type scale ranging from 1 (strongly disagree) to 7 (strongly agree) to items such as, ‘My employer is affirming toward women’, ‘The leaders of this organization are committed to the equitable treatment of female employees’, and ‘The policies of this organization are equitable to women’, ‘This organization takes steps to ensure that women are treated just like men’ (see Appendix E for full measure). The reported reliability of the measure is excellent (α = .97). Please note that while these items sound generic, the directions on the measure were designed to ask participants to think about the employer that is associated with the stigma communication event they had just 76 described. The Cronbach’s alpha for the diversity climate perceptions scale in the current study was .90. Role hierarchy. Since participants’ identity management behaviors could look very different based on whether they were addressing stigma communicated by a peer, supervisor, or clients, I operationalized the stigma communicators’ position in the hierarchy by following King et al’s (2017) operationalization of the ‘organizational position’ variable: “Higher level”, “Same level”, “Below level”, and “Clients” (p. 487). If an event included receiving stigma from multiple people, the response options also allowed participants to select multiple options and this group was coded as “Others”. Thus, four dummy variables were created for role hierarchy. Control Variables. Trait anxiety, and trait conscientiousness were used as two control variables. The 20-item State-trait Anxiety Inventory was used to assess trait anxiety (STAI; Spielberger, 1983; see appendix E for items). The Cronbach’s alpha for this scale was .93. Nine items of the Big Five Inventory (John, Donahue, & Kentel, 1991) were used to assess trait conscientiousness. The Cronbach’s alpha for the trait Conscientiousness in the current sample was .78. Once centered, the Cronbach’s alpha for trait Conscientiousness was .82. Preliminary Analyses Results Prior to performing hypotheses testing analyses, all data were examined for accuracy, missing values, normality, and assumptions of multivariate analysis using SPSS 24.0. Analyses indicated that skew and kurtosis were within an acceptable range with the exception of a small deviation on prevention focus (i.e., absolute value of skew < 2, absolute value of kurtosis < 7; West, Finch, & Curran, 1995). Also, not seen were issues such as range restriction due to ceiling 77 effects or floor effects on any of the variables. Overall, it was deemed that variable transformations were not necessary to proceed with data analysis (see Table 4). Table 4: Means, standard deviations, skew, and kurtosis on variables in Study One. Variables Mean S.D. Skew Kurtosis Promotion focus Prevention focus Diversity climate Identity centrality Trait anxiety Self-Reg Depletion Coping Self-efficacy Conscientiousness Identity management accepting 3.77 4.45 3.60 4.58 43.84 4.86 95.05 3.95 2.90 .70 .54 1.43 1.04 11.15 1.31 27.12 .61 .76 -.50 -2.04 .25 .08 -.05 -.31 -.08 -.14 -.43 0.35 9.50 -.66 -.40 -.48 -.77 -.51 -.69 .18 Identity management challenging 3.01 0.83 -.30 -.28 (Note: Scale Ranges: Promotion focus (Min = 1, Max = 5, Range = 4); Prevention focus (Min = 1, Max = 5, Range = 4); Diversity climate (Min = 1, Max = 7, Range = 6); Identity centrality (Min = 1.88, Max = 7, Range = 5.12); Trait anxiety (Min = 20, Max = 74, Range = 54); Self- regulatory depletion (Min = 1.24, Max = 7, Range = 5.76); Coping self-efficacy (Min = 13, Max = 143, Range = 130); Conscientiousness (Min = 2.38, Max = 5, Range = 2.62); Identity management accepting (Min = .71, Max = 5, Range = 4.29); Identity management challenging (Min = 1, Max = 5, Range = 4) 78 Descriptive Statistics Correlations. The correlations between the 14 variables in this study are described (see Table 5). Self-regulatory depletion was correlated with (a) promotion focus (r =.13, p <.05) and also with prevention focus (r = .13, p <.05); (b) negatively correlated with diversity climate such that the women reported more self-regulatory depletion when their perceptions of the diversity climate were poor (r =-.20, p <.01); (c) negatively correlated with self-efficacy such that the women with higher self-efficacy reported less self-regulatory depletion (r =-.49, p <.01); (d) negatively correlated with accepting identity management strategies such that higher levels of acceptance strategies were related to higher levels of depletion (r =-.26, p <.01); and (e) supervisor’s role hierarchy such that if stigma is communicated by ones’ superiors it is related with higher levels of depletion (r =-.14, p <.05). Accepting identity management strategies were correlated with (a) self-efficacy (r =-.14, p <.01) such that women with higher self-efficacy tended to use less accepting identity management strategies; (b) when sexism was communicated by one’s supervisors women reported higher use of accepting identity management strategies (r =.11, p <.05); (c) however, when sexist behavior comes from individuals at either same level or juniors then women tend to endorse less acceptance strategies (r =-.12, p <.05) In contrast, challenging identity management strategies were correlated with (a) both promotion-focus (r =.31, p <.01) and prevention focus (r =.17, p <.01); (b) diversity climate such that poorer diversity climates elicited lower challenging identity management strategies (r = -.18, p <.01); and (c) women with higher coping self- efficacy were more likely to engage in challenge identity management strategies (r =.28, p <.01). 79 Promotion focus and prevention focus were both moderately correlated in the current sample (r =.39, p <.01). Higgins (1997, 1998) have theorized the two dimensions to be orthogonal, and prior empirical research has supported (Johnson & Chang, 2008b; Lockwood, Marshall, & Sadler, 2005). However, studies also show that promotion and prevention foci are positively correlated. Hmieleski and Baron (2008) found a small positive correlation between the two dimensions (r = .17, p< .05) and Neubert, Kacmar, Carlson, Chonko, and Roberts (2008) found a moderate positive correlation between the two dimensions (r = .52, p <.00). Control Variables. In the current study, two individual difference factors were examined as potential covariates. Associations between outcome variables and trait anxiety as measured by STAI-T and trait Conscientiousness were evaluated. Both control variables had significant and expected correlations such that those with high state-trait anxiety tended to exhibit more self- regulatory depletion (r = -.62, p < .01). As trait anxiety increases, participants’ responses on self- regulatory depletion tended to decrease (note, lower scores on self-regulatory depletion imply higher levels of depletion). Similarly, Conscientiousness was related to self-regulatory depletion (r = .35, p < .01) such that as conscientiousness scores increased, participants’ reported lower levels of self-regulatory depletion. Given the significant zero-order correlations, both the variables were included as covariates in subsequent analyses. 80 Table 5: Correlations between Study One variables. 5 2 3 4 1 .95 .77 6 7 8 9 -.09 12 13 .88 -.03 .87 -.07 10 11 -.03 .93 .90 .07 .14* .13* .13* -.12* .19** .26** -.20** -.62** .30** .09 .39** - .16** Variables 1. Promotion focus 2. Prevention focus 3. Diversity climate 4. Identity centrality 5. Trait Anxiety 6. SR Depletion 7.Self efficacy 8. Conscient. 9. IM accepting 10. IM challenging 11. Role: Supervisor 12. Role: Same Level 13. Role: Junior 14. Role: Client -.09 (Note: SR Depletion = self-regulatory depletion; Conscient. = Conscientiousness; ** Correlation is significant at the 0.01 level (2-tailed); *Correlation is significant at the .05 level (2-tailed). Note: Sample size for Role hierarchy levels for Supervisors (N = 158; 50.80%); Same level (N = 83; 26.69%); Juniors (N= 22; 7.07%); Clients (N = 31; 9.97%); and Others (N = 17, 5.47%; dummy coded referent group). -.28** -.12* -.50** -.54** .21** .22** .49** .35** .94 .34** .09 .07 -.19** -.17** -.18** -.15** -.26** -.17** -.61** -.28** -.34** -.20** .82 .28** .26** .31** .17** .24** -.12* -.12* -.14* -.14* -.14* .11* -.09 -.08 -.09 -.06 -.01 -.00 -.06 -.04 -.05 -.01 -.04 -.06 -.10 -.09 -.04 -.10 -.09 .03 .06 .02 .10 .04 .01 .09 .04 -.02 .07 -.07 .05 .00 .02 .09 - .10 .10 .07 .78 .09 .04 .03 .02 - - .66 .10 .02 81 14 - Vivid Recall Task. Overall, participants in the study provided 311 vivid recalls of a time when they had faced sexist events at work. A few examples of the vivid recalls are presented below (see Table 6). On average, participants wrote an average of 128.81 words (i.e., roughly one paragraph each; SD = 77.15). Table 6: Examples of vivid recalls of sexist events at work for Study One. Participant Examples of vivid recalls of sexist events A Participant B Participant C Many years ago I was an assistant to a male financial broker who was only a few years older than me. He never showed much respect towards women and used to have his stay at home wife model clothing for him in the office after shopping. One time he made sexual comments to me about wanting to swipe his credit card through my cleavage to see if anything came out. This made me uncomfortable and feel disrespected, when in fact my job was to make him look good which he profited from since he worked off of commission. Since we were affiliated with a bank with various branches I felt that he could have said something to any other bank employee which could cause potential legal ramifications and put our company in jeopardy. I said something to our supervisor so she was aware and could address it with him. Afterwards, he told other people that I caused problems for him and tried to get him fired. He eventually quit. I was glad he left but also felt slightly guilty since he had just had a baby. During the 1980's, I was working as a secretary at the Internal Revenue Service. The only other people in my department were three Criminal Investigators, two of which were men in their 40's. They had a reputation for being womanizers. They didn't act in a negative manner towards me, but in their office, they had posters of scantily clad women. They also had Playboy magazines in their office. I felt uncomfortable when entering their office, but I never made any comment to them about how I felt. There was also a photograph which one of them had taken of a naked woman in a compromising position. I'm pretty sure this woman was not one of their wives. I was embarrassed when I saw the photo. I just dealt with the situations by not saying anything to them, but telling my husband about it. While at work a male team leader was always flirting with me. As the days went on it became clear that he was trying to sleep with and telling me that basically if I did I could move up in the company. I really paid it no mind. I declined his offer of course I know how men are. I didn't report it I eventually found a better job so I never did anything about it. 82 Table 6 (cont’d) Participant My co-worker commented on my posterior. He mentioned that it looked good enough to bite. I was very appalled and looked around to see if anyone else had heard but we were alone and the nearest person was about 3 stations down. I felt powerless because even if I told it would be a 'he said she said' thing. I nicely told him that the comment was not appreciated and he complained that I didn't know how to take a compliment. As soon as I could, I changed shifts. D As is evident from above, the vivid recalls varied greatly in their content. As such, these were coded on the following dimensions for exploratory purposes: (a) type of sexist behaviors recalled – in line with prior research (Fitzgerald, Magley, Drasgow, and Waldo, 1999), reports of sexism were coded17 into three broad categories namely gender harassment (i.e., “negative gender-related comments and lewd sexual comments”), unwanted sexual attention (i.e., “unwanted physical contact and pressure to engage in romantic or sexual interactions”), and sexual coercion (i.e., “sexual bribery and threats”); and (b) identity management strategies that participants reported using to cope with the events they had recounted were also coded. Note that accepting and challenging identity management behaviors were coded18 in a manner consistent with the measures used in the empirical analyses. Also raters were allowed to code an incident into multiple categories. For example, a respondent could have engaged in challenging strategies in the moment, but then have described how they avoided the perpetrator later on. In this 17 Before coding data, two coders were trained to identify the types of sexist behaviors, identity management strategies, perpetrators, and emotions using the Pilot Study data. 18 Accepting strategies included (a) accepting sexism, (b) ignoring sexism, (c) keeping the incident to oneself, (d) avoiding sexism, (e) trying to conduct oneself in a manner inconsistent with the stereotypes of being a woman, (f) trying to bond with other women, and (g) trying to use self-deprecating humor. Challenging strategies included (a) speaking up, (b) trying to change things, (c) working harder to prove perpetrators wrong, (d) trying to educate colleagues about stereotypes about women, (e) trying to communicate the inaccuracy of the stereotypes about women, (f) trying to make favorable social comparisons, and (g) denying that there is stigma on me because of being a woman 83 example, raters coded the respondent’s use of both ‘challenging identity management strategies’ and ‘accepting identity management strategies’ to navigate the stigma communication. Upon examination of how the two raters had coded the open-ended data some inconsistencies emerged and were corrected by the author. First, when coding gender harassment, Rater One showcased a consistent pattern of noting unwanted sexual attention as gender harassment, and author corrected these instances. When coding identity management strategies, as compared to Rater One, Rater Two tended to use a very strict definition of challenging identity management strategies. Furthermore, even when respondents did not describe having used any identity management strategies, Rater Two consistently ascribed their lack of response as ‘accepting identity management strategies’, whereas Rater One consistently and accurately coded these vivid recalls as examples of non-responses. Thus the author corrected Rater Two’s patterns of coding described here. The kappa estimates are reported and show statistically significant moderate to substantial agreement between the coders (see Table 7). Table 7: Sexist incident vivid recall qualitative data coded for Study One. Gender harassment Unwanted sexual attention Sexual coercion Accepting identity management strategies used 39% (N = 120) k = .73 Challenging identity management strategies used 46% (N = 144) k = .77 77% (N = 239) k = .83 % (N =311) Interrater reliability Kappa (Note: Kappa values of .41 -.60 indicate moderate agreement, and .61 -.80 indicate substantial agreement). 12% (N = 36) k = .64 3% (N = 9) k = .71 In addition, participants’ reports on who the perpetrator(s) was/were were categorized. Data show that the most common perpetrators at work were supervisors (38.91%, N = 121), followed by peers (28.30%, N = 88), clients (11.58%, N = 36), and finally juniors (0.64%, N = 84 2). Finally, coders also categorized participants’ emotions into themes. Based on the Circumplex Model of Affect (cf. Posner, Russell, & Peterson, 2005; Russell, 1980), these categorized emotions have been represented graphically on a matrix of positive and negative affect with high and low activation (see Figure 10). A clear portrait of sexist behaviors at work emerges from these qualitative data. Figure 10: Emotions coded from qualitative data for Study One. 85 Empirical Model: Main Effects and Mediation Figure 11: Empirical paths with estimates for Study One. (Note: Only statistically significant paths are presented here to ensure a clear graphic; the greyed out paths were non-significant and beta weights for these are included in the hypothesis testing section). Path analysis was used to test the hypothesized mediation model using MPlus version 7.31 (Muthén & Muthén, 2015). Under conditions of stigmatization, accepting and challenging identity management strategies were tested as mediators of the association between the various individual difference factors (i.e., identity centrality, promotion and prevention focus; self- efficacy) and contextual factors (i.e., diversity climate and role hierarchy) and self-regulatory depletion as the dependent variable. The empirical model included paths from each of the independent variables (i.e., identity centrality, promotion and prevention focus, self-efficacy; diversity climate and role hierarchy) to the mediating variables of identity management behaviors, and then from identity management to self-regulatory depletion. A direct path between identity centrality and self-regulatory depletion was also tested. Including the full 86 empirical model yielded an unreadable figure, and as such a reduced empirical model with statistically significant estimates is presented (See Figure 11). The model fit was evaluated by assessing the following: Chi-square, comparative fit index (CFI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR) was used to test overall model fit. In general, nonsignificant chi-squares, CFI ≥ .95, RMSEA ≤ .05, and SRMR ≤ .05 suggest good model fit with the observed data (Hu & Bentler, 1999). In the current sample, non-significant chi-square test was obtained (c2= .033, p = .86). Also obtained were values of CFI = 1.00, RMSEA = 0.00 (p = .90), and SRMR = 0.00, indicating an excellent model fit. A model without control variables was also tested for fit as an alternative model, and the model fit remained stable. Specifically, non-significant chi-square test was obtained (c2 = .033, p = .87). Also obtained were values of CFI = 1.00, RMSEA = 0.00 (p = .90), and SRMR = 0.00, indicating an excellent model fit. The model with control variables was retained as the empirical model since that was the hypothesized model. Next, bootstrapped confidence intervals were computed to test mediation in MPlus by using the CINTERVAL command in conjunction with Bootstrap option (Muthen & Muthen, 1998-2012). Bootstrap option was used because it is (a) computationally stronger at finding indirect effects, (b) better than significance testing based on normal distribution because it makes no assumptions about the sampling distribution (Hayes, 2013; Hayes & Preacher, 2014), (c) most accurate in terms of power (i.e., ability of finding mediation when there is mediation), and (d) results in lowest Type I error rates. Thus, it provides the most accurate estimate of mediation over other methods like Barron & Kenny causal steps approach or the Sobel test of mediation (MacKinnon, 2002). Confidence intervals that did not include zero indicate significant mediation. In the current study, Maximum Likelihood Estimation was used. 87 Hypothesis Testing Hypothesis 2a predicted that the greater use of accepting strategies would be associated with lower self-regulatory depletion. Note self-regulatory depletion was scored such that lower scores indicate higher depletion. Contrary to our hypothesis, it was found that greater use of accepting strategies was significantly associated with higher self-regulatory depletion (β = -.20, p = .02). Hypothesis 2b predicted that greater use of challenging identity management strategies would be associated with higher self-regulatory depletion. This relationship was not statistically significant (β = -.08, p = .25). Hypothesis 3a predicted that individuals’ self-regulatory focus would impact their choice of identity management strategies such that greater promotion focus would be associated with greater challenging behavior. This hypothesis was supported (β = .25, p = .00). In contrast, Hypothesis 3b predicted that individuals’ self-regulatory focus would impact their choice of identity management strategies such that greater prevention focus would be associated with greater accepting behavior. This relationship was not statistically significant (β = -.01, p = .89). Hypothesis 4 predicted that participants with higher identity centrality would be more likely to show higher self-regulatory depletion. This hypothesis test was not statistically significant (β = -.09, p = .12). Hypothesis 5a stated that participants with higher identity centrality would be more likely to challenge stigma. However, this hypothesis test was not statistically significant (β = .03, p = .43). In contrast, Hypothesis 5b stated that participants with lower identity centrality would be more likely to accept stigma. This relationship was also not statistically significant (β = .01, p = .89). 88 Hypothesis 6a predicted that participants with higher self-efficacy would be more likely to challenge stigma. This hypothesis was supported (β = .01, p = .01). In contrast, Hypothesis 6b predicted that participants with lower self-efficacy would be more likely to accept stigma. This hypothesis was not statistically significant (β = -.001, p = .50). Hypothesis 7a predicted that people who perceive a supportive diversity climate would be more likely to challenge stigma. This relationship was not statistically significant (β = -.05, p = .17). Hypothesis 7b predicted that people who perceive a non-supportive diversity climate would be more likely to accept stigma. This hypothesis was not supported and results yielded a null effect (β = .00, p = .99). Hypothesis 8a predicted that individuals who receive stigma communication from those lower in the role hierarchy are more likely to challenge stigma. Four different dummy groups were created to test this hypothesis for each of the following: supervisors, same-level (i.e., peers), juniors, and clients. Results show non-significant findings for the use of challenging strategies when same-level individuals (β = -.06, p = .69), and juniors (β = -.001, p = .99) communicated stigma to respondents. Participants reported using fewer challenging strategies when supervisors communicated stigma to them, however this relationship was also not significant (β = -.18, p = .28). However, participants reported using fewer challenging strategies when their clients communicated stigma to them and this relationship was statistically significant (β = -.39, p = .04) Thus, overall this hypothesis was partially supported. Hypothesis 8b predicted that individuals who receive stigma communication from those higher in the role hierarchy are more likely to accept stigma. Results show that when respondents received stigma from their supervisors (β = .43, p = .00) and from their clients (β = .42, p = .01), 89 they tended to report higher use of accepting identity management strategies. In contrast, a statistically non-significant relationship emerged between using accepting identity management strategies when confronted with stigma communication from their peers (β = .20, p = .14) and their juniors (β = .04, p = .80). Thus, overall this hypothesis was partially supported. Hypothesis 9a to 14b tested for mediation effects19 and results show no statistically significant effects (see Table 8 that includes beta weights for the indirect paths). Table 8: Mediation results of Study One (Indirect paths were specified in MPlus). H# Hypothesis Beta-weight (p- value) for indirect paths 9a 9b Accepting stigma would partially mediate the relationship between diversity climate and self-regulatory depletion β = .00, p = .99 Challenging stigma would partially mediate the relationship between diversity climate and self-regulatory depletion β = .01, p = .48 10a Accepting stigma would partially mediate the relationship between role hierarchy and self-regulatory depletion Accepting identity management strategies did not mediate the relationship between receiving stigma communication from supervisors and participants’ self-regulatory depletion β = -.08, p = .07 Accepting identity management strategies did not mediate the relationship between receiving stigma communication from peers and participants’ self-regulatory depletion β = -.04, p = .23 Accepting identity management strategies did not mediate the relationship between receiving stigma communication from their juniors and associated self-regulatory depletion β = -.01, p = .82 19 To test the mediation effects, I specified indirect effects by using the ‘MODEL INDIRECT’ command. As mentioned earlier, bootstrapped confidence intervals were computed by using the ‘CINTERVAL’ command, and confidence intervals that did not include zero were used to indicate significant mediation. 90 Table 8 (cont’d) Accepting identity management strategies did not mediate the relationship between receiving stigma communication from clients and associated reports of self-regulatory depletion β = -.08, p = .09 10b Challenging stigma would partially mediate the relationship between role hierarchy and self-regulatory depletion Challenging identity management strategies did not mediate the relationship between receiving stigma communication from supervisors and participants’ self-regulatory depletion β = .02, p = .51 Challenging identity management strategies did not mediate the relationship between receiving stigma communication from peers and participants’ self-regulatory depletion Challenging identity management strategies did not mediate the relationship between receiving stigma communication from their juniors and associated self-regulatory depletion Challenging identity management strategies did not mediate the relationship between receiving stigma communication from clients and associated reports of self-regulatory depletion β = .01, p = .77 β = .00, p = .99 β = .03, p = .35 11a Accepting stigma would partially mediate the relationship between promotion focus and self-regulatory depletion β = -.01, p = .70 11b Challenging stigma would partially mediate the relationship between promotion focus and self-regulatory depletion β = -.02, p = .29 12a Accepting stigma would partially mediate the relationship between prevention focus and self-regulatory depletion β = .00, p = .90 12b Challenging stigma would partially mediate the relationship between prevention focus and self-regulatory depletion β = -.01, p = .55 13a Accepting stigma would partially mediate the relationship between identity centrality and self-regulatory depletion β = -.00, p = .90 13b Challenging stigma would partially mediate the relationship between identity centrality and self-regulatory depletion β = -.00, p = .62 14a Accepting stigma would partially mediate the relationship between self-efficacy and self-regulatory depletion β = .00, p = .53 91 Table 8 (cont’d) 14b Challenging stigma would partially mediate the relationship between self-efficacy and self-regulatory depletion β = .00, p = .31 Discussion Stigma communication has been associated with self-regulatory depletion in individuals (e.g., Bair & Steele, 2010; Inzlicht, McKay, & Aronson, 2006). Literatures on stigma and identity management also posit that when people face stigmatizing contexts, they pursue numerous identity management strategies to navigate these negative events (e.g., Meisenbach, 2010; Petriglieri, 2011; Richman & Leary, 2009). One of the key questions that emerged from the literature review presented in Chapter One, and the basis of this study, was if individuals who chose accepting identity management strategies in response to stigmatizing events were differentially depleted when compared to those who chose challenging identity management strategies. The specific goals of this study were to: (a) explore if using accepting versus challenging identity management strategies is associated with differential self-regulatory depletion; (b) underscore how various individual difference variables are implicated in the choice of identity management strategies and how these are associated with self-regulatory depletion under stigmatizing contexts; (c) assess how various organizational context variables were implicated in the choice of identity management strategies and how these are associated with self-regulatory depletion; and (d) explore participants’ vivid recalls of sexism at workplace with the goal to enhance our understanding of the extent and types of sexism faced by women at work, and to also explore the resultant choice of identity management behaviors. Thus, individual difference variables such as identity centrality, self-regulatory foci, and 92 self-efficacy were included as important antecedents of choice of the identity management strategies, which in turn were tested as mediators between the individual factors and self- regulatory depletion reported by respondents. Similarly, contextual variables such as role hierarchy and diversity climate were also included as important situational antecedents to the accepting and challenging identity management strategies. Identity management strategies and self-regulatory depletion under stigma communication Contrary to the hypothesis, it was found that greater use of acceptance strategies was associated with higher self-regulatory depletion. However, greater use of challenge strategies was not associated with lower self-regulatory depletion. The retrospective recall design that was used to cue stigma communication in this study could have also impacted these findings. Specifically, participants were asked to recall a sexist event, and then endorse whether or not they had used accepting or challenging identity management strategies in that particular event. It is likely that in hindsight, the participants who used accepting identity management strategies might perceive their behavioral choices as a failure in safeguarding their own interests, or in standing up against injustice at the time. It is likely that this perception of failure cued their self- regulatory depletion reports in the moment (the instructions asked them to report how they felt in this moment). I find some support of this explanation in the open-ended narrative comments shared by the respondents. For example, women who used accepting identity management strategies, tended to write vivid recalls that include assertions that they are still dealing with the negative outcomes of the sexism they encountered: One of the male co-owners of the business touched me inappropriately in an elevator. This affected me throughout my career towards high-level male leaders. This has always stayed with me and I feel that I should have spoken up when it happened. It was degrading and made me feel cheap. I shortly left the company because I did not feel appreciated for my duties that I performed in my job. 93 Today I feel disappointed in myself for not addressing this inappropriate behavior [emphasis added]. At the time of the action I did not think that I had any right to challenge the co-owner of the company. My young age, along with the desire to impress and maintain a good working relationship, interfered with me reporting this abuse. I have since gained confidence and know that this would not go by without reporting again in my career. Lack of experience was a big contributor to my silence. Women role models in recent years have set the path to take when abused sexually. (Participant A) I had a very sexist supervisor for the past 15 years. I put up with it because I liked my job and the company was great to work for. I still work there but my sexist supervisor was forced to retire. I'm not sure why but I'm glad he's gone. I felt so degraded and disrespected while working under him. Part of what made it bearable was that he would leave early almost everyday, early as in after lunch so I only had to deal with him for about half a day. Specifically he talked down to women and referred to them as bitches. He would make comments about women are this way or that…I was so glad when he left; I'm still working on getting over the sadistic abuse he dealt out though [emphasis added] (Participant B). At a previous job my boss would always grab on me and push me against the wall. I was totally disgusted by him. I was married and so was he and I felt sorry for his wife and family. At the time I needed the job so I just started to ignore his advances and did my job. I was not at that job for long after that because I moved out of state, but to this day it disgusts me [emphasis added]. (Participant C). In contrast, women who had used challenging identity management strategies at the time of the sexist event may have perceived those behaviors favorably when answering this survey. That is, it is likely that they interpreted their challenging identity management strategies as having done all they could do navigate the stigmatizing event and therefore feeling good about their actions. In the open-ended data, women who used challenging identity management strategies either tend to describe their coping behaviors in positive and confident manner or reference positive outcomes as a result of putting their foot down against the injustice – even if a lot of them ended up quitting the job. More importantly, those who challenged the sexist behavior, do not reference negative outcomes that haunt them to this day (unlike many such instances for women who used accepting identity management strategies). A few representative quotes are shared below: 94 I was a previous probationary firefighter with my local fire department, of a nearly full roster of 24 people; I was one of two women. The other woman had explained how unfairly she was treated, but I tended not to believe her because everyone was initially very accepting of me. My fire chief (who is also a co- worker of equal authority at our other job) was the one who had encouraged me to join, but also the one who was sexist against me. I was accepted as a member the same time as a man who was much older than me and not physically capable of fighting fire. He was always given special treatment over me. A specific incident occurred when we both happened to be moving to different towns outside of the department’s jurisdiction. Chief told him that there was always a place in their department for him, and they wouldn't kick him out just for moving away... When I went to the Chief he said I had to choose, either stay in my old place and be with the department, or move and lose my badge. Mind you, at that time I had attended 60% or our stations calls, and my counter part only had 12%. On top of that I completed all mandatory training, yet the other probationary guy was given a pass because he was out of shape. I was somewhere between furious and hurt. For a group that preached about being a family and having so much brotherhood, all my effort was cast aside. I've never taken offense to the word “brotherhood”, I like to think I put the "her" in it, and have always done well in male dominated fields. Physically I made myself sick with anxiety and nausea, when confronted by him face to face I cried from rage, had rapid heart rate. I thought like everything had been ripped out from under me, and even though I did nothing wrong, I couldn't stop him from doing it…Chief thought he could control me with his threat, but I wasn't going to let that happen. I did stay in my old house, but just to prove I wasn't going to let him push me around, I resigned on my own accord and joined a neighboring department that was hurting for people… (Participant D) I was working as a quality manager in a factory about 20 years ago. We had a new plant manager who was much older than the previous manager and was openly behind the times in respect of how woman took on roles in the work force. This manager came in and did not try to get to know anyone and upset our everyday operations causing work flow problems. We (myself and the production supervisor tried to talk to him about our concerns but we fell on deaf ears. On one occasion he made the comment to me in front of another woman and two men that he had never met a woman who did not constantly complain and that did not have a pocket book for a brain. He then told us that women needed to keep in their place. This made me angry and determined to make it very clear to him that I would not tolerate his attitude toward me and the other women in the factory. I approached him about the comment and he blew me off. He continued to be very arrogant and the comments continued. I then went to his supervisor and spoke to them about the incident. He never did accept us women in our roles and made it very difficult for us. At one point I had given this managers supervisor my notice that I would be leaving the company because I was not willing to take his abuse until he was finally replaced. (Participant E) The results of this study should also be interpreted in context of the #MeToo movement – 95 it is likely that women who reported having challenged sexism felt more confident about how in that moment in time it was the “right thing to do” due to the national narrative surrounding sexual harassment issues and how best to deal with such issues; in contrast women who recalled having used accepting strategies, may in the moment, have felt guilt and/or a sense of “not having done enough” thereby endorsing higher depletion. A stronger test of this hypothesis is presented later in Study Three where the issue of retrospective recall of sexist incidents to cue stigmatizing contexts was not a limiting factor. It is also likely that the current findings are in fact not a design limitation, but instead point to how the underlying theory that accepting stigma communication in the moment may lead to less depletion, and challenging stigma in the moment may lead to more depletion, is inaccurate. Role of individual differences in choice of identity management strategies Three key individual difference variables were included in this study and a discussion of the main effects and choice of identity management strategies under stigma communication is presented below: Self-regulatory foci. As expected, respondents with greater promotion focus tended to report using greater challenging identity management strategies. This finding is consistent with literature (Higgins, 1998) that promotion oriented individuals are more likely to be approach oriented in dealing with problems. For example, promotion oriented individuals have been shown to perform better when faced with a difficult task than prevention oriented individuals who tend to quit more readily and promotion focused individuals, as compared to their prevention focused counterparts, also tend to come up with many more alternatives when encountering a problem (operationalized as the unique number of characteristics generated in sorting different members of an overarching category; Crowe & Higgins, 1997). Similarly, 96 promotion oriented individuals have been shown to be more open to change, while prevention oriented individuals tend to prefer stability (Liberman, Idson, Camacho, and Higgins, 1999). This strategic proactivity towards coping with stigmatizing events was evident in our sample as well, and women with higher promotion focus were more likely to use challenging identity management strategies. However, self-regulatory prevention focus did not predict greater accepting behavior. While the measure had good reliability, the moderate positive correlation between promotion and prevention focus suggests that it failed to capture the underlying theory on self-regulatory foci that outlines these as orthogonal dimensions. Identity centrality. There was no relationship between identity centrality and self- regulatory depletion, and between identity centrality and the choice of identity management strategies. Identity centrality is often theorized as an important antecedent in stigma, discrimination, and identity management literatures (Sellers and Shelton, 2003; Settles, 2004). A recent meta-analytic investigation found that people who endorsed higher racial centrality were more likely to report perceived discrimination, and described an indirect link to higher perceived distress (Lee and Ahn, 2013). However, the meta-analysis also reported several mixed effects for the role of racial centrality and distress (in discrimination contexts) that are pertinent to the current findings. First, authors found that as the number of women in the sample increased, the effects of racial identity (as operationalized by two facets of pre-encounter/assimilation and Afrocentrality/racial centrality) and distress significantly decreased. Authors did not offer any explanation of why this effect may have occurred, but given that this study utilized a women only sample, it is possible that identity centrality based on gender may have showcased a similar attenuated effect with self-regulatory depletion. Second, the meta-analysis yielded inconclusive results regarding the role of racial 97 centrality in predicting distress, and authors asserted that the facet of racial centrality could both buffer and exacerbate perceived discrimination and reported distress. While authors found that racial centrality positively related to perceived discrimination (r = .20, k = 17), it was negatively related to reported distress (r = -.04, k = 15). Note that in the current study, physiological and psychological depletion items comprised the self-regulatory depletion scale and were, at least in form, similar to the notion of distress in the meta-analysis. This can be explained by taking a developmental/experiential view of dealing with various “–isms” in life. Having a stigmatized identity, and knowing that it is a stigmatized identity, can also help people override the negative effects of stigma based on developmental experiences at navigating such experiences. Third, meta-analytic findings also show that while racial identity predicted distress and perceived discrimination, ethnic identity did not. This is important for the current study because an assumption embedded in the design of our study was that identity centrality based on gender would have similar explanatory power in predicting outcomes as racial centrality tends to have. This assumption that different forms of identities are just as salient in discrimination or stigmatization contexts was unwarranted, and needs to be put to test in future research. While the “self” can have multiple layers, and all these layers may be important to one’s sense of identity, these layers are also said to live as an “organized structure” (Stryker and Serpe, 1994; p. 16). As we navigate a complex world, in any given moment, how likely is it for a person to know to what extent his/her behaviors are guided by all the different identity attributes that coalesce together to produce a sense of one’s “whole self”? Perhaps identity salience, that is when people confront situations when some aspect of one’s identity is made more salient, might be a better predictor than identity centrality? Stryker and Serpe (1994) asked a more fundamental question about whether or not identity centrality and identity salience are equivalent, orthogonal, or if they 98 overlap, or if one is a better construct to include in identity related research, but found no clear answer. Their recommendation to use both salience and centrality would perhaps have helped the current study. Future research should investigate the role of identity centrality by shining a light on its construct space, and creating empirical research agendas that do not assume all forms of identities are equally “central”. Another possibility for the lack of expected relationships in the current study is that participants had difficulty in responding to the measure. Stryker and Serpe (1994) in their discussion note that the measurement of psychological centrality about one’s identity requires “a self-conscious or self-aware actor whose own specification of what identities are more or less important constitutes the essence of the concept” (p. 34). It is likely that in the current research study, the respondents were not as self-conscious about their gender identity, or alternatively were unable to respond to the more abstract items on the measure. While the identity centrality measure used in the current research has been previously used, it is likely that some of the more abstract items on the measure such as “my destiny is tied to the destiny of other women”, “being a woman is unimportant to my sense of what kind of woman I am”, “overall, being a woman has very little do with how I feel about myself” may have not translated well for lay audience. Coping Self-efficacy. In Chapter One, a case was made to view identity management strategies as expectancy bound responses (Eccles and Wigfield, 2002; Carver and Scheier, 1981). Self-efficacy (operationalized as coping self-efficacy) was theorized as an important attribute that may explain why women may choose accepting or challenging identity management strategies to navigate gender-based stigmatizing situations. In line with the expectations, results show that women who reported higher self-efficacy to navigate sexist situations, tended to use more challenging identity management strategies. In contrast, there was no relationship between 99 lower self-efficacy and choice of identity management strategies. To conclude, individual difference constructs such as the ones used in the current study are often tested in lab-based research that uses neatly conceptualized tasks. Such tasks and research paradigms invariably exclude the complex milieu of extraneous life factors that in the real world would influence an individual’s decision. For example, promotion focus and prevention focus studies have been primarily lab-based studies (cf., Crowe and Higgins, 1997; Lee and Aaker, 2004; Liberman, Idson, Camacho, and Higgins, 1999; Liberman, Molden, Idson, and Higgins, 2001). In contrast, the current study was a correlational design, and the open-ended recalls make it evident that the respondents were describing a whole host of factors that influenced their decisions to accept or challenge the stigma communication. It is possible that statistical significance of the effects on constructs like self-regulatory foci of promotion and prevention, and identity centrality were dampened due to a combination of such factors. How might these different factors interact to create a set of conditions that guide an individual’s response in stigma events is a question that needs to be evaluated in future research. Investigations that explore the role of self-regulatory foci and identity centrality in accepting and challenging identity management behaviors using mixed-methods designs are needed as these would allow for a more thorough examination of the phenomenon. While it may seem to be a herculean undertaking to extract all the possible factors that may be implicated in one’s decision to accept or challenge stigma in the moment, I do think that using grounded theory methods to inform subsequent empirical work are sorely needed in identity management research. For example, for the current question, researchers could start with interview studies of individuals who have faced stigma, and who can therefore shed light on a complex of factors that tend to exist in such stigmatizing situations. 100 Role of context in choice of identity management strategies Studies show that diversity climate is an important attribute in a person’s decision matrix of whether or not to speak up at work (e.g., Clair, Beatty, and MacLean, 2005; DeJordy, 2008; Hitchcock and Wilson, 1991; King, Reilly, and Hebl, 2008; Swim and Thomas, 2006). Though I had expected that perceptions of diversity climate would influence the choice of identity management strategies when women confronted sexist events, this was not supported in the current study. In contrast, in line with prior research, role hierarchy emerged as a better predictor of choice of identity management strategies (Phillips, Rothbard, and Dumas, 2009). Specifically, when women reported that clients were the perpetrators of sexist behaviors, they were less likely to use challenging identity management strategies. In contrast, when they received stigma communication from their supervisors or from their clients they reported using more accepting identity management strategies. Taken together, findings from Study One point to the possibility that women who encounter sexism-related events (i.e., gender harassment, unwanted sexual attention, and sexual coercion) may react to a “micro-climate” where they are much more sensitized to the power differential between the perpetrator and themselves, than they are to the overall diversity climate of an organization. Thus, power differential inherent in the role hierarchy may emerge as a moderator such that the overall diversity climate of an organization as seen in policies against such behaviors (distal force) is likely to be magnified or stymied depending on the immediate threats at hand (proximal force). Recently, Yang, Mossholder, and Peng (2007) found that power distance moderated the relationship between procedural justice climate and organizational commitment and organizational citizenship behaviors such that lower power distance magnified the positive effects of procedural justice climate. This is not to say that diversity climate does not 101 matter – if anything, diversity climate may play an important role in how power manifests in interpersonal relationships at work. That said, future research should examine if it is the case that the written policies and diversity climate are predictive of general feelings of safety at work, and/or predictive of how power equations manifest in a system. That is, in a system marked by a positive diversity climate, is it the case that even the most powerful individuals practice responsibility and accountability (such that they approach their roles in a manner that ensures that the power equations are never so tilted that their juniors/colleagues are rendered unable to engage in voice behaviors)? Research also needs to examine if employees who have confronted a negative event, are more likely to focus on the more proximal variables such as role hierarchy in making decisions about when to speak up against an affront to self, and if so, how to speak up. Limitations This study had two key limitations, both related to the retrospective recall design. First, the design is a retrospective recall design where participants were asked to think about an event from their past when they faced sexism. Issues of retrospective recall are documented elsewhere in reviews and are therefore not discussed here (cf., Arvey and Cavanaugh, 1995; Hassan, 2005; Miller, Cardinal, and Glick, 1997). It is likely that participants’ ability to accurately report on some variables such as diversity climate of the organization where they had faced the sexist event could have resulted in the non-significant findings for diversity climate and choice of identity management strategies. I did not ask participants to report on when the sexist event occurred, so it is unclear if some were describing an event from recent past or from several months or years ago. The issue of retrospective recall outlined earlier for diversity climate, could just as well impact the finding that vivid recalls of sexism were related to self-regulatory depletion. In fact, 102 when combined with another limitation of this study, the correlational design, this issue could be magnified. That is, participants’ self-reported self-regulatory depletion in response to writing the vivid recalls is just a correlational effect in the current study, and not a causal effect. However, the Pilot Study described in chapter three was an experimental study where participants were randomly assigned to vivid recall and control conditions. The current correlational study replicates the finding that vivid recall of sexism is positively associated with self-regulatory depletion that was obtained in the experimental design earlier. If retrospective recalls are implicated in attenuated effects, then the fact that this effect was replicated even with a retrospective recall design only bolsters my confidence in this finding. 103 CHAPTER 5: STUDY TWO This study is an attempt to answer if under differential goal conditions people select different identity management strategies, and how the different identity management strategies may in turn be related to differential self-regulatory depletion. I used both empirical and exploratory methods to test the key ideas in this study. As discussed in chapter one, people are likely to choose different identity management strategies depending on their goals of self- verification, self-enhancement, or belonging. Specifically, I posit that when people have self- enhancement goals they are more likely to choose challenging strategies (i.e., making favorable social comparisons, simple denial) and accepting strategies (i.e., recategorizing20). In contrast, when people have belonging goals they are more likely to choose accepting strategies (i.e., passive acceptance, self-deprecating humor, and bonding with others who are stigmatized), but no challenge strategies. When people have self-verification goals they are more likely to choose challenging strategies (i.e., making favorable social comparisons, logical denials, and educating others) and accepting strategies (i.e., bond with others who are stigmatized, and recategorizing). As before, identity centrality and self-efficacy will impact the choice of strategies and therefore the resultant self-regulatory depletion. Finally, as in Study One, stigma communication has not been included as a variable in the design, but is instead treated as an operating condition on all participants in the study. The conceptual model to be tested in Study Two is presented below (see Figure 12) and each of the specific hypotheses are also described (see Table 9): 20 Recategorizing was operationalized by the item ‘trying to conduct myself in manner inconsistent with the stereotypes of being a woman’ 104 Figure 12: Conceptual model for Study Two. Table 9: Study Two hypotheses. H# H2 H4 H5 H6 Hypothesis People who (a) accept stigma will show lower self-regulatory depletion when compared to people who (b) challenge stigma People with higher identity centrality are more likely to show higher self-regulatory depletion People with higher identity centrality are more likely to challenge stigma while people with lower identity centrality are more likely to accept stigma People with higher self-efficacy are more likely to challenge stigma while people with lower self-efficacy are more likely to accept stigma H15 An individual’s goals during a stigma communication event will be differentially related to the IM strategies an individual chooses to cope with stigma21 21 These hypotheses have not been explicitly crafted to include specific strategies under different goals because of two reasons. First, the relationships proposed in chapter one emerged from a rational mapping instead of previous empirical or theoretical work. Second, while I do expect that under different goals people are likely to select differential strategies, I expect them to select strategies from both accept and challenge strategy bins, with the exception of belonging goals where people are not likely to select challenge strategies. Under self-verification goals, I expect people to bond with stigmatized others and recategorize (Accepting stigma) and make favorable social comparisons, logical denials, and educate others (challenging stigma). Under self- enhancement goals, I expect people to recategorize (accepting stigma), and make favorable social comparisons and simple denials (challenging stigma). Finally, under belonging goals I expect women to use passive acceptance, self-deprecating humor, and bonding with stigmatized others strategies (accepting stigma) and no challenging stigma strategies. 105 (a) Accepting stigma will mediate the relationship between identity centrality and self- regulatory depletion and (b) Challenging stigma will partially mediate the relationship between identity centrality and self-regulatory depletion (a) Accepting stigma will partially mediate the relationship between self-efficacy and self-regulatory depletion and (b) Challenging stigma will partially mediate the relationship between self-efficacy and self-regulatory depletion (a) Accepting stigma will mediate the relationship between goals and self-regulatory depletion and (b) challenging stigma will mediate the relationship between goals and self-regulatory depletion H14 H16 Table 9 (cont’d) H13 Power Analysis A power analysis for a priori structural equation models (SEM) was conducted (using http://www.danielsoper.com/statcalc3/calc.aspx?id=89) for an effect size of .15 and alpha of .05. The power analysis calculator recommended a minimum sample size of 600 participants. Since the issue of power in SEM is not straightforward (Quintana & Maxwell, 1999), a power analysis may not yield the best sample size estimates. Thus, two other strategies were used to address this issue. MacCallum, Browne, and Sugawara’s (1996) guidelines on calculating power based on Root Mean Square Error of Approximation (RMSEA) fit index would require at least 258 participants to achieve a power of 0.8 (with 45 degrees of freedom; cf., MacCallum et al., 1996; p. 145). However, according to the rule of thumb of 10 participants per estimate (Bentler & Chou, 1987) the current study should have 260 participants (for 23 paths and three endogenous variables). This estimate qualifies the minimum sample size criterion suggested for SEM (Weston & Gore, 2006; Quintana & Maxwell, 1999) and has also been assessed as adequate in simulation studies (e.g., Chou & Bentler, 1995). Thus, the sample size of 260 participants seems adequate for the current study. To account for possible attrition or bad quality data an additional 50 participants were recruited for the study. 106 Participants 314 female participants were recruited via Qualtrics panels and were compensated based on the incentive structure that participants had previously negotiated with Qualtrics (see Pilot Study for details). The decision to limit the sample to women has already been described in chapter two. I also screened for respondents who were currently employed (full time and/or part time). Respondents who had already participated in Pilot Study and Study One were screened out22 of this study so they were not exposed to the study treatment. Participants ranged between ages 18 – 74 years (M= 36.46, SD = 11.99). Among the participants, 11.90% were African American, 78.78% were Caucasian, 7.72% were Hispanic, 2.89% were Asian/Pacific Islander, 1.61% Native American or American Indian, and 1.93% were from other racial/ethnic groups. Note, participants were allowed to check multiple racial/ethnic categories. Participants reported a total work experience ranging from less than a year to 51 years (M= 15.92, SD = 11.07). Participants reported the tenure at their current job ranging from less than a year to 41 years (M= 6.33, SD =6.99). Diverse jobs were represented across the participants ranging from managerial staff, teaching, technology, healthcare, customer service, inventory specialists, supermarket associates, retailers, serving and wait staff in food industry, and manufacturing etcetera. Design In this experiment, a posttest-only design with nonequivalent groups (cf., Shadish, Cook, & Campbell, 2002) was used. Participants were first randomly assigned to either ‘self- enhancement goals’, ‘self-verification goals’, ‘belonging goals’, or ‘no-goals’ condition. Then 22 Please note, this study did not require screening out participants who may not have sexist incidents to report, so unlike other studies, this was not included as a screening criterion here. 107 the identity management measures were administered, and finally the dependent measure was administered. Procedure Participants were emailed an individual survey link via Qualtrics. The survey took about 22.5 minutes (median time) to complete and was designed as follows: after consenting to participate in the study, participants were randomly assigned into one of four goal conditions: self-enhancement goals, belonging goals, self-verification goals, and no goals (please see the Appendix C for all the instructions). After assignment into conditions, participants read the performance discussion manipulation that was piloted previously (see Chapter Three; and see Appendix E for details). The next set of instructions asked participants to “think about the goal you were assigned to work with your virtual team. Given this goal, and given what you know about the internal team dynamics and norms, please tell us which of the following two behavioral strategies are you likely to use when you work with your teammates on the upcoming word scrambling task…” (See Appendix C and E for full instructions and measures respectively). Next participants selected from two bins of identity management strategies (i.e., accepting and challenging identity management), and then also rank-ordered different strategies within each bin from ‘most likely to use’ (7) to ‘least likely to use’ (1). They were given an option to write-in a strategy in an open- ended question if they did not see the strategies that would like to use in the options presented to them. Next participants rated the Self-control capacity scale (Ciarocco et al., 2016) to provide a measure of their self-regulatory depletion. Finally, measures such as identity centrality, self- efficacy, and control variables such as Conscientiousness and trait anxiety were also 108 administered. Finally, an anagram task was presented to the participants. Note, this task was used as a filler task to make the overall experiment seem realistic. Manipulation and Tasks Please refer Pilot Study for the manipulation (i.e., performance task; Chapter Three). Manipulation Check. A single item measure was used as a manipulation check where by the participants were asked to report on the goal they were assigned at the beginning of the survey. This item was established as a quality check criterion with Qualtrics and was also used as an attention check to ensure that the researcher only paid for data that could be used. If participants failed this check, Qualtrics excluded them as inattentive participants and their data were not included in the analysis. Sexism Check. The performance discussion prompt was used to communicate sexism and was assessed by using eight items such as, ‘team members seem rude to women’, ‘team members are aggressive towards women’, and ‘I am a bit worried about having to work with this team’ (M = 3.89; SD =.86). The internal consistency for the scale in the current sample was Cronbach’s alpha .91. Overall, participants were deemed to have seen their teammates as sexist. Believability Check. I also tested for how believable the performance discussion prompt was, and respondents rated two items ‘team members in the real world do not talk like this’, and ‘I do not believe the team members actually said these things’ (M = 2.53; SD = 0.95). Note lower scores on a five-point Likert type scale (strongly disagree to strongly agree) indicated that respondents in the current study tended to believe the manipulation as reflective of a performance discussion that may have occurred in an organization. The internal consistency for the scale was Cronbach’s alpha .72. Overall, participants were not excluded from the analysis, 109 since the manipulation was deemed to have produced the intended effect and was seen as a believable prompt. Measures Please see the Pilot Study and the Study One for all the measures (i.e., accepting and challenging identity management strategies, identity centrality, self-efficacy). Self-regulatory depletion. Details of this 25-item, self-report measure of self-regulatory depletion have already been presented in both the Pilot Study and Study One. In the current sample, the Cronbach’s alpha for self-regulatory depletion scale was .95. Identity management strategies. The 7-item measure for accepting identity management strategies and the 7-item measure for challenging identity management strategies were administered (see the Pilot Study, Study One, and the Appendix E for details on the measure). In the current sample, Cronbach’s alpha for the accepting identity management strategies was .66. The Cronbach’s alpha for the challenging identity management strategies was.78. Coping Self-efficacy scale. As in the Pilot Study and Study One, the Cronbach’s alpha for the coping self-efficacy scale was .94. Identity centrality. Details of this measure were presented in Study One. The Cronbach’s alpha for the identity centrality scale in the current sample was .77. Control Variables. As in Study One, this study used trait anxiety, and trait conscientiousness as control variables. The Cronbach’s alpha for trait anxiety in the current sample was .93. The internal consistency reliability for trait conscientiousness in the current sample was slightly higher than in Study One, with Cronbach’s alpha of .83. 110 Empirical Model, Main Effects, and Mediation Path analysis was used to test the hypothesized mediation model (See Figure 13) using MPlus version 7.31 (Muthén & Muthén, 2015). Under conditions of stigmatization, accepting and challenging identity management strategies were tested as mediators of the association between different goals (i.e., self-enhancement, self-verification, belonging, and no goals) and individual difference factors (i.e., identity centrality and coping with sexism self-efficacy) and self-regulatory depletion as the dependent variable. Two control variables were also included (i.e., trait anxiety and trait Conscientiousness). The empirical model included paths from each of the independent variables to the mediating variables of identity management strategies, and then from identity management to self-regulatory depletion. A direct path between identity centrality and self-regulatory depletion was also tested. Including the full empirical model yielded an unreadable figure, and as such a reduced empirical model with estimates is presented (See Figure 13). In the current study, identity management strategies were assessed in three ways. First, I was interested in exploring if based on proximal goals participants were more likely to select either accepting or challenging identity management strategies. This model was treated as the hypothesized model. 111 Figure 13: Empirical paths with estimates for Study Two. (Note: All the greyed-out paths in the hypothesized model were non-significant) Second, I wanted to explore more nuanced relationships, specifically I wanted to understand how depending on different proximal goals, participants’ choice of specific accepting or challenging identity management strategies changed. This model was used for exploratory analyses. Finally, a third exploratory model that treated the ordinal data as continuous data was also tested and the results are reported later. Results To achieve the first goal, participants were asked to first select between accepting strategies and challenging stigma strategies. This data was thus dichotomous in nature. That is, proximal goal conditions predicted a dichotomous dependent variable (either accepting or challenging strategies). The analytic plan therefore included running a logistic regression23 23 Logistic Regression operates by applying a non-linear log transformation to the predicted odds ratio, and can handle continuous, ordinal and nominal data as IVs. Logistic regression requires 112 model (alternatively, a logit model) to accommodate the binary dependent variable. In Mplus, the dichotomous mediator variable (dummy coded) was classified as ‘Categorical’, WLSMV estimator was selected, and then both overarching accepting and challenging strategies were regressed on the independent variables. Setting the dependent variable as categorical ensured that Mplus selected ‘Logit’ as the link function. Next, bootstrapped confidence intervals were computed to test mediation in MPlus by using the CINTERVAL command in conjunction with Bootstrap option (Muthen & Muthen, 1998-2012). Please note that Mplus does not return any fit statistics when running a logistic regression. Please also note, results reported for hypothesis testing use this model. To achieve the second goal, that is to explore more nuanced relationships between different proximal goals and participants’ choice of specific accepting or challenging identity management strategies, participants were asked to rank-order the identity management strategies they were most likely to use given their proximal goal from within the specific “bin” they had selected. Specifically, the participants were asked to rank-order the seven identity management strategies for each of accepting stigma and challenging stigma bins and the participants rated their “most likely” choice (7) to their “least likely” choice (1) given their proximal goals. Given the rank-ordered nature of the response options, the data were polytomous and ordinal (cf., Skrondal & Rabe-Hesketh, 2003). As such rank-ordered logistic regression24 with Maximum Likelihood estimator25 is typically used. However, one concern with using the ordered that the “factor level 1 of the dependent variable should represent the desired outcome” because it assumes that “P(Y=1) is the probability of the event occurring” (Statisticssolutions.com; assumptions of logistic regression). 24 Rank-ordered logistic regression is alternatively also known as the Plackett-Luce model, exploded logit model, and as the choice-based method of conjoint analysis (source: stata.com) 25 Source: https://stats.idre.ucla.edu/mplus/dae/ordinal-logistic-regression/ 113 logistic regression is that the analysis cannot be run if some of the cells are returned empty – therefore, first a crosstab of proximal goal conditions and response variables on rank-ordered identity management strategies was conducted. Data showed that there were several cells that were returned empty (specifically, on identity management strategies of ‘accepting what is said’, ‘keeping to myself’, ‘self-deprecating humor’, ‘trying to change things’, ‘make favorable social comparisons’, and ‘denying stigma attributions’). Thus, a rank-ordered logistic regression could not be run on the overall model. To deal with this issue, the data were analyzed by splitting the sample into two: (a) those who selected the accepting strategies and (b) those who selected the challenging strategies. Two independent path-models were run to include paths from all the independent variables, to the individual strategies of accepting (or challenging) and then to the dependent variable of self- regulatory depletion. The rank-ordered accepting variable returned an empty cell in ‘accepting sexism’ identity management strategy, and the rank-ordered challenging variable returned an empty cell in ‘speaking up’ identity management strategy. As such the two strategies with empty cells were excluded from further analyses. To run the ordered logistic regression, the WLSMV estimator was used and identity management strategies were regressed on the independent variable (i.e., proximal goals). Results from this exploratory model are presented later (see Table 13 and Table 15). Finally, given that the accepting and challenging identity management strategies were rated on a ‘most likely to use’ (7) to a ‘least likely to use’ (1) response scale, the data were also analyzed by treating the ordinal data as continuous data. Findings from this model are also tabulated as a comparison to the main model later (see Table 20 and Table 21). 114 Preliminary Analyses Prior to performing hypotheses testing analyses, all data were examined for accuracy, missing values, normality, and assumptions of multivariate analysis using MPlus SAMPSTAT command. Analyses indicated that skew and kurtosis were within an acceptable range (i.e., absolute value of skew < 2, absolute value of kurtosis < 7; West, Finch, & Curran, 1995); therefore, transformations were not necessary (see Table 10). Descriptive Statistics Table 10: Means, standard deviations, skew, and kurtosis of Study Two variables. Variables Self-reg. depletion Identity centrality Self-efficacy Trait anxiety Conscientiousness Mean 5.00 4.46 100.79 40.38 4.04 S.D 1.28 1.02 25.02 10.58 .66 Skew -.28 .07 -.27 .07 -.30 Kurtosis -.87 -.08 .10 -.33 -.66 Correlations. The correlations between the nine variables in this study are described (see Table 11). Self-regulatory depletion was correlated with (a) self-efficacy such that higher the self-efficacy, lower the self-regulatory depletion scores (r =.51, p <.01); and (b) negatively correlated with identity management strategies such that the women who chose more accepting strategies, in comparison to challenging identity management strategies, showcased higher depletion (r =-.16, p <.01). Identity management strategies were correlated with (a) identity centrality (r =-.13, p <.05) such that women with higher identity centrality tended to use more challenging identity management strategies in comparison to accepting identity management strategies; (b) women with higher coping self-efficacy tended to use more challenging identity management strategies in comparison to accepting identity management strategies (r =-.12, p <.05); and (c) belonging goals such that women in belonging goal condition tended to use more accepting identity 115 management strategies as opposed to challenging identity management strategies (r =.14, p <.05). In addition, manipulation check (on whether or not the team was seen as sexist) was correlated with identity centrality (r =.28, p <.00), and identity management strategies (r =.23, p <.00). Manipulation’s believability (i.e., manipulation as a real-world performance discussion) was correlated with coping self-efficacy (r =.12, p <.04), and with trait anxiety (r =- .13, p <.02). Table 11: Study Two correlations. 2 1 3 4 6 7 8 9 .95 .06 .51** -.65** .61** .77 .09 -.13* .13* .94 -.54** .51** .93 -.60** 5 .83 -.16** -.13* -.12* .16** -.21** - .00 -.03 .03 -.01 -.02 -.02 - - -.35** - Variables 1. Self-reg. depletion 2. Identity centrality 3. Self-efficacy 4. Trait anxiety 5. Conscientiousn. 6. IM Strategy (1 = Accept, 0 = Challenge) 7. Goal: self enhancement 8. Goal: self verification 9. Goal: belonging .01 .03 .13* -.09 -.01 -.07 .02 .01 .00 .02 -.07 .14* -.34** -.31** (Note: ** indicates p<.01; * indicates p<.05); Conscientiousn. is a label for trait Conscientiousness). Control Variables. In the current study, two individual difference factors were examined as potential covariates. Associations between outcome variables and trait anxiety as measured by STAI-T and trait Conscientiousness were evaluated. Both control variables had significant and expected correlations such that those with high state-trait anxiety tended to exhibit (a) more self- regulatory depletion (r = -.65, p < .01) such that as trait anxiety increases, participants’ responses on self-regulatory depletion tended to decrease (note, lower scores on self-regulatory depletion 116 imply higher levels of depletion); (b) lower identity centrality (r =-.13, p <.05) such that as trait anxiety increased, reported identity centrality decreased; (c) lower coping self-efficacy (r =-.54, p <.01) such that as trait anxiety increased, respondents also reported lower coping self-efficacy; and (d) lower identity management behaviors (r =-.12, p <.05) such that those with higher trait anxiety tended to use more accepting identity management behaviors in comparison to challenging identity management strategies. Similarly, Conscientiousness was related to self-regulatory depletion (r = .35, p < .01) such that as conscientiousness scores increased, participants’ reported lower levels of self- regulatory depletion. Given the significant zero-order correlations, both the variables were included as covariates in subsequent analyses. Hypothesis Testing Hypothesis 2 predicted that the greater use of accepting strategies would be associated with lower self-regulatory depletion while greater use of challenging identity management strategies would be associated with higher self-regulatory depletion26. While this effect was in the same direction as in Study One, this effect was not statistically significant in the current study (β = -.04 p = .55; note: accept and challenge identity management strategies in the current study were dummy coded as categorical variables where accepting strategies were coded as ‘1’ and challenging strategies were coded as ‘0’, hence there is just one beta weight). Hypothesis 4 predicted that participants with higher identity centrality would be more likely to show higher self-regulatory depletion. This relationship was not statistically significant (β = -.06, p = .24). 26 Lower scores on self-regulatory depletion indicate higher depletion 117 Hypothesis 5a stated that participants with higher identity centrality would be more likely to challenge stigma. In contrast, Hypothesis 5b stated that participants with lower identity centrality would be more likely to accept. This hypothesis was also not supported (β = -.12, p = .12). Hypothesis 6a predicted that participants with higher self-efficacy would be more likely to challenge stigma. Hypothesis 6b predicted that participants with lower self-efficacy would be more likely to accept stigma. This hypothesis was not supported (β = .01, p = .85). Hypotheses 13, 14, and 16 predicted mediated effects between the independent variables and self-regulatory depletion, however none of these relationships were statistically significant. Specifically, Hypothesis13a predicted that accepting stigma would partially mediate the relationship between identity centrality and self-regulatory depletion, while Hypothesis 13b predicted that challenging stigma would partially mediate the relationship between identity centrality and self-regulatory depletion. These hypotheses were not supported (β = .01, p = .64). Hypothesis 14a predicted that accepting stigma would partially mediate the relationship between self-efficacy and self-regulatory depletion, while Hypothesis 14b predicted that challenging stigma would partially mediate the relationship between self-efficacy and self-regulatory depletion. These hypotheses were not supported (β = .00, p = .93). Hypothesis 16a predicted that accepting stigma would mediate the relationship between each of the three goals and self- regulatory depletion and Hypothesis 16b predicted that challenging stigma would mediate the relationship between each of the three goals and self-regulatory depletion. These hypotheses were also not supported (βSE = .00, p = .92; βSV= .00, p = .88; βB = -.01, p = .65; note for subscripts: SE = self-enhancement goals, SV = self-verification goals, and B = belonging goals). 118 Hypothesis15 predicted that an individual’s goals during a stigma communication event would be differentially related to the identity management strategies an individual chooses to cope with stigma. While a different pattern of relationships emerged for different goals and identity management strategy choice, these relationships were not statistically significant (i.e., self-enhancement and identity management (β = -.01, p = .86); self-verification and identity management (β = -.07, p = .77); and belonging and identity management (β = .36, p = .16). These relationships between goals and choice of identity management strategies were also tested outside the path analytic model (see section below). Goals and choice of general identity management strategies. In chapter one, I had posited that under different goals, people would be likely to select differential identity management strategies. While I had expected that respondents in self-verification and self- enhancement goals may choose from both accepting and challenging identity management strategies, I had expected that respondents in belonging goal condition would likely not choose challenging identity management strategies. Data show that participants in all goal conditions endorsed both accepting and challenging identity management strategies as possible ways to navigate the stigma situations at work, although those in self-verification and self-enhancement goals tended to choose challenging identity management strategies more often than they did accepting identity management strategies. Specifically, it was posited that under belonging goals more people would endorse accepting identity management strategies compared to those who were in self-enhancement or self-verification goals. Descriptive data show that 55.41% of women who had belonging goals indicated that they would choose accepting strategies, as compared to 39.44% who had self- enhancement goals and 37.21% who had self-verification goals. Also, 44.59% of women who 119 had belonging goals indicated that they would use challenging identity management strategies, as compared to 60.56% in self-enhancement goals, and 62.79% in self-verification goals (see Table 12). Table 12: Comparison of accepting and challenging identity management by goals. Accepting IM Challenging IM Goals Self-enhancement Belonging Self-verification None 39.44% 55.41% 37.21% 40.54% 60.56% 44.59% 62.79% 59.46% A chi-square analysis was conducted to test the relationship between goals and identity management strategies chosen. Results show a statistically significant chi-square test of association between self-verification goals and identity management strategies (c2 = 4.43, p =.04). The phi correlation shows that women in self-verification goals were more likely to choose challenging identity management strategies than accepting identity management strategies (f = .35, p = .04). All other relationships between goals and identity management strategies were non-significant. Proximal goals and participants’ choice of specific accepting or challenging IM strategies The results of the overarching model were further explored with an analysis to assess whether or not participants chose differential sub-strategies of identity management in response to the goal conditions they were in. Since the identity management strategies were listed as ordered categorical strategies in this test, the model had to be broken up into two different pieces. That is, the data were analyzed separately for accepting identity management strategies as a mediator, and then for challenging identity management strategies as a mediator due to the issues described earlier. 120 Accepting identity management strategies. In this model, six different accepting identity management strategies were tested as rank-ordered data, namely: (a) ignoring, (b) keeping to myself, (c) avoiding, (d) acting in a manner inconsistent with others’ expectations, (e) bonding with others like me, and (f) using self-deprecating humor. Results show that no significant differences emerged in endorsement of various identity management strategies based on the goal conditions (see Table 13). None of the personality variables predicted identity management strategies with the exception of a small effect where women who had higher self- efficacy to manage sexism tended not to use the identity management strategy of ‘acting in a manner inconsistent with others’ expectations’. None of the mediated paths between the key independent variables, different accepting management strategies, and self-regulatory depletion were significant. Table 13: Sub-strategies of accepting identity management strategies from exploratory model. Conscien- Accepting identity tiousness management strategies Ignoring Self- regulatory depletion Identity centrali ty Self-efficacy Trait anxiety Goals β = .06, p = .55 β = .06, p = .55 β = .00, p = .60 β = -.01, p = .51 β = .03, p = .90 βSE = -.17, p = .11 βSV = -.06, p = .88 βB = -.63, p = .11 βSE = .21, p = .09 βSV = .16, p = .72 βB = .02, p = .96 121 Keeping to myself β = -.05, p = .57 β = -.05, p = .62 β = .00, p = .52 β = -.01, p = .70 β = -.11, p = .60 β = .00, p = .98 β = .36, p = .16 β = .00, p = .94 β = .01, p = .95 β = .10, p = .40 **β = -.01, p = .05 β = -.02, p = .14 β = -.11, p = .60 β = -.00, p = .97 β = -.00, p = .85 β = -.00, p = .76 β = .01, p = .97 β = -.09, p = .42 β = -.00, p = .81 β = .00, p = .99 β = -.03, p = .89 Table 13 (cont’d) Avoiding β = .15, p = .10 β = .04, p = .69 Acting in a manner inconsistent with others’ expectations Bonding with others like me β = .01, p = .94 Using self- deprecating humor β = .03, p = .76 βSE = .02, p = .83 βSV = .21, p = .58 βB = -.13, p = .71 βSE = -.07, p = .55 βSV = -.33, p = .42 βB = -.08, p = .82 βSE = -.10, p = .34 βSV = -.61, p = .17 βB = .05, p = .90 βSE = .03, p = .82 βSV = -.07, p = .87 βB = .28, p = .48 (Note: ** indicates statistical significance at p<0.05; βSE = beta weight for self-enhancement goals, βSV = beta weight for self-verification goals, and βB = beta weight for belonging goals) Goals and choice of accepting identity management strategies. Furthermore, data show that based on the different goals, respondents chose different accepting identity management strategies (see Table 14). While participants have marked all the strategies, I present the top four identity management strategies chosen under each goal condition. 122 Specifically, when respondents were under the press of self-enhancement goals, they chose ‘bonding with others’ as their top identity management strategy. This was followed by ‘ignoring what is said’, followed by ‘trying to act in a manner that is not stereotypical of women’, and finally using ‘self-deprecating humor’. In contrast, when respondents had self- verification goals, the top four accepting identity management strategies endorsed were ‘ignoring what is said’ (i.e., top choice), ‘avoiding my teammates’, ‘trying to act in a manner that is not stereotypical of women’, and using ‘self-deprecating humor’. For belonging goals, respondents indicated preference to use ‘bonding with others’, followed by ‘trying to act in a manner that is not stereotypical of women’, ‘using self-deprecating humor’, and using ‘accepting what is said by teammates’. Table 14: Accepting strategies based on goals. Self- Accepting IM sub-strategies A MANOVA was conducted to assess the relationship between goals and choice of accepting identity management choices. The null hypothesis that the observed covariance matrices of the dependent variable are equal across groups was not supported (p =.003; note for 123 enhancement goal (average (N)) 3.68 (25) Belonging goal (average; N) 4.03 (40) verification Self- goal (average; N) 3.69 (29) No goal (average; N) 3.46 (28) 4.25 (28) 3.61 (41) 4.72 (32) 4.80 (30) 3.78 (23) 3.70 (23) 4.22 (23) 3.51 (41) 3.66 (41) 4.46 (41) 4.72 (29) 4.38 (29) 3.97 (29) 4.80 (28) 4.07 (28) 3.79 (28) 4.46 (24) 4.71 (41) 3.48 (29) 3.89 (28) 3.79 (24) 4.10 (41) 3.57 (30) 4.00 (28) Accepting what is said by my teammates Ignoring what is said by my teammates Keeping to myself Avoiding my teammates Trying to conduct myself in a manner inconsistent with my teammates' expectations Trying to bond with others like me Trying to use self-deprecating humor significance, p-value needs to be lower than p =.005). In addition, the Wilke’s lambda was also examined, and there was no statistically significant difference in choice of accepting identity management strategies based on an individual’s goals, F (21, 308) = 1.04, p < .42; Wilk's Λ = 0.82, partial η2 = .06. Therefore, further analysis was not conducted. Challenging identity management strategies. In this model, six different challenging identity management strategies were tested as rank-ordered data, namely: (a) changing things, (b) working harder to prove them wrong, (c) educating others, (d) trying to communicate inaccuracies in their thinking, (e) making favorable social comparisons, and (f) denying attributions my teammates make of me. Results show that no relationship between goal conditions and endorsement of various challenging identity management strategies (see Table 15). None of the personality variables predicted identity management strategies with the exception of a moderate effect where women with higher trait conscientiousness tended not to endorse the strategy of changing things. None of the mediated paths between the key independent variables, different challenging identity management strategies, and self-regulatory depletion were significant. Self- regulatory depletion Table 15: Challenging identity management strategies from exploratory model. Challenging Self-efficacy Trait identity management strategies Changing things Identity centralit y β = -.10, p = .15 β = -.14, p = .11 β = .00, p = .40 β = -.00, p = .67 anxiety Conscien- tiousness **β = -.38, p = .035 Goals βSE = .14, p = .15 βSV = .04, p = .89 βB = .23, p = .45 124 β = -.01, p = .86 β = -.00, p = .89 β = .01, p = .55 β = .12, p = .54 β = .01, p = .87 β = .01, p = .15 β = .01, p = .54 β = -.14, p = .44 β = .36, p = .16 β = .36, p = .16 β = .36, p = .16 β = .36, p = .16 β = .01, p = .91 β = -.01, p = .34 β = -.01, p = .61 β = -.07, p = .72 β = .10, p = .27 β = -.01, p = .23 β = -.01, p = .57 β = .14, p = .47 Table 15 (cont’d) Working harder to prove them wrong β = .05, p = .46 Educating others β = .04, p = .49 β = .07, p = .24 Trying to communicate inaccuracies in their thinking β = -.01, p = .17 Making favorable social comparisons β = -.03, p = .74 Denying attributions my teammates make about me βSE = .02, p = .83 βSV = .03, p = .93 βB = -.26, p = .45 βSE = .07, p = .42 βSV = .13, p = .62 βB = .37, p = .24 βSE = .14, p = .15 βSV = .04, p = .89 βB = .23, p = .45 βSE = .11, p = .17 βSV = .39, p = .21 βB = -.28, p = .41 βSE = -.05, p = .54 βSV = -.12, p = .71 βB = -.22, p = .52 (Note: ** indicates statistical significance at p<0.05; βSE = beta weight for self-enhancement goals, βSV = beta weight for self-verification goals, and βB = beta weight for belonging goals). 125 Goals and choice of challenging identity management strategies. Similarly, data show that based on the different goals, respondents chose different challenging identity management strategies (see Table 16). As before, while participants have marked all the strategies, I present the top four identity management strategies chosen under each goal condition. Specifically, when respondents were under the press of self-enhancement goals and under the press of self- verification goal, they chose ‘speaking up’ as their top identity management strategy. Participants rated ‘educating others’ as their second choice across both these goals, followed by ‘changing things’, and finally using ‘working harder to prove them wrong’. In contrast, for belonging goals, respondents indicated a preference to use ‘speaking up’, followed by ‘educating others’, ‘changing things’, and communicating ‘inaccuracies in perpetrators’ thinking’. Table 16: Challenging strategies based on goals. Challenging IM sub-strategies Self- enhancement goal (average; (N)) 4.49 (43) Belonging goal (average; (N)) 5.45 (33) Self- verification goal (average; (N)) 5.15 (54) No goal (average; (N)) 5.05 (44) 4.39 (41) 4.50 (32) 4.25 (53) 4.12 (43) 4.22 (41) 3.66 (32) 4.13 (53) 4.40 (43) 4.44 (41) 4.76 (33) 4.43 (53) 4.21 (43) 4.15 (41) 4.34 (32) 4.04 (53) 4.14 (43) 3.29 (42) 2.63 (32) 3.65 (54) 3.49 (43) 3.83 (41) 3.39 (33) 3.79 (53) 3.28 (43) Speaking up Trying to change things Working harder to prove my teammates wrong Trying to educate my colleagues Trying to communicate the inaccuracies in their thinking Trying to make favorable social comparisons that make me look good Denying the attributions my teammates make about me 126 A MANOVA was conducted to assess the relationship between goals and choice of challenging identity management choices. The null hypothesis that the observed covariance matrices of the dependent variable are equal across groups was not supported (p =.002; note for significance, p-value needs to be lower than p =.005). In addition, the Wilke’s lambda shows that there was no statistically significant difference in choice of accepting identity management strategies based on an individual’s goals, F (21, 454) = .94, p < .54; Wilk's Λ = 0.89, partial η2 = .04. Therefore further analysis was not conducted. Open-ended data findings. Participants’ open-ended strategies for identity management behaviors were also coded (N = 142). First, each response was mapped on to theoretically defined identity management sub-strategies for each of accepting and challenging domains. Next, the sub-strategies that mapped on to distinct accepting or challenging domains and distinct sub-strategies within each of the domains were coded as “Accept” (e.g., ignoring) and “Challenge” (e.g., speaking up). 26 open-ended responses emerged for accept pure category, and 54 open-ended responses emerged for challenge pure category. Next, instances where a respondent had selected two different sub-strategies from the same overarching category were coded as “Accept complex” (e.g., ignoring and bonding with others) and “Challenge complex” (e.g., educating others and speaking up). In total, four respondents shared strategies that mapped on to ‘accept complex’ and seven respondents shared strategies that mapped on to ‘challenge complex’ (See Table 17 and Table 18 for examples). 127 Table 17: Accepting identity management strategies. Total N (N/142) 30 (21%) Accept IM sub-strategies and frequency of endorsement • Accepting • Avoiding • Behaving in ways that are inconsistent with stereotypes of being a woman • Bonding • Ignoring • Keeping to myself • Accept complex Exemplars from the open-ended responses “Find another team” (avoiding) “Try to find common ground” (bonding) “Honestly if I can't work with them to improve their mindsets, at that point I would stop trying and just tag along with them and let them take the lead and the credit and try and not be at fault for their flawed thinking” (ignoring) “I would more than likely keep to myself and avoid my teammates” (keeping to myself and avoid; ‘accept complex’) “I might just have to go along with their way of thinking and keep my mouth shut” (Ignoring and accepting: ‘accept complex’) 128 Exemplars from the open-ended responses “I would speak up for myself” (speaking up) “I would need to change their entire attitudes” (changing things) “Be honest and upfront about what is acceptable in teams” (communicating) “I would work on the assignment to the best of my ability and in doing so prove their assumptions about all women wrong” (working harder to prove them wrong) “Well I think it's important to speak up and let them know their behavior towards women is unacceptable and explain that it is damaging behavior for the whole team because if women are being looked down upon they can not preform at their best” (challenge complex: speaking up and educating) “I would speak up when needed and show them what a good job I am able to do. I would let them see how talented and educated I am by giving good ideas” (Challenge complex: speaking up and working hard to prove them wrong) Table 18: Challenging identity management strategies. Total N (N/142) 60 (42%) Challenge IM sub-strategies and frequency of endorsement • Speaking up • Changing things • Communicating • Educating • Working harder to prove them wrong • Challenge complex I also coded the responses for new identity management strategies that have not been explicitly addressed in the literature, or strategies where respondents offered ways to connect the different identity management strategies in interesting ways (N = 43). In cases where strategies have not been explicitly addressed in the literature, I mapped them to the existing strategies by connecting them to Meisenbach’s (2010) typology (See Table 19). 129 Table 19: Coding open-ended data for new identity management strategies. Total N (N/43) Exemplars from the open- ended responses New IM strategies and frequency of endorsement Focusing on the task 15 (34.90%) Mapping to IM strategies in the literature Accepting strategies Rationale for mapping These were coded as accepting identity management strategies given that these respondents show a willingness to simply avoid the stigma communication by hoping that people judge them on their work so they do not have to address the stigma communication. Information seeking was coded as a form of accepting strategies because of the willingness showcased by respondents to accept their team’s sexist feedback as meaningful enough to act on and make requisite changes to how they navigate work. • “Come up with a plan or splitting up the work” • “I’ll use professional, dispassionate, unassailable logic and excellent quant skills to show those guys that they’re wrong. And if they try to say I’m an exception (not like other women), I’ll jokingly say that I feel sorry for them for having such limited experience with women, then suggest that we get some work done” • “Working together to achieve and accomplish one goal regardless of our preferences” • “I’d try to work the best I could with my team to try and get the job done well and accurately” • “Ask the virtual team for more information” • “I would take into consideration what they said and try to adjust my behavior to their standards” • “Listen and question what they are trying to accomplish” • “Ask questions” • “Be open minded” • “Listening” 8 (18.60%) Information seeking Accepting strategy 130 Neither accepting nor challenging Challenging strategy Emotional regulation is not traditionally included as an identity management strategy, and instead is treated as a coping behavior. These were coded as challenging because the respondent showcase a willingness to change something about the situation by putting in place norms of what it would mean to work with them and/or work as a cohesive team Table 19 (cont’d) 5 (11.62%) Emotional regulation • “I would also be very neutral emotionally so that the team does not think I am being over reactive, or hyper sensitive” • “First I would get mad” • “Deep breaths and positive thoughts” 3 (.07%) Norm setting • “I would let them know • “Pray that the lord move mountains in my work place” • “You don’t want to know what I’d do” that they would lose me as a teammate with valuable insight if they continued to act this way” • Would try another approach like using examples of other teams or individuals and how they made it work 131 Table 19 (cont’d) 3 (.07%) Showcasing past performance (proven track-record) 2 (.05%) Requesting training • “Use my education and work history and my proven track record of strengths to show them they were wrong” • “Be assertive and explain my proven track record of success in my career” • “In this instance I would relish the opportunity to prove them wrong. Math was my strong suit and was offered a full scholarship to engineering college in the 60's when women were not even seriously considered for engineering. I'd let my work do the talking and confident I would succeed” • “Ask for more training” • “When people have those attitudes and mind sets it’s very difficult to get a group to change their thinking. They need to go thru [sic] training to make them see what their [sic] doing and how destructive their behavior and thought process is” Challenging strategies Challenging strategies This strategy most closely aligns with logical denials where people use rhetorical or behavioral strategies to showcase how stigma does not apply to them (cf., Meisenbach, 2010). In the current study, respondents described relying on their past performance to navigate the sexist situation at hand This strategy speaks to respondent’s attempts to educate others and thereby change their environment. While Meisenbach’s (2010) typology lacks educating others as a sub- strategy, other identity management frameworks include it (cf., Corrigan and Penn, 1999). Thus given the active nature of behaviors, this was coded as a challenging strategy. 132 Table 19 (cont’d) 2 (.05%) Ingratiating • “I would be complimentary” • “Deferring to other team members, let them take the lead, try to smooth out any hurt feelings between members” Accepting strategies While ingratiating behaviors are a form of impression management tactics (cf., Leary and Kowalski, 1990), I coded these as a form of accepting strategies, because respondents showcase a desire to defer to their teammates despite their stigmatizing behaviors. Some strategies could not be mapped to either accepting or challenging identity management strategies. For example, emotional regulation is typically treated as a coping strategy, but is not referenced as an identity management strategy. However, the manner in which an individual enacts identity management strategies – that is layering on an emotional regulation lens on accepting or challenging behaviors in response to stigma situations – may in fact change the manner in which the identity management strategies are likely perceived by others. Supplementary analyses: Treating ordinal data as continuous data A final model was tested where identity management strategies were treated as continuous variables since both accepting and challenging identity management strategies were rated on a scale of ‘most likely to use’ (7) to ‘least likely to use’ (1). These results are tabulated below (see Table 20 and Table 21). Results show that under conditions of self-enhancement goals women tended to use significantly less accepting identity management strategies. Apart from this relationship, no 133 significant differences emerged in endorsement of various identity management strategies based on the goal conditions (see Table 20). Some personality variables predicted the choice of identity management strategies. Specifically, self-efficacy of coping with sexism emerged as a statistically significant predictor of endorsing ‘acting in a manner inconsistent with others’ expectations’ such that higher coping self-efficacy was positively related to participants’ endorsement of this identity management strategy (although this relationship is small). Conscientiousness and trait anxiety did not predict any of the identity management choices. As before, none of the mediated paths between the key independent variables, different accepting management strategies, and self-regulatory depletion were significant. Self- regulatory depletion Table 20: Accepting identity management strategies treated as continuous variables. Accepting identity management strategies (as continuous data) Accepting Identity centrality Self- efficacy Trait anxiety β = -.16, p = .17 β = -.18, p = .08 β = .00, p = .94 β = .01, p = .46 Conscien- tiousness β = .25, p = .21 Goals **βSE = -.36, p = .00 βSV = -.28, p = .48 βB = -.12, p = .76 βSE = -.17, p = .10 βSV = -.03, p = .94 βB = -.62, p = .10 134 Ignoring β = .00, p = .97 β = .04, p = .67 β = .00, p = .66 -β = .01, p = .45 β = .04, p = .83 β = -.05, p = .61 β = .00, p = .54 β = -.01, p = .55 β = -.12, p = .53 β = .06, p = .61 β = -.00, p = .92 β = -.00, p = .75 β = -.04, p = .86 β = .10, p = .32 **β = .01, p = .05 β = -.02, p = .17 β = -.09, p = .64 β = .01, p = .91 β = .00, p = .97 β = -.00, p = .83 β = .01, p = .98 β = -.08, p = .39 β = -.00, p = .85 β = .00, p = .85 β = -.02, p = .89 Table 20 (cont’d) Keeping to myself β = -.08, p = .51 Avoiding β = .11, p = .36 β = -.01, p = .96 Acting in a manner inconsistent with others’ expectations Bonding with others like me β = .04, p = .77 Using self- deprecating humor β = -.01, p = .97 βSE = .19, p = .06 βSV = .14, p = .70 βB = -.02, p = .96 βSE = .00, p = .99 βSV = .14, p = .71 βB = -.23, p = .52 βSE = -.05, p = .61 βSV = -.29, p = .44 βB = -.02, p = .95 βSE = -.04, p = .62 βSV = -.37, p = .35 βB = .29, p = .44 βSE = .02, p = .82 βSV = -.03, p = .93 βB = .29, p = .39 (Note, ** indicates statistical significance with p<.05; βSE = beta weight for self-enhancement goals, βSV = beta weight for self-verification goals, and βB = beta weight for belonging goals). 135 For challenging identity management strategies, results show that there was no effect of being in different goal conditions on choosing identity management responses (see Table 21). Some personality variables predicted the choice of challenging identity management strategies. Specifically, Conscientiousness was significantly and negatively related to the use of ‘changing things’ identity management strategy. As before, none of the mediated paths between the key independent variables, different accepting management strategies, and self-regulatory depletion were significant. Conscien- tiousness Goals Self- regulatory depletion Table 21: Challenging identity management strategies treated as continuous variables. Challenging identity management strategies (as continuous data) Speaking up Identity centrality Self- efficacy Trait anxiety β = -.07, p = .42 β = -.07, p = .47 β = .00, p = .82 β = -.02, p = .14 β = -.17, p = .41 βSE = .02, p = .70 βSV = .37, p = .22 βB = .46, p = .15 βSE = .12, p = .12 βSV = .02, p = .92 βB = .20, p = .43 βSE = -.02, p = .80 βSV = -.02, p = .95 βB = -.22, p = .45 Changing things β = -.10, p = .18 β = .03, p = .70 Working harder to prove them wrong *β = -.13, p = .09 β = .00, p = .39 β = -.00, p = .71 **β = -.33, p = .03 β = -.02, p = .80 β = -.00, p = .88 β = .01, p = .61 β = .11, p = .53 136 β = .01, p = .89 β = .01, p = .098 β = .01, p = .40 β = -.13, p = .43 β = .06, p = .37 β = .01, p = .13 β = .01, p = .57 β = .01, p = .92 β = .00, p = .99 β = -.01, p = .24 β = -.01, p = .45 β = -.07, p = .71 β = .09, p = .31 β = -.01, p = .16 β = -.01, p = .45 β = .15, p = .44 Table 21 (cont’d) Educating others β = .02, p = .82 β = .06, p = .43 Communicati ng inaccuracies in their thinking β = -.11, p = .17 Making favorable social comparisons Denying attributions β = -.04, p = .54 βSE = .05, p = .54 βSV = .11, p = .67 βB = .32, p = .27 βSE = .01, p = .93 βSV = -.09, p = .69 βB = .14, p = .56 βSE = .13, p = .095 βSV = .45, p = .12 βB = -.17, p = .56 βSE = -.04, p = .63 βSV = -.09, p = .76 βB = -.18, p = .60 (Note: βSE = beta weight for self-enhancement goals, βSV = beta weight for self-verification goals, and βB = beta weight for belonging goals). 137 Discussion One of the key questions that emerged from the literature review presented in Chapter One, and the basis of this study, was if individuals are likely to choose different identity management strategies based on their goals in the workplace, and whether or not these choices would be implicated in predicting self-regulatory depletion differently. The specific goals of this study were to explore: (a) if the different identity management strategies may in turn be related to differential self-regulatory depletion, (b) how individual difference variables of self-efficacy and identity centrality impact the choice of strategies and resultant self-regulatory depletion, (c) if under different goal conditions of self-enhancement, self-verification, and belonging goals, people select different identity management strategies, and (d) themes that emerged from the open-ended data provided by participants by coding for the different identity management choices used to navigate sexist situations. Thus, individual difference variables such as identity centrality and self-efficacy were included as important antecedents of choice of the identity management strategies, which in turn were tested as mediators between the individual factors and self-regulatory depletion reported by respondents. Similarly, self-enhancement, self-verification, and belonging goals were included as important situational antecedents to the accepting and challenging identity management strategies. Unfortunately, some of the design decisions in this study – that is, model specification problems due to different variable distributions and resultant power issues, combined with the failure of the performance discussion manipulation to yield enough variance on self-regulatory depletion (the dependent variable) under different goal conditions led to non-significant findings for the hypothesized model. As such, in the discussion below, I discuss findings by aggregating across both the hypothesized and alternative models that were tested. Where applicable, I also 138 discuss key findings from the exploratory analyses and include ideas extracted from the open- ended data that are central to the questions under investigation. Identity management strategies and self-regulatory depletion Neither accepting nor challenging identity management strategies predicted self- regulatory depletion in the current study. This was true across all the models tested in the current study. To better contextualize this finding, a broader discussion will be presented later in the General Discussion that explains relationships between identity management strategies and self- regulatory depletion by aggregating across the three studies (see Chapter Seven). Role of individual differences in navigating stigma communication Two key individual difference variables of identity centrality and coping self-efficacy were included in this study and a discussion of how these are related to self-regulatory depletion under stigma communication and choice of identity management strategies is presented below. Also presented below is a discussion on the role of control variables, trait anxiety and conscientiousness. Individual differences and self-regulatory depletion. As in Study One, no relationship was found between identity centrality and self-regulatory depletion. As in Study One, coping self-efficacy was implicated in self-regulatory depletion such that women who had higher levels of coping self-efficacy, tended to be less depleted in a sexist event. This finding is consistent with the literature; specifically, individuals report less stress and anxiety if they have higher perceived coping self-efficacy (cf., Chesney et al., 2006). However, the largest effect sizes were obtained for the control variables. Specifically, while women who reported high levels of trait anxiety tended to report more self-regulatory depletion, women who reported high 139 conscientiousness, tended to report less self-regulatory depletion. Trait anxiety has been associated with higher focus on threat and higher stress reactions to threatening stimuli, and it is clear why women high in trait anxiety may process sexist events as threatening stimuli (cf., Mathews and MacLeod, 1985) and therefore feel more depleted (Inzlicht, McKay, and Aronson, 2006). Studies show how trait anxiety is related to higher amygdala reactivity to stress-related cues (Fakra, Hyde, Gorka, Fisher, Munoz, Kimak, and Hariri, 2009). Trait conscientiousness resulted in the positive association with self-regulatory depletion such that at higher levels of trait conscientiousness, respondents’ endorsed being less depleted (higher scores are indicative of less depletion). This is also consistent with previous research that finds that individuals with high conscientiousness tend to persist longer when faced with self-regulatory fatigue (Nes, Carlson, Crofford, Leeuw, and Segerstrom, 2011). Identity centrality and identity management strategies. As in Study One, the relationship between identity centrality and choice of identity management strategies were not statistically significant. When a rank-ordered model for accepting identity management strategies was specified, it was found that none of the tested relationships were statistically significant. However, when the data were treated as continuous, higher identity centrality was negatively related to accepting strategies (marginal significance); no other statistically significant effects emerged. When a rank-ordered model for challenging identity management strategies was specified, it was found that none of the tested relationships were statistically significant. It was expected that women with high identity centrality would seek to use challenging identity management strategies more because they treat their identity as a source of pride and self-esteem and would like to take actions that proactively protect their identities (e.g., Bombay, 140 Matheson, Anisman, 2013). In line with literature, I had expected that high identity centrality would also translate to concerns about how one’s in-group is perceived, and therefore translate into endorsing strategies such as engaging in attempts to ‘change things’, and ‘working harder to prove others wrong’ about their stereotypes of one’s identity group. However this was not the case. Perhaps, there is a moderating role of one’s own stigma beliefs about one’s identity group. For example, if a woman does not believe that her identity is stigmatized, then it would make sense for her to state that there is nothing there to change or nothing to prove in the first place. This is just a speculation and would make an interesting research question to pursue. Literature on group identification and perceived discrimination (cf., Crocker and Major, 1989; Sellers and Shelton, 2003) would suggest that it is also likely that high identity centrality women are more vigilant and/or more knowledgeable about the power structures that are in place in organizations, and might assume a fatalistic view such as ‘why bother, nothing I do would change things anyway’. This may explain why women with high identity centrality also showcased a willingness to engage in accepting strategies such as ignoring and avoiding sexism when anticipating working with the sexist virtual team. Given these findings, on the whole, identity centrality did not emerge as a predictor across both the studies. A broader discussion of these non-significant findings is presented later in the General Discussion (see Chapter Seven). Coping Self-efficacy and identity management strategies. While in Study One, higher coping self-efficacy emerged as a statistically significant predictor of using more challenging identity management strategies, this was not the case in the current study. When the rank-ordered model for accepting identity management strategies is tested 141 however, self-efficacy emerged as a statistically significant predictor of one of the identity management choices (although the effect size was small) – specifically higher levels of coping self-efficacy were negatively related to ‘acting in a manner inconsistent with others’ expectations’. Perhaps it is the case that women who have higher coping self-efficacy to navigate sexism, view attempts at positive distinctiveness strategies as shining too bright a light on their stigmatized identity, and therefore choose not to rock the status quo by acting in ways that are inconsistent with people’s expectations at work. That said, in treating the ordinal strategies as continous data, the direction of this finding flipped such that higher coping self-efficacy was related to higher use of ‘acting in a manner inconsistent with others’ expectations’ (note this finding was also statistically significant, although the effect size was small). While the ordinal data model is the better test of this effect given alignment between the response scale (i.e., ordinal distribution of responses) and model specification, this finding should be cautiously interpreted. While it would make intuitive sense that women who have the confidence to cope with sexist situations would likely not prefer to use self-deprecating humor and bonding with other stigmatized individuals because both these strategies may illuminate their own stigmatized identity. Seeking the company of others who have been ‘Othered’ can offer social support (Ashforth and Kreiner, 1999), but if one’s ability to ride out the negative effects of sexism is strong, then traveling with ‘birds of a feather’ can also be perceived as compounding the risk of stigmatization (akin to stigma by association; Kulik, Bainbridge, and Cregan, 2008). Similarly, using self-deprecating humor as a strategy may also have the potential to draw unnecessary attention on the stigmatized part of one’s identity (Holmes, 2000; Holmes and Schnurr, 2005). However, none of the accepting identity management choices were related to coping self- 142 efficacy in the current study. Women who are confident about navigating sexism, may feel less press to protect their own ego needs by putting down other identity groups by using strategies such as ‘making favorable social comparisons’. Self-efficacious individuals were also expected to be less likely to be distressed and to have higher perceived control of negative situations (Sumer, Karanci, Berument, and Gunes, 2005) – thus high self-efficacious women could be expected to take the sexist events and comments in the workplace on the chin and persist with their tasks at hand by using strategies such as ‘denying attributions others make about me’. However, when the rank- ordered model for challenging identity management strategies was tested, coping self-efficacy did not predict any of the specific challenging identity management strategies. One key insight to emerge from this discussion is that clearly there is significant variance in the choice of identity management strategies, and the question of why women may select different identity management strategies remains an important one. While this study was designed as a test of some of the antecedents that were seen as theoretically important in elucidating why women chose one strategy over the other, this question still remains open. In the current theorizing, identity centrality and coping self-efficacy were thought to be two important predictors of these choices, but it turns out that the control variable of conscientiousness had better explanatory power than the focal constructs of interest. In the hypothesized model, it was found that conscientiousness was negatively related to choosing challenging identity management strategies (moderate effect). Furthermore, in the rank-ordered data model for challenging strategies, ‘changing things’ identity management strategy was significantly and negatively predicted by trait conscientiousness (moderate effect). 143 Research on employee voice shows that conscientiousness predicts voice behaviors, especially when voice behaviors are directed towards one’s supervisors (e.g., Nikolau, Vakola, and Bourantas, 2008). However, in the current study, high conscientiousness seemed to hold women back from challenging sexism at work; this finding is consistent with a recent multi- source, time-lagged study that investigated how conscientiousness might buffer the impact of stressors at work (Lin, Ma, Wang, and Wang, 2015). It was found that high conscientiousness exacerbated employees’ psychological strain to stressors at work (Lin, Ma, Wang, and Wang, 2015). Given the meta-analytic findings on the efficacy of conscientiousness as a predictor of performance, organizations actively select for this trait (Judge and Ilies, 2002). To the extent a work environment is sexist, women would be at an apparent disadvantage in challenging their colleagues and/or supervisors on their sexist behaviors. Given these findings and the many speculative claims presented above, perhaps future research should build answers to the question of why women choose one identity management strategy over the other from a grounded theory perspective. Specifically, I posit that the focus needs to be on evaluating the psychological perceptions – both cognitive and affective – of each of the identity management strategies. When faced with having to navigate sexism, how do women view these identity management strategies, and could we isolate the reasons for why women endorse one strategy over the other? Role of Goals in choice of self-regulatory depletion and identity management strategies Goals are an important attribute in a person’s decision matrix of how to navigate their day-to-day organizational lives (cf., Barrick, Mount, and Li, 2013; Harris, Daniels, and Briner, 2003; Maier and Brunstein, 2001). Though I had expected that proximal goals would influence 144 the choice of identity management strategies and resultant self-regulatory depletion when women confronted sexist events, this was not supported in the current study. Main effects of self- enhancement, self-verification, and belonging goals on self-regulatory depletion were non- significant, and so were the mediated effects via identity management strategies. I highlight some design issues in the limitations section that shed light on why this may have been the case. It was also predicted that goals would be differentially related to endorsement of identity management strategies, and while a differential pattern of identity management choices was observed, the findings were non-significant in the current study. In the hypothesized model, participants in self-enhancement and self-verification goals tended to report using more challenging identity management strategies as opposed to accepting identity management strategies. However, those in belonging goal conditions tended to choose accepting identity management strategies at a much higher rate. Specific accepting identity management strategies and goals. When the rank-ordered model with specific accepting identity management strategies was tested, under self- enhancement goals women tended to endorse ‘keeping to myself’ identity management strategy (statistically significant positive relationship). With the exception of this finding, under self- enhancement and self-verification goals, women showcased a similar pattern of endorsing different accepting strategies (all other relationships were statistically non-significant). Specifically, under both goals women tended to use more of (a) ‘keeping to myself’, (b) ‘avoiding’, (c) ‘using self-deprecating humor’; and less of (a) ‘ignoring’, (b) ‘acting in a manner inconsistent with others’ expectations’, and (c) ‘bonding with others like me’. One would think that when asserting one’s identity, the use of self-deprecating humor may not be a realistic choice – however, it is likely that this strategy is perceived as a ‘non-confrontational’ way 145 (Holmes, 2000; Holmes and Schnurr, 2005) by which women can comment on their identity and bring it into focus. It remains unclear why women in self-enhancement goal would also choose self-deprecating humor – one explanation may be that when women play up their performance at work, or act in agentic ways, they are at a higher risk of backlash (Phelan, Corinne, & Rudman, 2008; Rudman & Glick, 2001; Ridgeway & Correll, 2004). Perhaps women use self-deprecating humor as a strategy to pre-emptively thwart the perception that they may be seen as too agentic which they know from experience can quickly detract from their self-enhancement goals? However, research also shows that the prescriptions of role congruity theory and the backlash effect may apply only to women who are in lower to middle levels of the organizational structures, whereas top women leaders who are seen as more agentic and more communal may have an advantage over male leaders (Rosette & Tost, 2010). In contrast, women in belonging goals tended to use less of (a) ‘ignoring teammates’, (b) ‘avoiding teammates’, and (c) ‘acting in a manner inconsistent with teammates’ expectations’; and more of (d) keeping to myself, (e) bonding with others like me, and (f) using self- deprecating humor. While these relationships are not statistically significant, they intuitively make sense about how the press of belonging goals would manifest itself in behavior. Those with belonging goals would in fact find it difficult to ignore and avoid people who are likely to fulfill their need to belong (Baumeister and Leary, 1995). Similarly, women could worry that by ‘acting inconsistent with teammates’ expectations’ they may be seen as rocking the proverbial boat and hence upsetting the status quo at work, which is something individuals with a goal to belong would likely avoid. This explanation is consistent with research that shows that non- prototypical group members in teams tend to adhere to group norms when they have a high need to belong (Steinel, Van Kleef, Van Knippenberg, Hogg, Homan, and Moffitt, 2010). Finally, it 146 would make sense for women to use ‘bonding with others’ and also ‘self-deprecating humor’ to gain empathy and understanding from peers. Specific challenging identity management strategies and goals. When the rank-ordered model with specific challenging identity management strategies was tested, as before, under self- enhancement and self-verification goals, women showcased a similar (but statistically non- significant) pattern of endorsing different challenging strategies. Specifically, women selected more of (a) ‘changing things’, (b) ‘working harder to prove them wrong’, (c) ‘educating others’, (d) ‘trying to communicate inaccuracies in teammates’ thinking’, and (e) ‘making favorable social comparisons’; and less of (a) ‘denying attributions my teammates make about me’. In contrast, for belonging goals women endorsed using more of (a) ‘changing things’, (b) ‘educating others’, and (c) ‘trying to communicate inaccuracies in their thinking’; and less of (a) ‘working harder to prove them wrong’, (b) ‘making favorable social comparisons’, and (c) ‘denying attributions my teammates make about me’. While the choice of most challenging strategies here make sense given the goals, it is unclear why women in belonging goal condition would endorse less of ‘working harder to prove them wrong’. Perhaps the use of “them” in the item made the respondents feel that they would be casting their coworkers as separate from self rather than as a group. Insights from descriptive data. Overall, data show that given different goals, respondents do choose different identity management strategies. While some strategies emerged as top contenders, there were others that were not endorsed as often. For example, ‘keeping to myself’ accepting identity management strategy was not utilized as often as the others. Similarly, from challenging identity management strategies, respondents tended to not endorse ‘making favorable social comparisons’ and ‘denying the attributions made by teammates’ as often as they 147 endorsed the other strategies. The language for both of these is not user-friendly, and in hindsight, it is likely that respondents did not endorse these items simply because they did not understand them well enough. While hypotheses were not explicitly crafted to include which specific strategies would be endorsed under different goals, I did present preliminary expectations (see Footnote 18, p. 105). Under self-verification goals, I had expected women to ‘bond with stigmatized others’ (Accepting stigma) and ‘make favorable social comparisons’, ‘logical denials’, and ‘educate others’ (challenging stigma). It was found that while respondents did use ‘bonding with stigmatized others’ as a strategy, ‘ignoring what is said’ emerged as the top ranked strategy. It is interesting that even under self-verification goals women endorsed ‘ignoring’ as a way to assert their identity. This is counterintuitive to the assigned goal, but I offer two preliminary explanations. First, is it the case that women see ignoring as the least depleting choice? Perhaps by ignoring they feel they are conserving their resources and will not be as depleted as they would be if they were to take action. Alternatively, could it be that in work contexts, it may be difficult for those confronting stigma about a given identity to then assert that same identity? Could this difficulty translate into forms of disengagement behavior such as ignoring sexism? DeJordy (2008) have theorized this difficulty for individuals with invisible stigmatized identities, whereby lack of self-verification can precipitate disengagement from social context. Overall, while we could argue that ignoring stigma communication is not an effective strategy, women may see this strategy as the most “efficient” even if it is not effective in eradicating stigma. Future empirical research should investigate if this holds for individuals with visible identities as well. 148 Under self-enhancement goals, I had expected women to recategorize (accepting stigma), and ‘make favorable social comparisons’ and ‘simple denials’ (challenging stigma). Here again, ‘ignoring sexism’ emerged as the top choice for accepting identity management strategies, and ‘speaking up’ emerged as the top challenge identity management strategy. Finally, under belonging goals I had expected women to use ‘passive acceptance’, ‘self-deprecating humor’, and ‘bonding with stigmatized others’ strategies (accepting stigma) and no challenging stigma strategies. Women did use ‘passive acceptance’, ‘bonding with others’, and ‘self-deprecating humor’ from the accepting identity management choices, but they also used ‘speaking up’ as a challenge identity management strategy in the belonging goal. It is interesting how women in belonging goal condition also endorsed ‘speaking up’ as a strategy – while theoretically it does not seem to make much sense, the open-ended data provided by participants, does shed light on preliminary explanations. Specifically, ‘speaking up’ does not always have to be about challenging perpetrators; instead women mentioned using communication with their teammates to set the right norms, to glean more information from them, and to engage in more sense making (see open-ended data discussed later). This version of non-threatening ‘speaking up’ could very well fit with the belonging goals. Insights from open-ended data. Women in the current study also shared open-ended data on the different types of identity management strategies they would have used in dealing with the sexist team at work. While 21% women reported wanting to use accepting identity management strategies, 42% women reported wanting to use challenging identity management strategies to navigate the sexist events at work. Participants also provided six new strategies that have so far not been explicitly included in identity management frameworks, but that can be mapped on Meisenbach’s (2010) framework 149 (with the exception of emotional regulation). Focusing on the task at hand emerged as one way to navigate sexist events at work. For example, one respondent stated that, “I would do my work and not worry about no one else I can be a team player but I'm not kissing no one's ASS [sic]”. Another eight respondents spoke about using information seeking behaviors to navigate sexist events at work. For example, “I would take their criticism as a chance to improve myself” and “I don’t believe I would ignore them, however, challenging what they say wouldn’t help me meet my goal. I might ask them to explain or elaborate on how it is they feel”. A few women referenced the use of emotional regulation as a possible strategy to cope with sexism, while others spoke of norm setting. For example, “I would value each other's opinion and not have criteria that is unprofessional”. Women also indicated that they would rely on their past performance at work to show how the stereotypical stuff does not apply to them, and two women in the sample also referenced seeking policy support from the organization, example by asking HR to engage in training. These strategies can be mapped to the positive distinctiveness categories (not included in Meisenbach, 2010). Limitations The performance discussion manipulation in the current study was designed to ensure that participants saw the team they would work with as sexist. In line with research, it was expected that reading about a sexist performance discussion (i.e., ambient harassment), women would feel depleted due to the stigma communication. While the prompt worked as intended such that participants rated the team as sexist, and indicated that they would be concerned about working with the virtual team, this concern did not translate into variance on self-regulatory depletion as it related to the different goal conditions. It may be the case that the depletion that was reported was not connected to the goal in an artificial task, but the sexism story in the 150 performance discussion task could still have primed individuals to recall feelings and/or experiences that were depleting, creating the variability in depletion. It is clear that the performance discussion manipulation clearly had the intended effect because participants rated the team as sexist, and even though the consent form clearly stated that the study was only 20 minutes long, participants still rated that they would not like to work with this virtual team. Where the synthetic nature of the task could have attenuated the effects is in participants’ “buying-into” the assigned goals. Although a manipulation check was used to establish whether or not the participants remembered the goal condition they were in, this is not the same as having committed to the goal. Typical goal setting studies use some form of cash incentives to create genuine goal commitment for respondents (Bonner, Hastie, Sprinkle, & Young, 2000), and in this study, no cash prizes were offered to help respondents feel committed to their goals. So while people were aware of the goal, they may not have showcased enough commitment. Also, I had not tested for these specific relationships in the Pilot Study for concerns about testing the overall methodology, but given the novelty of the task, in hindsight, the experimental paradigm should have been piloted more extensively. There were significant issues created due to the different types of scale of variables used in this study. Accepting and challenging identity management strategies were treated as categorical, and then as a next step, participants were instructed to rank-order each of the strategies on a scale of ‘most likely to use’ to ‘least likely to use’. A combination of these two choices created a particularly onerous design such that three different distributions were represented across the same design: (a) binary distribution for the accepting and challenging variables, (b) ordinal distribution for the rank-ordered identity management strategies, and (c) continuous distribution for the dependent variable of self-regulatory depletion. Specifying all 151 three distributions in one SEM model proved to be a challenge and these three distributions could not be accommodated in one model. To manage these issues, first an overarching model was tested where the identity management strategies were treated as binary categorical variables. While the direction of effects was consistent with the first study on hypotheses that were being replicated, this model returned non-significant findings. The new hypotheses that were being tested either returned non- significant or null findings. One possible reason for the non-significant findings could be that participants were making a decision to either accept or challenge without really having read the definitions of what these bins meant (i.e., it is likely that they were guided by the words “accepting” and “challenging”). Once they were in each of the bins, they did not have an option to go back and change their decisions if they confronted strategies that did not align with what they might have thought about the meaning of “accepting” or “challenging”. This is just a speculation though, as I do not have the means to test this idea with data. To better understand the nature of the data, a nuanced model with the individual identity management strategies was run – however, because participants had rank-ordered these data, I had to specify an ordinal regression model and the sample had to be split into two halves: one for accepting identity management, and another for challenging identity management. While the mean differences showcase possible relationships, this model was underpowered and unable to detect effects. Finally, since the identity management strategies were rank-ordered from ‘most likely to use’ to ‘least likely to use’, I tested a final model where identity management strategies were specified as continuous variables. There were some findings obtained for this model, but once again, since the data had to be split into accepting and challenging, this model was also 152 underpowered. Given these issues, the findings of this study should be interpreted as preliminary, and a better test of these ideas is explained in the ecologically valid Study Three presented next. 153 CHAPTER 6: STUDY THREE This study is an attempt to answer if stigma communication impacts other key outcomes (e.g., self-regulatory depletion, psychosomatic health indices, and strain) depending on differential identity management strategies used. Evidence shared in chapter one sheds light on how accepting identity management strategies may be related to lower strain, lower psychosomatic complaints, and lower self-regulatory depletion. In contrast, challenging identity management strategies may be related to higher strains, higher psychosomatic complaints, and higher self-regulatory depletion. The current field study builds on the previous laboratory cross-sectional studies by testing these relationships using an experience sampling method (ESM). The ESM paradigm is useful here because (a) it allows for an exploration of the within-person, day-to-day variation in phenomena such as self-regulatory depletion and identity management, taking into account the role of goals that also may vary with time, (b) avoids issues of retrospective recall and memory lapses in reporting about the stigma communication and the resultant identity management strategies used and/or outcomes of interest, (c) allows for episodic or experiential reports over the more semantic or conceptual reports, and (d) allows for scholars to study a person-in-context without constraining expression of behavior using synthetic laboratory paradigms that can be contaminated by demand characteristics and/or may suffer from problems of ecological validity (cf., Christensen, Barrett, Bliss-Moreau, Lebo, & Kaschub, 2003; Fisher & To, 2012). Given that the interest here is to study the impact of stigma communication events on the resultant identity management strategies and the associated outcomes, event-based ESM is a 154 natural fit for scheduling data collection instead of using signal27 based ESM or interval28 based ESM. Briefly, event-based ESM is used when the data collection is organized around a specific event’s occurrence and when there are concerns that collecting the data at a later time may lead to memory lapses. For example, event-based ESM has been used to study predictors of identity management strategies in a sample of lesbian, gay, and bisexual workers (cf., King, Mohr, Peddie, Jones, & Kendra, 2017). In this study, I am interested in examining sexist events that may communicate stigma to women. The conceptual model to be tested in this study is presented below (see Figure 14). Figure 14: Conceptual model for Study Three. (Note: All study variables are Level 1 variables, i.e., within-person variables29). 27 Signal based ESM is used when the researchers are interested in gathering data about experiences in real time. As the nomenclature suggests, researchers send out a signal during the course of a day that instructs individuals to respond to the survey(s) and report on their experiences at that given moment of time. Typically signal based ESM studies are used to study “states that may be susceptible to retrospective memory bias” and “states that are susceptible to cognitive or emotional regulation” (Christensen et al., 2003, p. 60). For example, Csikszentmihalyi & Hunter (2003) used a signal based ESM study to study the environmental factors and individual behaviors that are associated with happiness. 28 Interval based ESM studies are used when scholars know beforehand when the data can and should be collected and are thus able to pre-specify intervals (e.g., morning, noon, evening, night) at which participants should take the survey(s). For example, Ilies and Judge (2002) used an interval based ESM study to explore the dynamic relationship between mood, satisfaction, and personality and administered surveys by programming an electronic interface that accepted “only one set of responses during specified time intervals” (p. 1125). 29 Research has shown that self-regulation varies within-person (cf., Hofmann, Baumeister, Förster, Vohs, 2012) and so do the well-being variables of psychosomatic health indices and strains (cf., Repetti, 1993). 155 Table 22: Hypotheses for Study Three. H# H8 Within-individuals, role hierarchy would impact an individual’s decision to accept or challenge stigma such that when an individual receives stigma communication from those (a) lower in the role hierarchy she would be more likely to challenge stigma but when she receives stigma communication from those (b) higher in the role hierarchy she would be more likely to accept stigma. Hypothesis H10 Within-individuals, (a) accepting stigma would partially mediate the relationship between role hierarchy and self-regulatory depletion, and (b) challenging stigma would mediate the relationship between role hierarchy and self-regulatory depletion. H15 Within-individuals, the proximal goal during a stigma communication event would be differentially related to (a) accepting and (b) challenging strategies. H16 Within-individuals, (a) accepting stigma would mediate the relationship between proximal goal and self-regulatory depletion, and (b) challenging stigma would mediate the relationship between proximal goal and self-regulatory depletion. H17 Within-individuals, accepting identity management strategies would be related to (a) low strain, (b) low psychosomatic complaints, and (c) low self-regulatory depletion. H18 Within-individuals, challenging identity management strategies would be related to (a) high strain, (b) high psychosomatic complaints, and (c) high self-regulatory depletion. Power analysis 90 North American women were recruited for this event-based sampling study using Qualtrics panels. Longitudinal designs call for viewing the “repeated measurements as nested30 within individual subjects” (Hox, 1998). Carter’s (2016) thesis on ESM studies finds that because calculating power for longitudinal studies is difficult (often authors do not have a priori knowledge of how predictor and response variables vary between and within-persons), most ESM studies are conducted without offering a priori power analyses. Furthermore, for non- randomized control trial ESMs we do not yet have formulae to calculate power a priori (Carter, 2016). In her simulation, Carter (2016) asserts that power to find effects in an ESM study depends on the number of subjects (level 3), the number of days subjects are in the study (level 30 The participants in my sample will report on their role hierarchy, which is an organizational level variable. Since all the participants come from different organizations, I will not account for ICCs to aggregate perceptions on this variable. 156 2), and the number of observations gathered per day (level 1). Carter (2016) notes that increasing the number of participants, or increasing the number of moments in each day increases the power of the study instead of simply increasing the number of days. Carter’s (2016) simulations also find that 40 participants provide sufficient power to detect an effect size of 0.1. Further, Lu, Han, Chen, Gunzler, Xia, Lin, and Tu (2013) also assert that for longitudinal studies, power depends on (a) the within-subject correlation coefficient (r) such that a smaller sample size is needed as r approaches 0 and larger sample size is needed as r31 approaches 1, and (b) the number of post- baseline assessments such that smaller sample size is needed with greater number of assessments, and (c) the level of missing data can impact the sample size requirements. Accounting for a 15% attrition, Lu et al. (2013) recommend inflating the sample size n’ = n/(1- 15%); and note that with four assessments post-baseline, for a r = 0.3 the minimum sample size accounting for attrition was 65, and for a r = 0.5, the minimum sample size accounting for attrition was 89. In the current study, the data collection will unfold over four weeks to capture event-based data. Given the event-based design, it was difficult to predict how many data points would be returned, and the initial sample size estimate was based on the following precedents: in line with Lu et al. (2013), a sample size of 65 – 89 people should be enough. King et al. (2017) study recruited 61 participants for both within- and between-person effects), and as such a sample size of 60 for the current study was deemed reasonable (if four events of sexism are reported per person over the four week period, that would amount to 240 observations). To account for attrition, another 30 participants were recruited. 31 A higher r would indicate that the repeated measurements within a person yield data that is completely independent, as if the data is being captured from multiple other individuals, and therefore data is redundant because it yields no further information. 157 Screening Criteria To ensure accurate sample size estimate, it was critical to test the one sexism event per week assumption. Based on the Pilot Study (see Chapter Three) it was found that on average there were 1.2 sexist events that occurred over a period of four-weeks for each participant who stated that they faced sexism at work. Thus, when recruiting participants, I embedded a question on frequency of sexism experienced at work. Only respondents who indicated that they had experienced sexism at least four times over the past one month were recruited into the study. Additional screening criteria included recruiting women who had a full-time job, and making an effort to ensure that they were employed in industries ranging from manufacturing to service sectors to professional jobs, and also ensuring that they were specifically enmeshed in jobs that call for high levels of interactions (e.g., high levels of customer-facing requirements, and/or having to work in large teams with coworkers who are domestically or internationally dispersed). The effort to recruit women from a wide variety of organizations located in different geographies of United States was to maximize representativeness by getting a more diverse set of participants in terms of age, experience, and industry. The decision to limit the sample to women has already been described in chapter two. Respondents who had already participated in the Pilot Study, Study One, and Study Two were screened out of this study. Compensation Strategy One of the best practices in designing ESM studies is to compensate participants at multiple time points (cf., Christensen et al., 2003; Fisher & To, 2012; King et al., 2017). Participants were compensated at three time points. First, participants were compensated for filling out the recruitment survey where they shared their contact information if they agreed to participate in the research study on sexism at work, and upon agreement, were also administered 158 a baseline survey. As in earlier surveys, Qualtrics compensated participants based on pre- established modes of payment that were approved/preferred by participants. Once the initial recruitment was over, I compensated participants for the first ten days of participation in the study ($10 Amazon gift card). At the end of the study, participants were paid their remaining $30. One key issue in the sampling strategy outlined here is that of self-selection. In the recruiting materials, Qualtrics was instructed to explicitly state that this is a study about sexism at work. Furthermore, only women who had reported facing four or more sexist events over the course of a month were selected as participants. Generally speaking, self-selection bias is a problem for most studies, but specifically when designing ESM studies these concerns need attention in the design phase because research shows that a specific kind of person is attracted to time intensive studies – in addition to being highly motivated, conscientious, agreeable, engaged in white-collar jobs, etcetera (cf., Scollon, Kim-Prieto, & Diener, 2003), participants who responded to this study may view this study as a form of cathartic release. Clearly, this represents concerns of generalizability. If the goal is to understand how event-based stigma communication (in this study sexist events) precipitates identity management behaviors and various related outcomes then there was a practical need to find individuals who actually face a preponderance of sexist events. Thus, even though there was self-selection based on the levels of sexism experienced at work, it balanced the negative outcome of generalizability concerns with the positive outcome of having gained access to more insights about participants’ identity management behaviors and outcomes thereof. 159 Participants 90 female participants were recruited for this ESM study. Of these 60 participants participated in the study regularly and provided 269 events of sexism. Participants ranged between ages 19 – 68 years (M=35.54, SD =10.45). Among the participants, 6% were African American, 77% were Caucasian, 9%were Hispanic, 1% were Asian/Pacific Islander, 4% Native American or American Indian, and 3% were from other racial/ethnic groups. Note, participants were allowed to check multiple racial/ethnic categories. Participants reported a total work experience ranging from less than a year to 46 years (M= 14.77, SD = 9.71). Participants reported the tenure at their current job ranging from less than a year to 46 years (M= 6.89, SD =7.03). Diverse jobs were represented across the participants ranging from scientists, inventory counter, customer service, product specialists, teachers, service workers, etcetera. Method In their event-based ESM study, King et al. (2017) studied the identity management strategies used by lesbian, gay, and bisexual individuals in field settings by surveying them over a period of three weeks. To ensure that the participants are able to document sexism-based stigma communication events, the current study relied on King et al’s (2017) event-based ESM study precedent by extending their three-week data collection window to four weeks. Typically event based ESM studies rely on the use of paper-pencil measures or personal data assistants (PDAs; cf., Christensen et al. 2003; King et al., 2017) or more recently apps (e.g., Life Data) to log participant data. However, both PDAs and apps have their limitations. The operational requirements of PDAs can be fairly onerous (e.g., charging the batteries, repair issues, malfunctioning devices, inability to capture open-ended data, item types available in fixed formats) and the costs of PDAs can be prohibitive (e.g., costs of the devices, cost of data 160 recording software programs; Christensen et al., 2003). While apps are more convenient and less expensive than PDAs, most service providers charge a fee (e.g., Life Data charges $1750 for six month subscription for 500 participants; open source experience sampler app charges an iOS enterprise developer membership of $299/year in addition to paying for a server for data storage; the PIEL survey is open source and free). The PIEL survey has been used in previous ESM studies, however given its open-source nature, ensuring data privacy can be difficult. Instead, in the current study Qualtrics surveys were leveraged to create a virtual diary for participants. This “diary” was delivered via a simple email link that respondents accessed on devices they prefered. Operationally this choice was more practical because I had already retained Qualtrics’ services to create the participant panels, and also because aligning two distinct interfaces and/or web- services (i.e., Qualtrics panels with another ESM app that will be hosted on different web platforms/operating system) could have led to process inefficiencies such as compatibility issues (cf., Thomas & Azmitia, 2016). Thus, to avoid conditions for potential errors in data syncing, data storage, and data privacy concerns, Qualtrics platform was used. Once participants were recruited 32 and their informed consent was obtained, they were instructed to complete a baseline survey to gather their demographic data, information about their organization, and their base rates of psychosomatic complaints, strains, and self-regulatory depletion. At this point, participants were also emailed an instruction packet containing the following: (a) survey schedule (i.e., goal surveys sent each morning at 9 am and sexism surveys sent each evening at 4:30 pm); (b) instructions on identifying stigma communication events (i.e., any instances of sexism), and (c) instructions on understanding the different types of goals. In 32 Participants will be recruited via Qualtrics panels. Qualtrics manages the communication related to initial screening and extends the study invitation at their end. Thus the recruitment email has not been included in the appendix. 161 line with King et al’s (2017) event-based ESM approach, different sexism-related scenarios were described (see Appendix H). Event-level measures (Within-person; Level 1 variables). Each evening during the four-week period at the close of work-day (4:30 pm), participants were sent an email reminding them to document if any sexism-related events had occurred that day. Because participants resided in four different time zones of the United States, four different contact lists were created so the surveys were always received at 4:30 pm regardless of participants’ location. In addition, each morning during the four-week period, participants were asked to select among the following goals for the day at 9 am (e.g., ‘My goal is to showcase my performance in a positive light today’ (self-enhancement goal), ‘I aim to get along with my colleagues today’ (belonging goal), and ‘I aim to assert my identity as a woman at work today’ (self-verification goal). As before, four different contact lists were created to account for the time zone differences. Finally, participants were instructed to fill the following measures after each occurrence of the stigma communication event. Proximal goals. Participants were asked to select their most important goal for the day. The reason for this choice stems from the discussion on goals and identity management strategies in chapter one where I had created a rational mapping of the different goals (i.e., self- enhancement goals, self-verification goals, and belonging goals) under which people are likely to choose differential strategies. The rational mapping had to be created because so far the literature lacks such a conceptualization. Given the limited understanding about this relationship in the literature, to the extent possible, the current study was designed to elucidate distinct and clear relationships between different goals and different identity management strategies. 162 With this background in mind, the research design could have taken three “paths” to investigating the goals-identity management link under conditions of stigma communication, participants could: (a) tell the researcher about their goals each day (i.e., the goal variable could be an open-ended response item), and inform the researcher about their identity management strategies used if a sexist event occurs; (b) select multiple goals from a “universe of goals” identified by the researcher that they intended to pursue on a work day (i.e., self-enhancement, self-verification, and belonging goals in the current study would be the universe of goals), and inform the researcher about their choice of identity management strategies when faced with sexist event; or (c) select one goal as their most valued goal from a “universe of goals” identified by the researcher, and inform the researcher about their choice of identity management strategies when faced with the sexist event. If the research design allowed participants to select from an infinite universe of goals then the ability to connect different goals to identity management strategies would be limited, and it is for this reason that the first path was rejected from consideration even though it was clearly the more comprehensive of the three possibilities presented so far. The second path where participants could be given the opportunity to select multiple goals from the universe of goals selected (i.e., asking them to rate “All that apply” from self-enhancement, self-verification, and belonging goals) was an appealing option given how this option may be more realistic to peoples’ work experiences. For example, it is possible that when confronted with stigma communication, we may want to not only self-enhance ourselves, but also want to do it in ways that enhances our belonging to our work team. The issue with having participants “select all that apply” is that this realistic strategy will limit the current study’s ability to distinctly map identity management strategies to each of the goals. For example, it would be impossible to explain when 163 people seek to both self-enhance and belong, which goal is implicated in their choice of ‘X’ identity management strategy? The third path where participants are restricted to choosing just one of the goals from a pre-selected universe of goals risks forcing people into choosing goals that are most critical for their individual work contexts. For example, it may be the case that in investment banking kind of jobs there may be a press for selecting self-enhancement goals over all else, while in policing for example, the press may be more on selecting belonging goals. Such overt emphasis on just one goal had the clear advantage of providing distinct mapping to the types of identity management strategies chosen, but it also came at the cost of artificially forcing people to choose one goal even when they may have subjectively felt the tug of two or three goals during the course of a work day. These issues were at the forefront of my decision-making in asking participants to select their most proximal goal from the pre-determined universe of three goals (i.e., path c) given this strategy33 allowed for a clear explication of whether or not under different goals participants were likely to choose different identity management strategies. Thus, participants were asked to rate their most proximal goal (i.e., self-enhancement, self-verification, and belonging) at the start 33 An alternative strategy was also considered to address the pros and cons listed: I could ask participants to select multiple goals, but in the decreasing order of importance. This strategy would allow the research design to maximize the benefits of both paths ‘b’ and ‘c’ outlined above – by asking participants to select their most valued, proximal goal I could distinctly map each of these goals to the common identity management strategies, and utilize the multiple- goals’ rank-ordered data in an exploratory manner to see if these differentially relate to identity management strategies. However, MPlus maxes out at ten levels of a variable, and the permutation on proximal goals variable with three levels that are rank-ordered yields a total of fifteen levels that will have to be analyzed. 164 of each workday34. Definitions for each of the three types of goals were provided in the training, and were also presented on the survey itself. For instructions and measures see Appendix E. Role hierarchy. See Study One for details (Items in Appendix E). Accepting identity management strategies. Presented in Study One. Challenging identity management strategies. Presented in Study One. Self-regulatory depletion measure. This measure has already been presented in Study One. In this study, a shortened version of the self-regulatory depletion measure was used. In addition to using author’s reliability information, reliability for the ten-item shortened measure was calculated and had a Cronbach’s alpha of .92. Psychosomatic health indices. The 12-item Somatic subscale of the Hopkins Symptom Checklist (HSCL; Derogatis, Lipman, Rickels, & Uhlenhuth, 1974) was used to assess psychosomatic health indices, hereafter referred to as psychosomatic complaints. The items are rated on a 4-point rating scale (1 not at all and 4 extremely), and examples include: ‘Headaches’, ‘Faintness or dizziness’, etcetera (see Appendix E for full measure). Higher scores reflect greater somatic complaints, and the measure has Cronbach’s alpha of .88. Psychological Strains. The 12-item General Health Questionnaire (Goldberg, 1972) was used to assess psychological strain35. Items were rated on a 4-point Likert-type scale (see 34 Even though I am asking participants to report on their work goals each day, it should be noted that the ESM study is still an event-based study because I am interested in capturing participants’ identity management efforts in response to the sexist events. 35 The 53-item Brief Symptom Inventory (Derogatis, 1977; Derogatis & Melisaratos, 1983) is most commonly used to assess psychological strain. The measure taps into nine dimensions: somatization, obsessive-compulsive, interpersonal-sensitivity, depression, anxiety, hostility, phobic anxiety, paranoid ideation, and psychoticism. The measure has excellent reported reliability (a = .97; cf., Jex, Bliese, Buzzell, & Primeau, 2001). However, it is too long for the purposes of an ESM study and to my knowledge, a shortened version has not yet been created. 165 Appendix E for instructions, response scale, and items) by asking if the participants had experienced the symptoms at any time during the day. The measure had good reported reliability (Cronbach’s a = .82 - .90, test-retest = .73; cf., Hyman, 2001), and in the current sample Cronbach’s alpha was .88. Analytic approach Experience sampling method (ESM) data are hierarchically structured, that is, multiple observations are nested within the person. Excellent reviews on best practices to analyze nested data have been presented elsewhere (cf., Nezlek, 2001; Schwartz & Stone, 1998) and the analytical approach outlined below borrows from a close reading of multiple ESM studies that were reviewed (cf., Kane, Brown, McVay, Silvia, Myin-Germeys, Kwapil, 2007; Knouse, Mitchell, Brown, Silvia, Kane, Myin-Germeys, Kwapil, 2007; Nielsen and Cleal, 2011;). Hierarchical Linear Modeling (HLM), which is intuitively like a series of regressions, was used to study the relationship between the event factors and individual factors (Bryk, Raudenbush, & Congdon, 1994). A best practice in analyzing ESM data is to first conduct intra class coefficients (ICCs) that allow for an exploration of how the variance may be distributed at the between and within-person levels (cf., Nielsen & Cleal, 2011; Reise, Ventura, Nuechterlein, & Kim, 2005). To do this, first an empty model was specified without adding any independent variables, and assessed to see how the variance may be distributed around the mean of dependent variables at the event level and the individual level variables (Morrison, Payne, & Wall, 2003). Multilevel analyses should be used only if the variance is distributed at different levels. Intra class coefficients (ICC’s) were calculated and it was found that for all three dependent variables there was variance at both between and within person levels. Specifically, 68.7% of the variance for self-regulatory depletion is at the between-person level, and 31.3% variance was at the within- 166 person level. For strain, the between person variance was 55.4%, and within-person variance was 44.6%. Finally, 66.1% of the variance for psychosomatic health indices is at the between-person level, and 33.9% variance was at the within-person level. Once this condition was fulfilled, predictor variables were added to the model, and within-individual variables were centered relative to individuals’ scores (group-centering). There were no between-individual variables here to grand-mean center (since this is a 1-1-1, within- pereson model; Bryk & Raudenbush, 1992). At the event level of analysis (level 1) it was investigated which factors in the event predicted the different identity management strategies used, and the associated self-regulatory depletion, strains, and health indices for each group of within-person observations. Then, at the between-person level of analysis (level 2), the intercepts and slopes estimated at level 1 were regressed on the level 2 variables – that is, individuals’ mean values for independent variables, mediators, and outcomes for the day. To sum, in MPlus, a single-level mediation random effects model was specified for each of the dependent variables (i.e., for self-regulatory depletion, strain, and psychosomatic complaints; see syntax in Appendix G). Specifically, all within level variables were entered and participant ID was entered as the clustering variable (i.e., between level variable) so observations could be denoted as being nested within a person. The syntax specified the model that MPlus should run followed by regressing combinations of dependent variables on independent variables per the hypotheses presented earlier (see Table 22). In addition, please note that the proximal goal was a categorical independent variable with three different levels (participants chose one goal as the most salient goal out of self-enhancement, self-verification, and belonging goals in the stigmatizing situation). Identity management strategies were measured as a choice, that is, 167 participants indicated which strategy they used in navigating the stigma communication event (see Appendix E), and this data was classified as accepting or challenging stigma strategies. In case participants chose to write an open-ended response, the response was classified as accepting or challenging strategy before data analysis. Thus, identity management strategy was a categorical dependent variable in this study and as such a logistic regression in MPlus was used (Tofighi & Thoemmes, 2014). Including the full empirical model yielded an unreadable figure, and as such a reduced empirical model with statistically significant estimates is presented (See Figure 15). Figure 15: Empirical paths with estimates for Study Three. (Note: Goals were dummy coded and the referent is Belonging goals). 168 Preliminary Analyses Results Prior to performing hypotheses testing analyses, all data were examined for accuracy, missing values, normality, and assumptions of multivariate analysis using MPlus SAMPSTAT command. Analyses indicated that skew and kurtosis were within an acceptable range (i.e., absolute value of skew < 2, absolute value of kurtosis < 7; West, Finch, & Curran, 1995); therefore, transformations were not necessary (see Table 23). Table 23: Means, Standard deviations, Skew, and Kurtosis of Study Three variables. Variables Strain Psychosomatic complaints Self-regulatory depletion Mean 26.2 1.66 4.11 S.D 6.21 .53 1.49 Skew .03 1.01 .27 Kurtosis .53 1.48 -.63 Descriptive Statistics. All three scale variables in the current study showcased acceptable skew and kurtosis. However, on psychosomatic complaints, a floor-effect is evident. Respondents rated a four-point Likert type scale for psychosomatic complaints (Min = 1, Max = 4), however both the mean and standard deviation for this variable are fairly low (M = 1.66, SD = .53). Of the accepting identity management strategies, ‘ignoring what is said by my teammates’ was most frequently selected, followed by ‘accepting what is said by my teammates’, and ‘keeping to myself’ (see Table 24). Please note, here “teammates” was not defined for the participants, so they may have had difficulty with choices because of interpretation of the item – i.e., it is unclear if respondents thought of supervisors and clients as their teammates, or if they reverted to thinking about their peers and juniors as their teammates, or if they thought of teammates as everybody they interacted with. 169 Table 24: Frequency of endorsement of accepting identity management strategies. Accept IM Strategies Description Accepting what is said by my teammates Ignoring what is said by my teammates Keeping to myself Avoiding my teammates Trying to conduct myself in a manner inconsistent with my teammates' expectations Trying to bond with others like me Trying to use self-deprecating humor Frequency 74 87 64 33 21 19 28 Of the challenging identity management strategies, top three identity management strategies were ‘speaking up’ (the highest endorsed strategy), followed by ‘trying to communicate inaccuracies in their thinking’, and ‘trying to educate my colleagues’ (see Table 25). Table 25: Frequency of endorsement of challenging identity management strategies. Challenge IM Strategies Description Speaking up Trying to change things Working harder to prove my teammates wrong Trying to educate my colleagues Trying to communicate the inaccuracies in their thinking Trying to make favorable social comparisons that make me look good Denying the attributions my teammates make about me Frequency 71 20 36 44 49 13 23 Correlations. The correlations between the study variables are described (see Table 26) and key relationships are described here. As would be expected, strain was significantly correlated with psychosomatic complaints (r = .37, p <.01) such that women who reported high levels of strain also reported higher psychosomatic complaints. Strain was significantly and negatively correlated with self-regulatory depletion (r = - .64, p <.01), such that women who reported high levels of strain, also reported being more depleted at work. Women who had received sexist behavior from their supervisors or from their juniors tended to report higher 170 levels of strain (rsupervisors = .29, p <.01 and rjuniors = .13, p <.05). In contrast, when women encountered sexist behavior from their peers at the same level, they tended to report lower levels of strain (r = -.16, p <.05). Women who had received sexist behavior from their supervisors also tended to report higher levels of depletion (r = - .29, p <.01). Finally, psychosomatic complaints were significantly correlated with self-regulatory depletion (r = -.54, p <.01) such that women who reported high levels of psychosomatic complaints also reported high depletion at work. Psychosomatic complaints did not correlate with other study variables. When women received sexist behaviors from their juniors, they were more likely to engage in challenging identity management strategies over accepting strategies (r = .16, p <.01). Women who selected self-enhancement goals reported using more accepting identity management strategies than challenging identity management strategies (r = -.14, p <.05). Table 26: Correlations of study variables. 3 1 .88 2 .37** .88 4 5 6 -.64** .29** -.16* .13* -.08 .02 -.54** .10 .01 .12 -.07 .06 .92 -.24** .06 -.06 .10 .01 - -.38** -.13* -.25** -.09 - -.17** -.30** .01 - -.15* .16** Variables 1. Strain 2. Psychosomatic complaints 3. Self-regulatory depletion 4. Supervisor 5. Same level 6. Junior 7. Client 8. IM strategy 9. Goal: Self- enhancement 10. Goal: Self- verification 7 - -.06 8 - 9 10 -.01 -.10 .06 -.04 .06 -.18** .04 -.14* - .03 .03 - (Note: IM = identity management strategies, dummy coded Accepting = 1 and Challenging = 2; Goals are dummy coded; ** indicate p<.01 (two-tailed); * indicate p<.05 (two-tailed) -.57** -.13* -.08 -.01 .09 .01 .05 171 Main Effects and Mediation Hypothesis 17a predicted that accepting identity management strategies would be related to low strain and hypothesis 18a predicted that challenging identity management strategies would be related to high strain. Results show a statistically non-significant effect for both these relationships (β = -.24, p = .52; note: accept and challenge identity management strategies in the current study were dummy coded as categorical variables, hence there is just one beta weight). Hypothesis 17b predicted that accepting identity management strategies would be related to low psychosomatic complaints and hypothesis 18b predicted that challenging identity management strategies would be related to high psychosomatic complaints. Results show a statistically non-significant effect for these relationships as well (β = -.04 p = .28; note: accept and challenge identity management strategies in the current study were dummy coded as categorical variables, hence there is just one beta weight). Hypothesis 17c predicted that accepting identity management strategies would be related to low self-regulatory depletion and hypothesis 18c predicted that challenging identity management strategies would be related to high self-regulatory depletion. Results show a statistically non-significant effect for both these relationships (β = .07, p = .34; note: accept and challenge identity management strategies in the current study were dummy coded as categorical variables, hence there is just one beta weight). Hypothesis 8 predicted that within-individuals, role hierarchy would impact an individual’s decision to accept or challenge stigma such that when an individual receives stigma communication from those (a) lower in the role hierarchy she would be more likely to challenge stigma but when she receives stigma communication from those (b) higher in the role hierarchy 172 she would be more likely to accept stigma. Results show a non-significant relationship between facing sexism from those who were lower in the hierarchy (i.e., their subordinates and/or juniors) and use of challenging identity management strategies versus accepting identity management strategies (β = .38, p = .06). It was also found that when stigma communication was received from supervisors and clients, that is those higher in the role hierarchy, participants were more likely to use accepting identity management strategies than they were to use challenging identity management strategies (β = -.26, p = .28). While this relationship was not statistically significant, it was in the expected direction. A null effect was found for the relationship between respondents’ use of accepting identity management strategies and encountering sexist communication from those at the same level (i.e., peers and colleagues; β = .01, p = .96) Hypothesis 15 predicted that within-individuals, the proximal goal during a stigma communication event would be differentially related to (a) accepting and (b) challenging strategies. Results show that, when participants selected self-enhancement goals in comparison to belonging goals, they were more likely to choose accepting identity management strategies as compared to challenging identity management strategies in navigating sexism (β = -.39, p = .03). In contrast, no statistically significant relationships emerged between participants’ choice of self- verification goals and their choice of identity management strategies (β = -.16, p = .52). Hypothesis 10 and Hypothesis 16 predicted mediation36 between the key independent variables and dependent variables. Specifically, Hypothesis 10 predicted that within-individuals, (a) accepting stigma would partially mediate the relationship between role hierarchy and self- 36 To test the mediation effects, I specified indirect effects by using the ‘MODEL INDIRECT’ command. As mentioned earlier, bootstrapped confidence intervals were computed by using the ‘CINTERVAL’ command, and confidence intervals that did not include zero were used to indicate significant mediation. 173 regulatory depletion, and (b) challenging stigma would mediate the relationship between role hierarchy and self-regulatory depletion. Hypothesis 16 predicted that within-individuals, (a) accepting stigma would mediate the relationship between proximal goal and self-regulatory depletion, and (b) challenging stigma would mediate the relationship between proximal goal and self-regulatory depletion. None of the mediations were significant in this study (see Table 27). Table 27: Mediation hypotheses for Study Three. H# Hypothesis H10a Within-individuals accepting stigma would partially mediate the relationship between role hierarchy and self-regulatory depletion H10b Within-individuals challenging stigma would mediate the relationship between role hierarchy and self-regulatory depletion. H16a Within-individuals accepting stigma would mediate the relationship between proximal goal and self-regulatory depletion H16b Within-individuals challenging stigma would mediate the relationship between proximal goal and self-regulatory depletion. Supplementary Analysis Beta-weight (p- value) for indirect paths β = .01, p = .42 β = .01, p = .42 β = .00, p = 1.00 β = .00, p = 1.00 While the main effects between role hierarchy and dependent variables were not specified, the empirical paths were tested in the path analytic framework. Results show that when receiving sexist communication from supervisors and clients, participants tended to report more strain (β = 2.10, p = .01) and higher self-regulatory depletion (β = -.37, p = .07). Non-significant findings were obtained for the relationship between role hierarchy and psychological health indices (β = .02, p = .85). These results should be interpreted in context – that is, in this study, 174 participants were working in environments where they confronted frequent sexism. Discussion One of the key questions that emerged from the literature review presented in chapter one, and the basis of this study, was if women, when confronted with sexism, are likely to choose different identity management strategies based on their goals in the workplace, and whether or not these choices would be implicated in predicting self-regulatory depletion, strain, and psychosomatic complaints at a within-person level. Thus the specific goals of this study were to explore: (a) if stigma communication impacted outcomes such as self-regulatory depletion, psychosomatic complaints, and strain as a function of role hierarchy and goals; (b) if there may be differential impact of goals on choice of accepting and challenging identity management strategies used when faced with sexism; and (c) if accepting and challenging strategies may differently impact key outcomes. As such, goals and role hierarchy were included as important antecedents of choice of the identity management strategies, which in turn were tested as mediators between goals and role hierarchy and the outcome variables (i.e., self- regulatory depletion, strain, and psychosomatic complaints) reported by respondents. Stigma communication, role hierarchy, and outcomes Stigmatization has been related to a host of negative outcomes such as strain and deleterious health outcomes (cf., Stuber, Meyer, and Link, 2008), and studies show deficits in psychological well-being, interpersonal mistreatment directed at those with stigmatized identities (Carr and Friedman, 2005), and work discrimination (Ragins and Cornwell, 2001) to name a few. Findings in the current study are in line with this body of work, and also replicated findings from the experimental study outlined earlier where it was shown that vividly recall a sexist event was related to self-regulatory depletion. Specifically, the current results show that women facing 175 sexism reported higher levels of strain and self-regulatory depletion when their supervisors or clients engaged in sexist behaviors towards them. However, no significant effects emerged for psychosomatic complaints37. As in Study Two, power differential embedded in perpetrator’s position in the role hierarchy with respect to the women reporting on stigma emerged as an important attribute in predicting women’s outcomes at work. Together these results speak to the powerful and deleterious impact of sexism on women. The fact that even retrospective recalls of a sexist event created a press strong enough for women in the moment to feel psychologically depleted (Study One), and the associations of daily reports of sexism and consequent strain and depletion (current study) are testament to why organizations need to take the problem of stigmatization in the workplace seriously. Stigma communication, daily goals, and identity management strategies The theoretical model highlighted in chapter one specified the importance of evaluating how proximal goals may be implicated in how individuals with stigmatized identities navigate the stigma communication events. While the first test of this question in Study Two did not yield statistically significant findings, in the current study, it was found that daily self-enhancement goals, in comparison to daily belonging goals, were implicated in women’s choices of identity management strategies. Specifically, as expected, when women selected self-enhancement goals as proximal for the day, they chose more accepting identity management strategies as compared to challenging identity management strategies in navigating sexism at work. In contrast, there was no statistically significant relationship between self-verification goals and choice of identity 37 Research shows that psychosomatic complaints are a result of strain, with strain mediating the sexism-complaint relationship. However, in the current study, this link was not tested. 176 management strategies (this complements the non-significant patterns observed in Study Two). Two key insights emerge from this study. First, self-enhancement goals were more frequently endorsed (N = 128) than belonging (N = 62) and self-verification goals (N = 58). Thus, it is evident that women endorsed self-enhancement goals as the more important goals to fulfill at work. In light of this perceived importance, it is not surprising to see women revert to using the “low cost” accepting identity management strategies over the more “high cost” challenging identity management strategies. As was theorized in chapter one, challenging identity management strategies may detract from one’s efforts to showcase performance at work and focus on getting the job done. This is further corroborated by the fact that in the open-ended data from Study Two, women recommended adding more task-focused identity management strategies to the presented strategies. Limitations Only respondents who indicated that they had experienced sexism at least four times over the past one month were recruited into the study. This was done so the data collection yielded enough sexist events to ensure that statistical power to find effects was not eroded. Alternatively, to increase the power to find effects, the number of participants could have been increased. However, since this study was an experience sampling methodology, it was costly and resource intensive, and the practical constraints had to therefore be handled creatively. Furthermore, the screening criteria cut-off of four sexist events per month was arbitrarily chosen here. Clearly, respondents in the current sample were women who work in sexist environments as they are experiencing greater levels of sexism. Two open questions emerge from this limitation. First, it would be interesting to explore whether or not the frequency of use of strategies is affected by the frequency of experiences of sexism. That is, if a woman works in an 177 environment where sexism is more frequent, is she less likely to put forth the effort and energy into challenging every occurrence of stigma communication? Second, could goals become influenced by stigma levels such that women who work in high sexism environments tend to give up on their belonging goals because the perceived discrepancy between the need to belong and the possibility of ever belonging is just so high? Based on the theory outlined in the Chapter One, I would expect to see this, but this idea of goal abandonment in the face of large goal discrepancies will need to be put to test in future empirical work. 178 CHAPTER 7: GENERAL DISCUSSION In this chapter, I close the loop on how the three studies (i.e., Study One, Study Two, and Study Three) taken together contribute to the literature on identity management and stigma. In the discussions for each of the studies presented so far, the key findings were interpreted and discussed with respect to the literature, and limitations for each study were also highlighted. In contrast, this General Discussion elucidates how the multi-study, multi-method agenda contributes to the literature overall. In synthesizing the findings across the three studies, the theoretical model conceptualized in Chapter One is treated as a reference. A multi-study, multi-method research agenda was crafted with the aim of contributing to the literature on identity management and stigma in three key ways. First, it was posited that when people face stigma communication in work settings their options to choose various identity management strategies may be constrained. Second, it was posited that work environments may make some goals such as self-enhancement more salient than other goals, thereby further imposing a press on the type of identity management strategies that an individual can choose in response to stigma communication. Specifically, it was argued that ascribing self-verification motives and identity centrality to those who confront stigma communication is an unwarranted assumption, and identity management choices at work may instead unfold as a function of the more proximal self-enhancement goals given how organizations are driven by strong productivity norms. A theoretical case was also made to study aspects of peoples’ work experiences such as choosing between multiple goals that may not align with each other, and/or persisting with self-enhancement or productivity goals even in the face of stigma threat. While multiple goal choice was not tested in the empirical agenda, the influence of self-enhancement goals in addition to belonging and self-verification goals was tested. Finally, it was highlighted 179 that to explicate responses to stigma communication and resultant identity management strategies of accepting or challenging stigma communication we need to take an event-based view of stigma communication. Specifically, it was posited that in work contexts in particular, role hierarchy (of the perpetrator with respect to the target of stigma communication) would emerge as an important antecedent to decisions on accepting or challenging stigma communication. Thus, a case was made that viewing identity management behaviors as self- regulated, goal-oriented, norm based choices in work contexts is important if we are to understand how people respond to stigma communication in work contexts and how receiving stigma communication may impact key outcomes such as self-regulatory depletion, strain, and psychosomatic health indices. Specific findings from each study have already been presented in previous chapters (see Chapter Four, Chapter Five, and Chapter Six). In this General Discussion, my goal is to pull the various threads underlying the overarching research agenda together and weave them in a cohesive narrative. Thus, the goal is to shed light on how the key ideas that formed the backbone of the theory-building attempt outlined in this dissertation faired overall when empirically tested. Thus, in the first section below, instead of discussing each specific component of the overall model, I discuss key findings across the three studies to illuminate the construct of identity management as it relates to stigma communication and goals at work. In the second section, I discuss how identity management strategies of accepting and challenging stigma relate to key outcomes at work such as self- regulatory depletion, strain, and psychosomatic health indices, and what these relationships mean when they are contextualized in broader body of literature. In the third section, I discuss the key limitations of the overall methodological choices utilized in the current research agenda and highlight areas of future research. Finally, in the fourth section, I discuss practical implications 180 of the findings as they relate to organizations specifically. Identity management under stigma communication in work settings Identity management strategies are typically described as a vast array of options that are available to an individual when s/he confronts stigma communication (cf., Goffman, 1963; Meisenbach, 2010), but this assumption may not hold true in work settings. It was evident in the current studies that work environments constrain the choice of available identity management strategies to women. Data show that women reported using greater challenging identity management strategies when describing what they would do when faced with stigma communication that was sexist in nature, but in writing their vivid recalls many also ended up accepting stigma communication. This discrepancy is a form of Illusion of Control bias (Thompson, 1999) whereby people tend to overestimate the extent to which they may be able to influence events that are not in their control. This general tendency has been observed in the context of gender harassment as well. Research shows that women, when perceiving and coping with sexual aggression threats, tend to show an “unrealistic optimism”, “exaggerated perceptions of mastery”, and “overly positive self-evaluations” when thinking about how their future self would navigate threats (Norris, Nurius, and Dimeff, 1996; p. 12). However, I posit that while there is a great press on women to say that their future-self would challenge stigmatizing communication, in reality the situational press of the work environments may push them into accepting stigma communication. This claim was supported in Study Three, a field-based diary study, where 61.34% women reported using accepting identity management strategies in response to stigma communication on the job. In addition to the empirical data presented here, this attitude-behavior discrepancy has also been famously documented in LaPiere’s (1934) study where it was evident that people’s 181 endorsed attitudes on a questionnaire do not always line up with their behaviors under the press of an actual situation. In fact, the open-ended data in the Pilot Study and Study One further illuminated this discrepancy. Women felt that they did not have as many degrees of freedom in deciding what to do about the sexist communication directed their way, and felt that their livelihood may be endangered and/or their career would take a backseat if they attempted to do something about the stigma communication at hand. Here are a few examples: I do homecare and one of my clients was a man around 65 [years] with colon cancer…He touched my arm as he was talking and he said, I touched you. Even though it was just an arm, that gesture with his comment made me feel like a sexual prey…Even though I was by myself with him in his house I wasn't scared as I knew he couldn't go any further physically but he kept on touching me and caressing my arm skin often. I never reported him nor said anything to him. I don't know why. At the time I wasn't working many hours and I wanted to keep working without making any waves. I didn't like it but I didn't stand for myself. (Respondent A; Emphasis added) My male boss treats the men differently from the women in my office. He is [sic]. speaks to the women, me included, with disrespect and asks them to do things that he does not ask the men to do. He asked me to do an administrative task that was outside of my job description when I was busy and overwhelmed doing my main job. When I asked him for help, he spoke to me in a very dismissive way. When I suggested that a male coworker who was not busy could help me, he didn't even consider it and would not give me help. This made me mad, and frustrated and disrespected. I don't have a good way of coping with his sexist behavior. I ended up doing what he asked for, not saying anything to his boss and basically just accepting it. (Respondent B; Emphasis added) I was up for a promotion and would be placed on a team that was all men except me. It was insinuated by the lead of the team that I would not be a good fit since there would be travel involved, and I would be traveling with all men, and what would my husband think about that? I felt awful because I knew I was the best person for the job yet was likely turned down for the role since I was a woman and would be working with all men. I felt anger, hurt, disappointment and a myriad of other emotions. I coped the way most women cope I suppose. I just accepted it since I couldn't prove that was the reason I wasn't given the position. There's really not much you can do in a situation such as mine since the reason they gave me for not getting the position was that the team simply preferred another applicant who was probably equally qualified but was a man. (Respondent C; Emphasis added) 182 Together, the empirical findings and the qualitative data here show that the assumption that individuals will have a diverse array of identity management strategies available to them in work settings needs to be questioned explicitly in future research undertakings. In the #MeToo era, questions are often raised about why women do not speak up against stigma communication at work. I posit that in that question lies a form of victimization (cf., Koss, 1990), for it is the unfortunate case that challenging stigma communication may not always be perceived as a viable strategy in work contexts for all women. In fact, these findings lead me to wonder to what extent is challenging stigma communication a form of privilege? People may be embedded in jobs where sexism (or other forms of stigma communication) is just assumed as the norm. In such environments, to speak up against stigma communication may not be seen as an act of self-verifying one’s identity, but may instead be seen as an attack on the power structures that are in place. Women may view challenging stigma communication in such environments akin to painting a red dot on themselves, that is they may see challenging identity management strategies as a way to further “mark” themselves as a target, which would put them in further “peril” (Meisenbach, 2010). In the open-ended data, respondents spoke freely about how they viewed challenging stigma communication as a strategy that may exacerbate the negative outcomes on the job such as threats of loss of job, punishment (e.g., in the form of bad performance appraisals), or even threats of physical and bodily harm. For example, one respondent noted: “I was propositioned to have sex with my boss, when I didn’t I was bullied and fired”. Another respondent recounted her experience as follows: My boss was attracted to me…he just kept on sending me flowers…he [would] bump my breasts, my butt, and I didn’t like it, but I was scared to lose my job, [until] he just grabbed me somewhere else that was really really inappropriate, so 183 I finally stood up to him and [told] him if you do not stop, I’m going to get you in trouble; he said, “Why you ain’t going to have a job!” (Emphasis added) In line with research on sexual harassment, threats of physical violence, and/or actual violence (cf., Fitzgerald, 1993; Fitzgerald, Drasgow, Hulin, Gelfand, & Magley, 1997) were also reported, and some women who had suffered prior attempts at sexual assaults, noted facing subsequent assaults. For example, one respondent shared the following incidents at two different jobs: I was locked in a freezer with a male employee who told me that if I didn't make him happy he would make sure that I was fired. He constantly was grabbing me in front of customers and always made me feel very uncomfortable. I was in my late teens when this took place and when I was able to talk about it I was told that I just needed to grow up and not live in the past. I went to corporation about it and nothing was ever done to the male employee. I was humiliated in front of many people and constantly lived my life in fear after that. I left that job and moved on. Once I got another job, the same thing happened only this time I was raped. When I went to my supervisor … she told me the same thing as before that I just needed to grow up and deal with life, that things happen. I was even suspended for a week for reporting it. I have never felt so disrespected in my whole life. A couple of years passed and I would just avoid all meetings and anything to do with my work life out side of work. All the while I was being called a whore and slut. I stood my ground and transferred out of that area to a different one. I am still with the company after 16 years and I still refuse to be a part of any after work activities out side of my daily shifts. (Respondent D) The studies outlined here consistently point to how in stigma communication events, self- regulation takes on an even more important role in organizational settings because an individual while dealing with the stigma threat also has to navigate goals at work, and work within the norms of her role. So the key premise here was that these constrained choices in response to stigma communication events impact the key outcomes of self-regulatory depletion, strain, and psychosomatic health indices (discussed later). People are active agents who monitor their environment closely and self-evaluate how to self-regulate their responses, i.e., which identity management strategies to pursue in the given stigma communication event. One interesting question is if all types of stigmatizing behaviors operate in the same way 184 with respect to the identity management behaviors they might precipitate. That is, might there be some types of sexism for which identity management could be potentially different across different types of harassment? While I could not find direct evidence for this, there is some indirect guiding evidence. Major (2004) found that the degree of identity management behaviors used varied with the degree of threat in an individual’s environment. Specifically, pregnant women who perceived greater degree of stigma threat went to much greater lengths in safeguarding their identities by utilizing several different identity management strategies at work. Future research should examine if the degree and/or types of identity management behaviors change in response to the degree and/or type of sexism encountered by an individual. Alternatively, a reader could also make the claim that because sexual harassment behaviors are strong situations, it is likely that these constrain an individual’s behavioral options to an extent that the role of self-regulatory processes is minimized or completely eliminated. For example, in a high threat situation, a person may simply react to the threat without rationally thinking about one’s goals in the moment. This affective reaction definitely occurs reflexively and has been explored in the literature (Pryor, Reeder, Yeadon, & Hesson-McInnis, 2004). However, most work settings place a high premium on modes of conduct where negative affective displays are to be minimized (Diefendorff, Erickson, Grandey, & Dahling, 2011) and it is for this reason that studying the role of identity management behaviors as self-regulated, goal- oriented responses also becomes important. Identity management under stigma communication at work: Role of goals In the identity management literature, when an individual faces stigma communication, the role of an individual’s goals is not studied. It was posited that the productivity norms that define most organizations constitute strong situational press thereby further constraining the 185 choice of identity management strategies available to people who encounter stigma communication in organizational contexts. With this in mind, three key goals of self- enhancement, self-verification, and belonging were tested with respect to an individual’s decision matrix on different identity management choices of accepting or challenging stigma communication. While Study Two did not find any relationships between the types of goal condition (i.e., belonging, self-verification, and/or self-enhancement goal conditions) and the choice of different identity management strategies in a path analytic framework (a detailed discussion was provided in Chapter Five), chi-square analysis found that under self-verification goals women were significantly more likely to choose challenging strategies. On the question of which goals were considered more important at work, the data from Study Three show that on days when women faced sexist events, they selected self-enhancement goals 48% of the time, in comparison to selecting belonging goals 23% of the time, and self- verification goals 22% of the time. If all days are included in the analysis, that is including days when respondents did not face sexism, women chose belonging goals 47% of the times, self- enhancement goals 37% of the times, and self-verification goals 16% of the times. It is important to note here that even though self-verification goals are chosen the least, in stigma and identity management contexts, these are often assumed to be acting on individuals. In the open-ended data in Study One, respondents noted focusing on effectively executing tasks at hand as a form of identity management strategy during stigma communication events (coded as accepting identity management); they indicated that instead of focusing on their identity they would showcase their positive job performance to their team-members and override the stigma. This is in line with the theory outlined in Chapter One where it was posited that much of identity 186 management literature fails to take into account how the norms of productivity may often restrict individuals from completely abandoning their self-enhancement and/or belonging goals in favor of self-verification goals. That is, at work, one needs to accomplish work tasks and get along with others in order to maintain employment. Instead, it is often assumed that most individuals encountering stigma communication ought to or would prefer to assert their identities. However, data here show that self-verification goals in work contexts were treated as less important than self-enhancement and belonging goals. This intuitive expectation of self-verification motives is evident socially in calls that ask women to speak up (aka #MeToo movement described earlier) and is also evident in organizational policies of speaking up against the stigma communication (for example, unless a target follows the process of an organization to complain about the stigmatizing events, she has no legal grounds to claim sexism/sexual harassment). In the current studies accepting stigma communication (versus challenging stigma communication) was related to negative (versus positive) outcomes on the job (described later in detail). So asking women to speak up is a well- intentioned advice. From a legal standpoint, organizations’ expectation that women report sexist behaviors also makes sense. And yet, an important point gets lost in this well-intentioned advice – that to the extent a woman is faced with a strong environmental press where her self- enhancement and/or belonging goals are overriding influences, her options to engage in challenging stigma communication are likely to be constrained. How oppressive might it be to then not only have received stigma communication for her gender identity, but to also be reminded of how she has failed at asserting herself or verifying her identity? How might a target process the fact that while socially and organizationally she is expected to speak up, that in failing to act to assert herself, she is doing a disservice to not just herself but others like her? 187 Research shows that targets often feel oppressed and alone in facing stigma communication, feel powerless in changing their circumstances, and often live with the threat of being subjected to harassment (Ferraro, 1996). The qualitative data mirrored these themes. In line with research on sexual assaults and impact on targets (cf., Koss, 2000; Weiss, 2010), the open-ended data also showcase how women had feelings of guilt and shame when they accepted stigma communication. Needless to say, in a lot of these cases women also referenced quitting. This is especially true for women who were sexually assaulted; here is a case in point: When I was 19, I was working at a perfume shop in the mall…My manager who is the owner of the store, asked me to come into his office…As soon as he closed the door, he was standing behind me, groping my boobs. I could feel his penis harden from behind me and I was instantly scared. No one was working and there was no one in the store who would hear me if I screamed! I just froze there because I wasn't sure what to really do… I was afraid that he was going to fire me and I felt disgusted with myself that I didn't prevent such a situation…He kept more distance from me after that. My roommate convinced me to quit, so I did. I ended up not pressing charges though now reflecting on it, I should've as he could of tried to take advantage of any woman after me! I was just so humiliated and I didn't want anyone to know at the time that I kept my mouth shut. (Respondent E) Another respondent described how she was sexually harassed as follows: The person who was sexist to me was the actual physician I worked for as a nurse. He constantly treated me better than some of the other employees who were male or even other females. He would text me outside of work hours asking if he could see pictures of me…He only wanted nurses that were "pretty" to work for him. I confronted another nurse about this who worked for him and she acted like it was no big deal because she too was spoken to that same way but she knew she had job security out of it. I felt ashamed. I felt used and not worthy that my actual talent got me the job, but instead my beauty did as a female. I coped with this by quitting this nursing job that was high paying and had great hours. (Respondent F) Furthermore, antecedents such as identity centrality are hailed as central to most conceptualizations in the literature on identity management and stigma (cf., Jones & King, 2013; Ragins, 2008). Instead, in the current research agenda, identity centrality failed to predict any of 188 the relationships; specifically, it was expected that higher identity centrality would be related to greater self-regulatory depletion, identity centrality did not predict depletion in both Study One and Study Two. Similarly, while it was expected that high (versus low) identity centrality would be related to using more challenge (versus accept) identity management strategies in response to stigma communication, both in Study One and in Study Two, the findings were statistically non- significant. These null findings were explained in the discussion sections presented earlier (see Study One and Study Two). Specifically, I had described a meta-analysis (Lee and Ahn, 2013) that reported several mixed effects for the role of racial centrality and distress in discrimination contexts (see Study One for detailed discussion). One of the central aspects of the meta-analytic findings was that racial centrality both buffered and exacerbated the perceived discrimination and associated distress. Might it be the case that in the current research, gender-based identity centrality had a similar impact such that it both buffered and exacerbated the perceived sexism and associated distress, thereby creating a null overall effect? Instead goals emerged as important proximal antecedents that constrained individuals’ identity management behavior choices over identity centrality. While it is true that women think about their identity, and also about their “sisterhood” with respect to how they navigated stigma communication at work – that is, in their open-ended data, women referenced how their lack of action against sexism and/or harassment may have impacted other women who would come after them – it seems they do so more as post-hoc rationalizations of why they engaged in the said behavior, instead of focusing on these higher order issues when confronting stigma communication itself. Stigma communication event is a strong, threatening, and demeaning situation, and cognitive and affective resources likely kick in to first navigate the situation. Consider this example: 189 I work full time in an office setting at a major University hospital, but I also work part time at a restaurant. I’m a single mom, so I do what I have to do; but I am highly educated. One time at the restaurant a customer told me that I had such a pretty face and sweet personality, it’s “too bad I didn’t get an education and get a real job.” He also said “it’s great there are jobs in the restaurants so women can use their looks to make money.” I felt disgusted, but unable to verbally defend myself because he was a customer. I feel like I let myself down, and all other women working in the industry. (Respondent G) Both identity centrality and self-verification goals in the current study did not emerge as proximal determinants of choice of identity management behaviors at work. Instead it was self- enhancement goals that held more explanatory power. So based on both the empirical and qualitative data in these studies, ascribing the assumption that women would think about their identity as a woman first (i.e., identity centrality), and would therefore naturally want to self- verify (i.e., self-verification goals) when faced with a stigmatizing event in organizational settings are both unwarranted assumptions. There is an overwhelming press of self-enhancement and belonging goals and having to abide by work norms, and these concerns need to taken as front and center constructs of interest when studying identity management choices that are deployed when confronting stigma communication at work. Identity management under stigma communication at work: Influence of role hierarchy To the third and final contribution that was intended in this body of work, it was highlighted that responses to stigma communication and resultant identity management strategies do not just unfold in the same way regardless of situational attributes. Specifically, a case was made that we need to take an event-based view of stigma communication to better explicate the various pathways that could underlie a decision to accept stigma communication or challenge stigma communication. In work contexts in particular, whom one receives stigma communication from (i.e., the perpetrator’s role relative to the role of the target) was expected to 190 emerge as an important antecedent to decisions on accepting or challenging stigma communication. The research agenda provides support for this idea. When supervisors and/or clients directed stigma communication, women were more likely to revert to accepting stigma communication. And when clients directed the stigma communication, women were less likely to engage in challenging identity management strategies. Once again, these findings speak to the main focus of this dissertation. It is unrealistic to expect that in organizational contexts, people are free to choose from a wide array of identity management strategies. The organizational hierarchies, organizational culture, and power structures etcetera are all strong forces that constrain norms of behavior at work. In fact, in socializing new employees, significant amounts of time and resources are invested to coach employees on specific work norms, rules of emotional expression, and rewards and punishments for meeting the said standards (cf., Rafaeli & Sutton, 1987). Given the many different levers that constrain authentic expression at work, those facing stigma communication events may feel limited in their ability to take action against the perpetrators. Note the role of power as another strong determinant of challenging stigma communication has already been discussed and therefore not included here (see Chapter Four). Ascribing the burden of action or non-action on to the same people who face the brunt of stigmatizing situations based on their identity affiliation has the potential to quickly translate into a form of victim blaming (cf., Ryan, 1976). To what extent is asking someone to challenge stigma communication akin to one’s own reliance on the Defensive Attribution Hypothesis (Shaver, 1970; Walster, 1966) that posits that well-meaning observers may want to retain their own sense of control on their life outcomes and therefore can find fault in those who face forms of indignities directed against their person? Is it a form of defensive attribution to recommend that targets of sexism and/or harassment need to engage in challenging stigma communication 191 because it would make the observers feel more in control of their own life trajectories? Open- ended data described earlier show that women who ended up not challenging the stigma communication and/or their perpetrators have felt the weight of that inaction and indecision for years at a time. This point further underscores the importance of considering identity management behaviors from a self-regulation perspective because not only do women navigate the indignities directed against them in the moment while also juggling their work norms and press of goals, but in many cases they also described their continued efforts at shouldering the burden, the guilt, and the pain for inaction over long time arcs (sometimes spanning several years). Research shows that women tend to showcase higher incidence of depression, chronic strain and rumination (Nolen-Hoeksema, Larson, & Grayson, 1999), and this was mirrored in the qualitative data as follows: This example is less of a specific moment and more a pattern of behavior. In my previous workplace, my boss would often profile applicants based on age and gender. For example, he would only hire women for clerical/reception positions and wouldn't even consider men. There were no women in positions of authority and we were routinely asked to take managerial responsibilities without compensation. He also rarely took women seriously when they had opinions or ideas about how to better the business. I often found myself going to other men I worked with and asking them to go to him with my idea because I knew he wouldn't listen to me and was actually more likely to do the opposite. Getting credit for good ideas was irrelevant most of the time. This behavior was extremely frustrating. I worked for that company for 8 years and cared deeply about it and the people I worked with. The more I was treated like I didn't matter, the more depressed and resentful I became until I ultimately left the company. (Respondent H) My welder instructor told me that being a black woman in welding would slow me down. He refused to pass me on several welding tests although he knew that I was sufficient. I felt discrimination but when I brought it to the higher ups, nothing was done about it. I completely felt the need to appeal to someone higher and was filled with sadness and anxiety. I ended up just giving up and sat out of school for couple of years, and have not re-enrolled yet. (Respondent I) 192 Identity management under stigma communication at work: Key outcomes In this section, I discuss how identity management strategies of accepting and challenging stigma relate to key outcomes at work such as self-regulatory depletion, strain, and psychosomatic health indices, and what these relationships mean when they are contextualized in broader body of literature. Findings show that while accepting identity management strategies were expected to be theoretically less depleting, the opposite was true such that accepting stigma communication was related to higher self-regulatory depletion. Perhaps it is the case that accepting stigma communication in the moment is more depleting because a person is required to repress her true feelings, and on top of the stigmatizing event, that repression is more psychologically costly. Challenging stigma communication showcased a null effect such that challenging identity management strategies were more depleting for some participants and buffered others from depletion. However, accepting versus challenging identity management strategies were not implicated in differential prediction of psychological health indices and strain. As was described earlier, under the press of self-enhancement goals at work and when encountering stigma communication from those higher in role hierarchy, women showcase a higher willingness to endorse accepting strategies. Also, accepting identity management strategies, often perceived as appealing choices to deal with sexist behaviors in the short term, precipitate psychological depletion at work. Furthermore, the open-ended data show clear pattern of comments where women in accepting identity management strategies either reference having quit a job where they were subjected to stigma communication, or reference having continuous 193 thoughts of quitting the job where they encounter hindrance stressors 38 such as sexism and/or sexual harassment. These findings are in line with meta-analytic research that shows that individuals encountering hindrance stressors showcase (a) reduced job satisfaction, (b) lowered organizational commitment, (c) higher turnover intentions and turnover, and (d) higher incidences of withdrawal behaviors (cf., Podsakoff, LePine, and LePine, 2007). Given these findings, women who face sexist environments at work may have a perfect storm at hand – discussions in the #MeToo era often reference why women did not do speak up against the sexist behaviors at work, and both empirical and qualitative data show that speaking up is easier said than done. In the theory building effort, perceptions of a positive diversity climate were retained in the model for their possible role in helping women engage in more challenging identity management strategies, but that was not the case (discussion of these findings was presented in Chapter Four). The only constructs that offered buffering effects were individual difference constructs of self-efficacy in coping with sexism and promotion focus (but not prevention focus). Both coping self-efficacy and promotion focus were related to higher use of challenging strategies. Given these findings, it is important to explore how might organizations help employees enhance their coping self-efficacy, and how might managers and HRM professionals create environments that could possibly cue promotion focus states? These issues are addressed in practical implications later. Limitations and Future Research Directions The current research agenda was the first empirical test of identity management as a goal- 38 Hindrance stressors defined as those stressors that “constrain personal development and work- related accomplishment” (Podsakoff, LePine, & LePine, 2007; p. 438). 194 oriented, self-regulated, norm-based response to stigma communication at work. Several interesting insights and research questions emerged from the data that were described in the discussion sections for each of the studies presented earlier (see Chapter Three, Chapter Four, Chapter Five, and Chapter Six). The goal here is to present the limitations and future research directions that emerge from taking a consolidated view of the findings described so far, so the discussion has been limited to four broad points discussed below. First, sexism was used as a frame to investigate stigma communication in this research agenda. The reasons why it was necessary to bound the study of stigma communication by positioning it within sexism as a grounding framework and an operational choice have already been presented (see Chapter Two). That said, the current findings are obviously limited in scope because of this operationalization and may not generalize well to other forms of stigmatization such as race-based stigma communication, or sexual identity-based stigma communication. Furthermore, the current research agenda focused on women’s identity management in response to sexist behaviors at work. While this choice was in line with the NIH guidelines that called for more research focusing on women’s experiences exclusively and the current research contributes to the literature by addressing this call, the focus on women as sample of choice further limits the generalizability of the current findings. The generalizability issues raise interesting research questions about how different types of stigmas may unfold, and of these, stigma communication for gender-based identities (i.e., if men are included as potential targets of sexual harassment) and race-based identities are discussed next. There are reasons to expect different effects of stigma communication based on gender. The history of sexual oppression, power differentials, threat of bodily harm, and the disenfranchisement faced in organizational (e.g., women were not allowed entry into universities 195 and workforce), social (e.g., the gender norms that suggest tropes such as woman’s place is in the kitchen), and political spheres (e.g., suffragists had to fight for women’s right to vote) have together yielded rich narratives of how women have had to struggle for their basic rights for centuries. These narratives may sensitize women to sexism or sexual harassment such that mere anticipation of sexism is aptly seen as a threat to one’s self and dignity. A key question is how might men who face sexual harassment (Bergdahl, Magley, & Waldo, 1996) manage their identities? While men have been spared the specific burden of history that women continue to endure, research finds that men are not immune from the pressures of likely stigma communication. For example, ‘manhood’, compared to womanhood, is perceived as a precarious identity whereby men have to defend their turf to safeguard their ‘earned’ status as a man (cf., Weaver, Vandello, Bosson, & Burnaford, 2010). There is an ever-present danger of loss of this earned status, for example, if a man were unable to provide for his family, then culturally his manhood would be in peril. Research shows that this precariousness is related to heightened aggression in response to gender threats because manhood is defined “with qualities such as competitiveness, defensiveness, and constant struggling to publicly prove worth and status” (Bosson, Vandello, Burnaford, Weaver, & Wasti, 2009; p. 624). Given that men have their own cultural and social dictates to fight to protect their identities, simply reporting that one was subject of sexual harassment can be stigmatizing for the male identity that is wrapped up in the notions of strength and virility (cf., West & Zimmerman, 1987), power (Collinson & Hearn, 1994; Holland, Ramazanoglu, Sharpe, & Thomson, 2004), ‘self-as-protector’ (cf., Harper, 2004), and machismo (cf., Mosher, 1991). These themes coalesce in research that shows that male respondents considered shame, guilt, or embarrassment as even 196 greater barriers to reporting rape than women (Sable, Danis, Mauzy, Gallagher, 2006). A qualitative analysis of male veterans who have experienced sexual assaults in the military found that one of the most prominent reasons for men to avoid medical care were stigma-related barriers39 (Turchik, McLean, Rafie, Hoyt, Rosen, & Kimerling, 2013). Future research may investigate how men navigate the press of their goals in the face of stigma communication at work, and what types of identity management strategies might they use. Similarly, might individuals encountering race-based stigma communication likely accept or challenge stigma communication in the same way as those who encounter gender based stigma communication? Roberts, Settles, and Jellison (2008) found that when individuals perceived that their groups were devalued groups, both women and African Americans used both claiming (i.e., positive distinctiveness) and downplaying (i.e., social recategorization) strategies in managing their identities. However, this research did not study the use of identity management strategies under different goals, and furthermore, it is also unclear if the nuanced identity management strategies that were tested in the current research agenda would be applied in the same way for gender identity and racial identity groups. As an example, women facing gender- based stigma communication in the current research agenda chose self-deprecating humor when under the press of self-enhancement goals, but it is unclear if in race-based stigmatization attempts, respondents would use strategies such as self-deprecating humor or not. Theoretically, while there are some similarities, doing gender identity and racial identity work can also be different. Identity is constructed as part of one’s developmental experiences, and different constraints and opportunities in one’s environment set people up for doing 39 Stigma related barriers were defined as “veterans’ personal discomfort or internalized beliefs about seeking care for MST, and concerns about social perceptions and consequences” (Turchik et al., 2013, p. 217) 197 qualitatively different identity work in various arenas of life (e.g., Black male masculinity unfolds differently than White male masculinity; cf., Harris, 1995). Furthermore, to the extent that identity work is about telling and re-telling of stories that are descriptive of one’s experiences, one can expect that narratives women tell themselves about their gender identity and the narratives a person of color uses are likely to shape and construct both their personal (i.e., self-referential) identities and their social (i.e., discursive) identities differently (McAdams, 1996; 1999). One can also expect differences based on stigma types that are perceived as more perilous and/or controllable (Meisenbach, 2010). However, there is not much strong theory on how the different types of identities and/or stigmas may impact differential choice of identity management, especially when the role of goals at work is taken into account. I posit that empirically testing whether or not there are likely mean differences on outcomes such as levels of depletion, strain, and/or psychosomatic health indices depending on the type of stigma events under consideration for different identity groups can yield nuanced insights for both identity management and stigma literatures. Second, in the Pilot Study it emerged that the base rate of reported sexism events was low (i.e., participants reported an average of 1.2 sexist events over the course of four weeks; see Chapter Three). As such, an effort was made to screen for respondents who had indicated having faced sexism at work (for both Study One and Study Three). For Study Three, women who reported facing less than four events per month were screened out so the data collection yielded enough sexist events to ensure that statistical power to find effects was not eroded. Finally, self- selection into the studies outlined here was expected to be an issue and it was anticipated that women who faced a high degree of sexism at work might be particularly attracted to this research agenda and Study Three in particular (discussion in Chapter Six). 198 While the aforementioned choices were necessary, they also limit the current findings in several ways. Even though our data shows good variability on all measures, by expressly selecting for women who have faced sexism and/or who reported facing high levels of sexism, it is unclear how this experiential effect of navigating sexism regularly impacts the current findings. Literature on individuals who face constant stressors suggests people can become resilient against these stressors (Wright, Masten, & Narayan, 2013). Does this resilience reflect an individual’s learned strategies to navigate such stigmatizing events – a form of experiential knowledge? If so, experiential knowledge on how to navigate stigma situations may imply positive outcomes. However, research also shows that exposure to constant stressors can precipitate a host of negative outcomes (Gouin, Glaser, Malarkey, Beversdorf, & Kiecolt-Glaser, 2012). I wonder if women who work in environments where there is low frequency of sexist behaviors respond with different identity management choices? It is likely that we may see a relative difference in rank ordering of choices depending on low frequency or high frequency of sexist behaviors encountered. Is it the case that women in environments where they are faced with high frequency of sexist behaviors have in-built resources such as learned resilience over time that helps them counter sexism? Or do the women who know that sexist behaviors are part of the job, and come with the territory, fair better when confronting stigma communication? It is important to note the causality issue here. It may well be the case that those women who do confront stigma communication by challenging it, then experience less sexism, while those that accept are then met with more sexism because they did not confront the perpetrators the first time around. Third, the current work highlighted how, under conditions of stigma threat, individuals may engage in making identity management choices that rely in part, on the individual’s goals 199 during a stigma communication event. I assessed this idea by including three distinct goals of self-enhancement, self-verification, and belonging in the empirical studies presented so far. However, as DeShon and Gillespie (2005) suggest, there are many other levels of goals that could have been included in this investigation. This focus on three, higher order goals came at the cost of participants specifying their own goals – and at some level created a synthetic situation, even in the experiential sampling methodology study (Study Three). In Study Two, participants did not commit to the goals they were presented, and this created a limitation for that study. How might respondents’ choice of self-related goals that they specify on their own impact identity management choices in response to stigma communication at work? If respondents were allowed to enter their own self-related goals our analyses could have yielded more ecologically sound insights. It is also a given that people in organizations function under the press of multiple goals, and often have to select from multiple, and at times, even competing goals. Unfortunately, the goal setting research has not addressed multiple goal selection well. While a theoretical heuristic was presented that attempted to highlight how individuals may select among multiple goals when confronted with stigma communication, these ideas were not tested in the current research agenda given the sparse empirical findings in the goal setting literature. Research is needed on how individuals select from goals that may differ in their motivational structure while they concurrently process stigma communication, especially as this relates to self-regulatory depletion. Additionally, in the current research, ideas on actual goal achievement were not tested. So while women may believe that accepting stigma communication was better for their goals in the moment, I did not have the data to investigate whether or not women were accurate in their assessments. For example, it may be the case that by challenging stigma, one’s goal of belonging 200 is jeopardized, but perhaps challenging identity management strategies are not bad for other goals at work. These ideas should be tested in future research. Similarly, missing from the empirical work described herein are the ideas on goal-performance discrepancies and associated goal revision. The current research agenda was a necessary first step to provide evidence that an individual’s goals are implicated in her decision matrix of when to accept or challenge stigma communication, and that goals can and do constrain individuals in picking some identity management strategies over others. However, these goals were assigned to the participants and it would be interesting to see as next steps, how individuals choose between multiple goals when they are left to their own devices. In including identity centrality as a construct of interest, the current research agenda suffered from several limitations. The issues around the measure of identity centrality have already been outlined and will not be discussed here again (see Chapter Four). In their paper on concealable identities that may be stigmatized, Quinn and Chaudoir (2015) used four different stigma identity factors, namely, anticipated stigma, centrality, salience, and cultural stigma (defined as the extent to which different identities are devalued culturally). In their multidimensional model of racial identity (Sellers, Smith, Shelton, Rowley, & Chavous, 1998) distinguish between ‘centrality’ and ‘private regard’ and note that individuals vary on both how central they consider their identity and how positive, negative, or neutral they are towards the said identity (i.e., private regard). In hindsight, when conceptualizing the current research, the choice of identity centrality as a sole construct of interest to assess how one’s identity was related to key outcomes in stigma contexts was too narrow a conceptualization. Perhaps tangled in this issue of identity centrality are questions of identity salience, private regard, anticipated stigma, and cultural stigma. While Stryker and Serpe (1994) recommend using both salience and 201 centrality variables for a strong methodological approach (already discussed in Chapter Four), it would seem that private regard and cultural stigma for a given identity of interest are yet other constructs that should be included in investigations of identity centrality. In the current research, an unwarranted and unstated assumption was made that if a woman considers her identity as a woman central to the self, she must also hold it in high regard. How might the current findings be different if these assumptions were questioned and clarified at the outset, and/or if these variables were included in the test of the broader ideas? Future research can focus on just examining identity in its all its complexity and assessing how it relates to self-regulatory depletion under different goals. Specifically, it appears that identity centrality can have different effects based on whether or not the stigma communication under discussion is ‘enacted stigma’ or ‘anticipated stigma’. Enacted stigma is defined as “perceived experiences of discrimination from others in the past or present” while anticipated stigma is defined as “expectations of discrimination from others in the future” (Earnshaw, Lang, Lippitt, Jin, & Chaudoir, 2016). For example, Earnshaw et al. (2016) found a buffering effect of HIV identity centrality but only for anticipated stigma, and not for enacted stigma. Specifically, at low levels of HIV identity centrality, anticipated stigma had an indirect effect on HIV symptoms via stress, but this was not the case at high levels of HIV identity centrality. However, when this relationship was tested for enacted stigma (where stigma was experienced; for example ‘Healthcare workers have treated me with less respect’) the authors reported null findings. In the current research, respondents were under a state of stigma communication – that is, when recalling a sexist event (Study One) respondents reflected on a time when someone had communicated stigma to them, and when reading a sexist experimental prompt they described the team as “sexist”, and responded to the study materials under the 202 impression that they were working with this virtual team (Study Two). To the extent that the research methods of both these studies were perceived as “enacted stigma”, the null findings of identity centrality are in line with Earnshaw et al’s (2016) findings. This raises an interesting question for the identity centrality construct – while it is often theorized as an important antecedent in stigma, discrimination, and identity management literatures (Sellers and Shelton, 2003; Settles, 2004), might it be the case that it holds higher predictive ability when participants respond to “anticipated stigma” prompts versus when they respond to “enacted stigma”? Perhaps it is the case that when confronted with enacted stigmatizing behaviors, the strength of the situation acts as a cue for other more stable tendencies such as trait anxiety to express in behaviors and/or reports of distress, over and above thinking about one’s identity centrality? Future research is needed to investigate if there are indeed such differences. The current research focused on depletion, but literatures on microaggressions (Salvatore & Shelton, 2007) and stereotype threats (Schmader & John, 2003; Schmader, Johns, & Forbes, 2008) often also utilize cognitive load as an alternative focus. In fact the current data show that the correlations of trait anxiety and depletion and those between self-efficacy and depletion likely hint at the value of cognitive load explanations. Future research should explicitly investigate if this is the case. In addition, in the current work I utilized self-reports as a way to assess self-regulatory depletion. Future research should attempt to investigate depletion using alternative methodologies such as functional imaging techniques and/or using biological markers. Research using functional imaging techniques finds that individuals who are depleted tend to display more amygdala activity in response to negative events, but not to positive events and showcased failure to regulate emotions (Wagner and Heatherton, 2012). It would be 203 interesting to see how personally valent negative events such as stigmatization impact people, and whether self-reports based findings can be replicated. Future research could also look at biological markers to assess self-regulatory depletion instead of using self-report methods. For example, glucose has been implicated as a biological marker of depletion – specifically, low levels of blood glucose precipitate depleted states (Gailliot, Baumeister, DeWall, Maner, Plant, Tice, Brewer, Schmeichel, 2007). These findings have recently been questioned as statistical artifacts (Vadillo, Gold, & Osman, 2016). Nonetheless, perhaps triangulating between the various methods could help researchers better explore depletion related effects. Finally, before closing this section, one note about the use of the term ‘accepting’ stigma, a common shorthand in the literature, is merited. This shorthand is used to imply ‘accepting stigma communication’. In writing this dissertation, it was a point of discomfort for me to use ‘accepting stigma’ to describe examples of identity management strategies where an individual is forced to acquiesce to the stigma communication. Accepting stigma communication seems to imply that the target is willingly accepting stigma. However, in most cases of harassment or stigmatization of any kind, people are forced into situations where they feel compelled to acquiesce. Clearly, acquiescing is very different from accepting, and is a more accurate portrayal of the feelings of reluctant acceptance of stigma or acceptance of stigma without protest. Perhaps as a field, we need to move from our use of the phrase accepting stigma to ‘acquiescing to stigma communication’ so our terminology can better communicate the burden inherent in this strategy. Practical Implications Given the diverse samples represented across the three studies outlined in this research agenda, it is clear that while sexism is a real and present threat in many different types of 204 organizations ranging from blue collar industries to white collar, knowledge work based industries, the base rates of reporting experienced sexism were low. That said, while the frequency of enacted sexist behaviors may have been low, for negative outcomes such as self- regulatory depletion to manifest for women, all it took was for respondents to recall one incident when they had experienced sexism (in Pilot Study, and Study One), or for women to even just read about men describing a female colleague they had never met in sexist ways and anticipating that they too would encounter stigma communication from the men in their virtual team (Study Two). When women were recruited based on a higher than average incident rate of having faced sexism (i.e., women who had faced at least four sexist events in the past month; Study Three) they not only reported self-regulatory depletion, but also reported high levels of strain. So while it may be the case that reported base rates of sexist events at work are fairly low, organizational leaders and HR practitioners cannot take consolation in this number. The current findings speak to the powerful effects of gender-based stigma communication and highlight the need for HR leaders to continue their work on creating equitable work environments where they establish zero-tolerance policy for sexist behaviors. Second, individuals with stigmatized identities can directly benefit from knowing that they indeed have an active role in managing stigma communication, and that they can and in fact do choose identity management strategies that best align with their goals, motives, and needs. Traditional perspectives of identity management that impose identity centrality and self- verification motives on minorities have the potential to make minorities feel guilty about accepting stigma communication. In contrast, an active, goal- oriented perspective on stigma management offers employees the option to enact identity management behavior depending on their identity centrality, self-verification motives, and other critical goals. There may be potential 205 benefits to a more active approach to identity management where people are likely to show better affective outcomes, satisfaction, and overall health outcomes in the long run as they navigate often stressful stigma communication situations at work. Third, viewing identity management behaviors as goal-oriented, choice behaviors can help HRM professionals better understand the different behavioral strategies people may choose in navigating stigmatizing situations and how these might be tied to critical organizational outcomes. This is particularly important when diversity related interventions are designed at work – well-meaning diversity interventions can sometimes have unintended consequences if HRM professionals assume that everyone is, and should be, comfortable operating in a space of challenging stigma communication (Tavakoli, Lumley, Hijazi, Slavin-Spenny, & Parris, 2010). Such approaches can end up creating an impression that choosing to do nothing about the stigma communication is a sign of being a pushover and a sign of letting “your people” down. Data presented here show that when individuals are under the press of self-enhancement or productivity goals, they are less likely to engage in challenging stigma communication because the focus is on getting one’s task completed. While this research shows that accepting stigma communication is related to negative outcomes at work such as self-regulatory depletion, and it is disheartening that women in the current research tended to choose accepting strategies when receiving stigma communication from someone higher in the role hierarchy, or because of the press of their self-enhancement goals, HRM professionals cannot assume that simply asking women to challenge stigma communication is the best strategy to counteract these problems. While it is important to advise women to challenge stigma communication, the HRM professionals need to exercise restraint in how and when they ask women to challenge stigma communication. Specifically, women who are already feeling constrained by the contextual 206 features into accepting stigma communication may find it particularly frustrating if well- intentioned HRM professionals push for voice behaviors. In such cases, it is important to realize that the women are likely doing the best they can, and have a more nuanced understanding of their own work attributes; so attempts at diminishing their own unique expressions of managing stigmatizing situations may not be the most helpful approach. Alternative ways of helping women navigate stigma communication are offered later in this section. Fourth, given that women were more likely to accept sexism when stigma was communicated by supervisors and/or clients, organizational leaders will have to do a lot more work at this end to ensure that people in positions of power in an organizational setting are current on their sexual harassment trainings, understand the crucial role they play in impacting climate against sexism at work, and understand how sexism impacts their employees in a negative manner. I am uncertain on how clients could be coached – but if an organization is embedded in business-to-business commitments, then perhaps instituting policies that all participating organizations have sexual harassment policies in place would be helpful. Alternatively, organizations could institute policies whereby clients may be denied service (e.g., some restaurants and airlines stipulate that customers who harass staff are not welcomed). The current work also integrated contextual features such as norms, and individuals’ goals, motives, and needs to shed light on when and why certain identity management strategies may unfold, and in doing so, it also offers HRM practitioners and leaders a more comprehensive lens through which to view stigma communication, its effects, and how it can be managed by minority employees who may face stigmatizing situations. While not all of these relationships were statistically significant in the current research agenda, these constructs were carefully chosen for inclusion in the theoretical model based on prior research. So when designing policy and/or 207 interventions, organizational leaders would do well to at least keep an eye on these elements in their problem definition phase, and embed markers so the correct levers may be used to solve for stigma communication related problems at work. Finally, data in the current research show that both self-efficacy for coping with stigma communication and promotion foci aided in helping women challenge stigma communication, which was related to better outcomes on the job (lowered depletion and strain). In light of these findings, two important questions for HRM leaders to consider would be how they could (a) enhance coping self-efficacy for women, and (b) cue promotion focused states at work? Creating training materials that coach potential targets of stigma communication on how to navigate different forms of stigma are recommended here. As was done in the training materials for the current research agenda, the learning materials can be easily catered to different individuals by treating the different identity management strategies as metacognitive prompts instead of hard recommendations. That is, by having people engage in different scenarios, vivid recalls, visualizing potential future scenarios and combining these with the various identity management strategies, trainees could be prepared to deal with the negative outcomes of stigma communication. The training materials are included here as a reference (see Appendix H). Conclusion In the current research agenda, I incorporated a self-regulatory perspective in thinking about an individual’s decision to accept or challenge stigma communication, and making an informed, goal-oriented choice in selecting from different identity management strategies. In integrating a self-regulatory perspective (a) the deleterious role of assumptions of identity centrality and self-verification motives was highlighted and challenged, (b) an attempt was made to open up the black box of identity management as it relates to self-regulation, (c) the critical 208 role played by organizational constraints in identity management was underscored, and (d) new research questions that have the potential to energize the scholarship on identity management research were presented. 209 APPENDICES 210 APPENDIX A: RATIONALE Rationale for the choice of self-regulatory depletion tasks for the Pilot Study One of the main foci of Hagger et al’s (2010) meta-analysis was to identify whether or not the effects of ego-depletion varied as a function of the spheres of depletion and the different tasks used to assess ego-depletion. As mentioned earlier, the ego depletion theory posits that since self-regulatory resources are drawn from one common source, self-regulating on prior tasks should lead to depletion of self-regulatory resources (i.e., ego depletion) on subsequent tasks irrespective of the domain or sphere of depletion 40 that tasks may tap into (Baumeister & Vohs, 2007). Hagger et al’s (2010) meta- analysis shows that this is indeed the case. Specifically, while the effect sizes for some spheres of depletion were higher than others, a generally consistent ego-depletion effect is seen across both depleting and dependent tasks. Additionally, authors also investigated if type of tasks used, the complexity of tasks used, and the duration of presenting the tasks influenced differential ego depletion effects. The key findings from the meta-analytic review are presented below (see Table 28). Note depleting tasks are those that are used to deplete an individual’s self-regulatory resources. When tasks are used as dependent variables they are used to assess the extent of self-regulatory depletion. While the same task could be used as both depleting task and dependent task, there are some tasks that perform better as one or the other. For example, the Stroop task tends to work best as the dependent task, but is 40 Spheres of depletion refer to how differential ego depletion tasks are designed to ensure that participants either have to control their emotions, thoughts, impulses, or their more general cognitive and social processing in response to experimental stimuli. Examples of spheres of depletion range from controlling thoughts, managing emotions, overcoming unwanted impulses, fixing attention, making many choices, guiding behavior, social and/or interpersonal responses such as self-presentation or impression management, among others (cf., Baumeister & Vohs, 2007; p. 353). Since self-regulatory reserves are considered a general, but finite resource, one should expect that regardless of the sphere of depletion implicated, people should show ego depletion effects. 211 not that effective as a depletion task. Table 28: Meta-analytic findings to inform task choice in the current research. Effect sizes and CI when used as depletion task Effect size and CI when used as Dependent task d = .62 CI95 [.50, .74] d = .71 CI95 [.64, .78] d = .63 CI95 [.53, .75] NA NA d = .55 CI95 [.46, .64 d = .71 CI95 [.64, .78] d = .65 CI95 [.50, .81] NA NA d = .82 CI95 [.62, 1.00] d = .22 CI95 [.11, .33] d = .54 CI95 [.29, .80] d = .60 CI95 [.52, .67] d = .75 CI95 [.65, .84] d = .69 CI95 [.60, .79] 212 Types of tasks used Study (k) Subtypes of Sphere of depletion tasks Sphere of depletion tasks Controlling emotions Controlling thoughts Controlling impulses Controlling attention Choice and volition Cognitive processing Social processing 24 25 61 21 12 13 38 20 19 19 19 13 18 10 d = .77 CI95 [.65, .90] NA NA d = .65 CI95 [.52, .78] NA NA d = .55 CI95 [.42, .68] NA NA d = .61 CI95 [.48, .74] NA NA d = .40 CI95 [.26, .55] d = .76 CI95 [.59, .94] NA NA d = .64 CI95 [.45, .84] NA NA d = .60 CI95 [.44, .76] 10 NA NA d = .50 CI95 [.31, .69] Table 28 (cont’d) Crossing out letters task Frequently used depleting tasks Wegner’s white bear paradigm Video watching affect regulation Video watching attention control Modified Stroop task Hand grip task Solvable anagram task Math or mental arithmetic task (Note: Extracted from Hagger et al., 2010). In addition, authors also tested to see if the duration of the task, interim activity, task difficulty, or use of control tasks had an impact on the observed ego-depletion effect. The duration of the task predicted ego depletion effect size (β = .11, z = 1.79, p = .07), but the effect is small and marginal (k = 148 studies). For the studies where participants completed questionnaires, filler tasks, or had a break between tasks the effect sizes (a) did not vary across the different conditions, and (b) were higher as compared to studies where participants were 213 given no interim period (dinterim = .71; dno-interim = .47). When compared to simpler tasks, complex tasks had higher effects sizes (dcomplex = .65, CI [.54, .75]; and dsimple = .35, CI [.15, .55]). Finally, no differences in effect sizes were found when studies used an easier version of the depletion task as control, or when studies used an alternative task as control task. Based on the effect sizes and confidence intervals reported in the meta-analysis (cf., Hagger et al., 2010), and based on the practical constraints (i.e., to gather data across a combination of online and panel studies, and to ensure that the tasks can be readily combined with sexist stigma communication manipulations) the current research will use tasks that do not require participants to spend a long time working on the overall experiment, and that also do not necessitate an experimenter in the room. While Wegner’s white bear task has the highest effect size, that task cannot be combined with a stigma communication framework. For both video watching affect tasks and video watching attention tasks the effect sizes range from .55 to .61 respectively. Bair and Steele’s (2010) video based cognitive depletion task maps on well to the former, has been successfully implemented in self-regulatory depletion studies, and was thus considered as one of the experimental paradigms for current research. This task requires participants to watch a racist episode unfold on video, and then a depletion dependent task is administered to assess the extent to which watching stigma communication unfold impacts participants’ self-regulatory depletion. Since the task is a video-based prompt the manipulation is standardized and can be easily delivered to participants in online studies. However, one disadvantage of using this task is that it is not self-referential – that is, it is possible that watching stigma communication unfold towards a stranger on a video might elicit a lower self-regulatory depletion than if the stigma communication was self-referential. Another concern with this paradigm is that while the theory presented in the current research takes an event-based view of 214 stigma communication, Bair and Steele’s (2010) task would eschew that in favor of tapping into participants’ general/global stigma and/or justice perceptions. Given the issues with Bair and Steele’s (2010) task, I also evaluated Inzchlit and Kang’s (2010) vivid recall task for its suitability for the current research. In the identity threat condition, participants were instructed to vividly describe “a time when they felt discriminated against because of their group membership in any one of the five categories” and to “recreate the experience as fully as possible by thinking about all the thoughts, feelings, and sensations they had at the time of the experience. They were asked to think about how they coped and how they felt, both physically and emotionally” (p. 475). The authors included gender, age, sexual orientation, ethnicity, and religion as the five categories. One disadvantage of this task is its novelty; the Hagger et al. (2010) meta-analysis did not include use of vivid recall tasks. Having said that, such a task would tap into multiple spheres of depletion (e.g., controlling emotions, thoughts, impulses, and cognitive and social processing) and would therefore be considered a complex task. As mentioned earlier, the Hagger et al. (2010) meta-analysis does show that complex tasks yield stronger effect sizes than the simpler tasks. In addition, Inzchlit and Kang (2010) used the task successfully (found statistically significant effects for religion, race/ethnicity, but not gender41). The theoretical underpinnings of why vivid recall of a 41 Inzchlit and Kang’s (2010) inability to find self-regulatory depletion based on gender can perhaps be ascribed to the higher salience of race/ethnicity in both their pilot (17 Black participants from a Canadian university) and study sample (highly diverse sample: 36.1% South Asian, 33.1% East Asian, 11% White, 8.4% Black, 5% Middle Eastern, 1.7% Latino, 1.7% Biracial, 4.2% Other). Moreover, participants completed this study within a classroom setting, so it is likely that depending on the framing of the class, they attached a higher significance/salience to the race/ethnicity based narratives. Authors’ reason for the non-finding was that participants only had five minutes to write a detailed and expressive narrative of the stigmatizing experience, thus there was insufficient content to explore any differences. However, research shows that women, in comparison to men, are more likely to include negative events directed at them as 215 stigmatizing situation could elicit similar self-regulatory depletion as when an individual is in a stigmatizing situation have been presented elsewhere (cf., Ackerman, Goldstein, Shapiro, & Bargh, 2009). Several other studies on social exclusion have used the vivid recall paradigm successfully (cf., Crescioni & Baumeister, 2009; DeWall, 2010; DeWall & Baumeister, 2006; DeWall, Baumeister, Mead, & Vohs, 2011). In addition to being easy to administer in online settings, this task is also better aligned42 with an event-based view of stigma communication, and more importantly, this task is inherently self-referential. Both the Bair and Steele’s (2010) task and Inzchlit and Kang’s (2010) tasks present with pros and cons. In my opinion, the Inzchlit and Kang’s (2010) task, for its complexity (maps on to several spheres of depletion) and its self- referential nature, may elicit higher effect sizes and thus gives me confidence that it is worth piloting. For self-regulatory depletion as a dependent variable, the solvable anagram task was piloted given its good effect sizes (see Table 28), ease of administration, but most importantly because this task was closest to the type of cognitive work people do in work contexts. central to their identity, and also present more vivid, emotionally intense, and visceral emotional reactions as outcomes to including negative events in their self-concepts (Boals, 2010). 42 A reader could object to this view and state that the retrospective nature of the task cannot be equated with stigma communication in the moment. This criticism would be correct, but in the absence of actually stigmatizing someone, experimental approaches on stigma communication tend to rely on indirect measures. In addition, I am conducting an ESM study that will allow me to examine stigma communication in a more proximal manner. 216 APPENDIX B: MAPPING Based on the mapping, challenging identity management strategies were assessed by asking participants to select the behavior or behaviors that describe how they dealt with sexism in the stigma communication event they described on a scale from never (1) to very often (5). Table 29: Literature review on how women may manage their identities. Author(s) Study context Women’s Identity Management Strategies Used • Integrating identity management strategies* • Did not use avoiding identity management strategies • Favorable social comparisons • Blaming the stigma* (as in the case of deemphasizing quantitative work) • Bolstering/refocusing* and transcendence strategies* (as in the case of focusing on emphasizing verbal skills – that is, focusing on the positive aspects of the stigmatized identity) Chrobot- Mason, Button, & DiClementi (2002) Mussweiler, Gabriel, & Bodenhausen (2000) Shih, Young, & Bucher (2003) LGBT identities; but authors also conducted an analysis for male and female respondents and their use of different identity management strategies when working in open group process (defined as “the extent to which group members express their views and include all members in decision making”; p. 326). The study found that when women received negative evaluations compared to other women in the sample, women chose to reframe their identities by focusing on their ethnic identity instead of referencing their gender. This allowed for a self-esteem protective effect because the low scoring women were now able to cast the high-scorers as inappropriate comparison standards. In their review, authors identify how women engage in stereotype reassociation by deemphasizing quantitative work and emphasizing soft skills and verbal abilities in work contexts 217 • Displaying stigma • No bonding with other stigmatized individuals • Making favorable social comparisons • Making favorable social comparisons • Bolster/refocus* • Transcendence* • Positive distinctiveness Table 29 (cont’d) Halbert (1997) Pronin, Steele, & Ross (2004) Rabe-Hemp (2008) Raman (2008) This qualitative study examined how female professional boxers engage in identity management both inside and outside the rink. The identity management strategies reported were “wear feminine uniforms…do not associate with stereotyped people; and emphasize feminine characteristics of appearance such as long hair, makeup, and feminine clothing in public appearances outside the ring” (p. 27). When women who viewed themselves as competent at math confronted negative stereotypes about women’s quantitative abilities, they engaged in selective disidentification or “bifurcation of identity” (p. 164). Specifically, in a series of three experiments, the authors find that the women disavowed the stereotypically feminine attributes (e.g., wanting to have children, being flirtatious), but maintained their identification with the warmth-component of the female identity (e.g., being nurturing and empathic) This qualitative study examined how female police officers manage their identities in a macho, male-dominated police culture. Identity management strategies used ranged from “distancing themselves from female group membership and aligning themselves with the dominant police culture”, “actively resisting the expectation that they be the “pansy police” by emphasizing masculine aspects of policing”, while at the same time they also “adopted expectations for caretaking, softness, and empathy”, “stressed communication, familiarity, and building trust and rapport between the police and community members” (p. 12) This unpublished dissertation examined how women’s consumption of televised portrayals of women related to their social identity management strategies. The dissertation is based on a social identity perspective and thus examined three key strategies: mobility, social creativity, and social competition. 218 • Bolster/refocus* • Bonding with others • Avoid situation* • Acceptance • Bolster/refocus* • Isolate* This study examined the preferred coping strategies used by women in STEM fields to navigate the perceived gender bias. Of the 24 items (Ways of Coping scale; Folkman & Lazarus, 1986), participants tended to endorse active coping strategies over passive coping strategies. For example, working harder, and seeking social support were most endorsed. In contrast, acceptance of the gender bias and behavioral disengagement were least endorsed. This qualitative study examined the identity management strategies of seafaring women. Cargo ships often have highly masculine cultures, and women officers reported engaging in differential identity management strategies: • Women who negotiated their identities vis-à- vis the male identity at sea reported being ‘obscurers of femininity, and reinforcers of masculinity’ – that is they tried to fit into the male world as much as possible. • Women who were focused on crafting their identities to be accepted as crew-members focused overwhelmingly on being ‘acquirers of masculinity’ – that is, they tried to behave like men. • Women who sought to keep their identities tended to use disguising strategies to hide their true selves by isolating themselves from the men, or they kept their ‘tom-boyish’ identities. • Finally, the author describes women engaged in strategies such as using their rank/status on the ship to be their true selves, or those who endorsed a gender-neutral view in defining their self-concept. Table 29 (cont’d) Robnett (2014) Kitada (2013) (Note * relates to strategies that have so far not been included in any of the measures on identity management selected). Accepting identity management strategies. Examples of behavioral strategies from Moody-Ayers et al. (2005) measure that map on to the acceptance identity management strategies included ‘accepting it’, ‘ignoring it’, ‘avoiding it’, ‘keeping it to yourself’. One item 219 for Meisenbach’s (2010) accepting strategy of recategorizing has been extracted from Roberts, Settles, and Jellison’s (2008) recategorization scale. For the remaining two strategies, that is ‘bonding with others’ and ‘self-deprecating’ strategies, that were included in Meisenbach’s (2010) review and were deemed important for the current work based on the literature review presented here, I have written two new items which are shared in Study One materials (see full measure in Appendix E). Challenging identity management strategies. Examples of behavioral strategies from Moody-Ayers et al. (2005) measure that map on to the challenge identity management strategies include ‘speaking up’, ‘trying to change things’, and ‘working harder to prove them wrong’. However, the measure does not include Meisenbach’s (2010) challenging strategies of making favorable social comparisons, using simple and logical denials, and educating others identity management strategies. Items from Roberts, Setttles, and Jellison’s (2008) measure on positive distinctiveness strategies scale were used to assess the logical denials and educating others strategies. For the remaining two strategies, that is making favorable social comparisons and simple denials, that were included in Meisenbach’s (2010) review that were deemed important for the current work based on the literature review presented earlier, I have written two new items that are shared in Study One materials (see full measure in Appendix E). Mapping out the Identity Management Strategies measure. (Moody-Ayers et al. (2005)43 43 Moody-Ayers et al’s (2005) items also included ‘Praying’, ‘Getting violent’, and ‘Forgetting it’ which were not included because they were unrelated to identity management strategies. 220 Table 30: Identity Management Strategies items used in the current study. Author(s) Item Accepting Strategies Challenging Strategies 1. Speaking up 2. Accepting it 3. Ignoring it 4. Trying to change things Mapping to Specific Identity Management Strategies from Meisenbach (2010) in the Current Conceptualization Positive distinctiveness (try to be an advocate)* Passive acceptance Passive acceptance Positive distinctiveness (try to be an advocate)* Isolate self * Bolster/Refocus* Avoid situation* Recategorizing Educate others Using logical denials Bonding with others Self-deprecation Make favorable social comparisons Using simple denials 14 7 7 - Moody- Ayers et al (2005) Roberts, Settles, Jellison (2008) Items created for the study to align with selected IM strategies Number of items 5. Keeping it to myself 6. Working harder to prove them wrong 7. Avoiding it 8. Trying to conduct myself in manner inconsistent with the stereotypes of being a woman 9. Trying to educate my colleagues about the accomplishments of other women 10. Trying to communicate the inaccuracy of stereotypes about my gender 11. Trying to bond with others like me 12. Trying to use self- deprecating humor to diffuse the situation 13. Trying to make favorable social comparisons that make women look better than men 14. Denying that there is stigma on me because of being a woman ✔ - ✔ - ✔ - - ✔ ✔ - - ✔ ✔ - ✔ ✔ ✔ ✔ ✔ - - ✔ ✔ - - 221 APPENDIX C: EXPERIMENTAL PROCESS FLOWS Experimental process flows for each study are presented in this appendix. Pilot Study. Participants were emailed an individual survey link via Qualtrics. The survey took about XX minutes to complete and is designed as follows: after consenting to participate in the study, participants will be directed to a page of instructions about the vivid recall task. Since this is an online study, effective instructions were needed to ensure participants take the content seriously and are motivated to work through the materials. As such the instructions focused on the following: 1. Read ALL the instructions carefully – throughout the experiment there are attention checks that you need to respond to, and if a pattern of careless responding is documented then you may not get full credit for participation. 2. The task presented here is XX minutes long. You need a quiet room to be able to reflect and write a narrative story about the situation we have asked you to think about. The quality of the narratives you write will drive the effectiveness of the research so please follow the instructions closely: a. Screening question: Can you describe a specific event where someone at work was sexist towards you? Participants will be exited from the survey if they say “No”; if they say “Yes”, the following instructions will appear on the next screen: i. Please write about a time when you experienced discrimination at work because of your gender (i.e., when someone was sexist towards you). Note we are asking you to think about a specific event where someone was sexist towards you, not your working conditions in general. 222 b. As you think about this event, please write all the thoughts, feelings, and sensations you had at the time of the experience. To structure your story-telling, please describe the following: (a) details on who was sexist towards you and your relationship with that person (i.e., was the person a peer or co-worker, a supervisor, a subordinate, a customer, or someone else affiliated with your organization), (b) how you felt when confronted with sexism by describing your physical, emotional, and psychological reactions, and (c) how you coped with the stigmatizing event. c. You have fifteen minutes to write your narrative. Please ensure that your responses are at least 600-800 words (i.e., a full single spaced page). 3. After the reflection you will be asked to work on a task. As with the previous instructions, your motivated and effortful performance is of utmost importance to yield meaningful data. 223 Study 1. Participants were emailed an individual survey link via Qualtrics. The survey takes about XX minutes to complete and is designed as follows: after consenting to participate in the study, participants will be directed to a page of instructions about the vivid recall task. Since this is an online study, effective instructions are needed to ensure participants take the content seriously and are motivated to work through the materials. As such the instructions will focus on the following: 2. Read ALL the instructions carefully – throughout the experiment there are attention checks that you need to respond to, and if a pattern of careless responding is documented then you may not get full credit for participation. 3. The task presented here is XX minutes long. You need a quiet room to be able to reflect and write a narrative story about the situation we have asked you to think about. The quality of the narratives you write will drive the effectiveness of the research so please follow the instructions closely: a. Please write about a time when you experienced discrimination at your current work because of your gender (i.e., when someone was sexist towards you). Note we are asking you to think about a specific event where someone was sexist towards you, not your working conditions in general. i. Screening question: Can you describe a specific event where someone was sexist towards you? Participants will be exited from the survey if they say “No”; if they say “Yes”, the following instructions will appear on the next screen: b. As you think about this event, please write all the thoughts, feelings, and sensations you had at the time of the experience. To structure your story-telling, 224 please describe the following: (a) details on who was sexist towards you and your relationship with that person (i.e., was the person a peer or co-worker, a supervisor, a subordinate, a customer, or someone else affiliated with your organization), and (c) how you felt when confronted with sexism by describing your physical, emotional, and psychological reactions. c. You have fifteen minutes to write your narrative. Please ensure that your responses are at least 600-800 words (i.e., a full single spaced page). d. Please select from the multiple behavioral options below. These behavioral strategies have been described as typical behaviors most people use to address stigmatizing situations such as the one you just narrated. If you are unable to find behaviors that you engaged in, you will then be given the opportunity to share how you coped with the situation. 4. You are now asked to select the behaviors that you used in navigating the stigmatizing situation you just wrote about. All of these are viable strategies, many women tend to report using these behaviors to navigate similar situations such as the one you have described. We are asking you to select from the two overarching options presented first. Once you have made that choice, you will be asked to rank order some examples of these behaviors. That is, drag the behavior that you are most likely to do as your top choice and then fill in with the second best option to you in the second spot, and so on and so forth until you have exhausted ranking all the behaviors. a. Set-up: Participants will be shown a choice of two bins of identity management choices – they will see a section with accepting identity management strategies listed and on the same screen they will be given all the options from challenging 225 identity management strategies. They will be instructed to pick ONE bin of strategies that they used in navigating the situation. Once they pick between these two options, they will then be asked to rank order the individual strategies in descending order of perceived effectiveness in managing the situation that they had just recalled. b. An open ended field will be included to give participants the opportunity to speak their mind in case they had handled the situation very differently from the strategies that I populated for them. 5. [Next, they worked on the anagram task]. a. Instructions: Next you will be asked to work on a task. As with the previous instructions, your motivated and effortful performance is of utmost importance to yield meaningful data. 226 Study 2. Participants will be emailed an individual survey link via Qualtrics. The survey takes about XX minutes to complete and is designed as follows: after the instruction page, participants will be directed to fill the psychological measures. They will then be randomly assigned in the differential goal conditions. Next, they will be directed to the manipulation, and manipulation items will be administered. Finally, participants will work on the self-regulatory task. The instructions will be as follows: “Given what you know about the internal team dynamics and norms, please share the strategies you would use to achieve your goal to [Insert goal self-enhance/belong/self- verify/work (no goal)] with Connor, Michael, Zach, and Dave. Note if there are alternative strategies that you would like to use than the ones listed here, you will get a chance to answer them in an open-ended answer at the end.” 1. Given what you know about the internal team dynamics and norms, please share the strategies you would use to achieve your goal to [Insert goal self-enhance/belong/self- verify/work (no goal)] with Connor, Michael, Zach, and Dave. Note if there are alternative strategies that you would like to use than the ones listed here, you will get a chance to answer them in an open-ended answer at the end. 2. After reading the team dynamics and norms you will be presented with a selection of multiple behavioral options that have been described as typical things most people do when working in virtual teams with such dynamics. You must pick one bin of strategies that you think would be most helpful in achieving your goal and navigating the unique dynamics of your virtual team. We are asking you to select from the two overarching options presented first. Once you have made that choice, you will be presented with several different behavioral options that are classified as being the same category as the 227 overall option presented first. Tell us which of these options you would be most comfortable in using. 3. After having selected your behavioral strategies, you will be asked to work on the task for your virtual team. As with the previous instructions, for the experiment to yield meaningful data your motivated and effortful performance is of utmost importance. 228 APPENDIX D: RATIONALE FOR SCALE Rationale for choosing the unpublished State Self-control Capacity Scale (SSCCS). Although this is as yet an unpublished measure44, so far 38 different studies have used the measure to assess resource depletion. The measure has been used primarily in social psychology experiments, and authors have found group differences based on this measure in a wide variety of experiments. Unfortunately, most of these effects have been tested using ANOVAS and in a lot of the studies parametric information has not been shared (cf., Barber, Barnes, & Carlson, 2013; Becker, 2008; Boom, 2009; Janssen, Fennis, Pruyn, & Vohs, 2008; Jany, 2009; Smolders & Kort, 2014; Salmon, Adriaanse, DeVet, Ferris, & DeRidder, 2014). As such, I have detailed a subset of the studies for the reader where both reliability and validity information was provided (see Table 30) to provide my rationale for using this unpublished measure to assess self-regulatory depletion as a dependent variable. In line with the literature, in the current sample, internal consistency for the SSCCS wasα= .95. 44 To my knowledge, there is one other published measure of self-regulatory depletion in addition to the unpublished measure used in the current research. At first glance, the self- regulatory fatigue measure (SRF-18; Nes, Ehlers, Whipple, & Vincent, 2013) seems like a viable option for the current study – the item content overlaps with the Ciarocco et al’s (2016) State Self-Control Capacity Scale (SSCCS) and the SRF-18 shows adequate psychometric properties from a reliability standpoint. However, SRF-18’s factor structure suggests that the behavioral dimension of the scale has subpar internal consistency and low factor loadings (see Table 31). The SRF-18 measure has also not been used as widely as the SSCCS (only ten studies have cited it; Wang, Tao, Fan, Gao, and Wei (2015) used it to assess how chronically depleted individuals manage their goals in an ESM study and found that people with high levels of self-regulatory fatigue showed decrements in goal adherence). Perhaps a bigger concern is that the SRF-18 measure was created to assess self-regulatory fatigue for patients suffering from chronic pain in clinical settings and the psychometric properties were established using a sample of patients who suffered from fibromyalgia, chronic fatigue syndrome, or both. Given the issues raised so far, the SSCCS stands out as a better measure for operationalizing self-regulatory depletion because it does not assume chronic self-regulatory depletion, was not created specifically for clinical settings, and has been more widely used and validated. 229 Table 31: Reported reliability and validity information for the unpublished SSCCS. Authors who have used the SSCCS Reported validity coefficients (correlation coefficient r) • • • • • • • • • • • • • • • • • • • • • • • r = .41* (Insomnia) r = .49 (Sleep deprivation) r = .17 (Deceptive behaviors) r = .39 (Negative affectivity) r = .29 (Undermining victimization) r = -.26 (Interpersonal justice) r = .22 (Moral disengagement) r = -.10 (Moral identity) r = .18 (Social undermining) r = .42** (Impression management) r = -.39** (Abusive supervision) r = -.24** (Supervisor coercive power) r = .11* (Items skipped on questionnaire) r = .28** (High goals) r = -.24** (Task performance: number of matrices solved correctly) r = .14* (Cheating on reporting the number of matrices solved) r = .24 (Gender; women reported higher depletion) r = .62* (Effortful control) r = .50* (Inhibitory control) r = .57* (Attentional control) r = .44* (Activation control) r =.22* (Leisure metabolic equivalents mins/week) r = -.22* (Depressive symptoms) scale Barber, Barnes, and Carlson, 2013 Welsh, Ellis, Christian, and Mai, 2014 Lee, Kim, Bhave, and Duffy, 2016 Reported reliability (Cronbach’s α) .96 .87 .85 (short-form; 10 items) Lian, Brown, Ferris, Liang, Keeping, and Morrison, 2014 Gibson (2016) Welsh (2014) .94 .82 (short-form) .90 – check if this is correct Padin (2016) .94 230 Table 31 (cont’d) Lin (2013) .82 (4-item measure) • • • • • • • • • • • r = -.29** (Age) r = .19* (Manage) r = .23** (Intrusions at time 1) r = .20* (Intrusions at time 2) r = .44** (Intrusions at time 3) r = .80** (Fatigue) r = .71** (Cognitive failure) r =.68** (Memory) r = .66** (Attention) r = .71** (Behavior) r = -.42** (Performance) A comparison of SRF-18 and the SSCCS scales’ psychometric properties: Table 32: Comparing the SRF-18 and the SSCCS scales. Nes et al. (2013): Self- regulation Fatigue Scale items (scored on 1- 5 Likert)45 Ciarocco et al. (2016): State Self- Control Capacity Scale items (scored on not true (1) to very true (7) Likert type scale) 1. I feel full of energy 2. It’s easy for me to set goals 3. I find it difficult to exercise as much as I should 4. I have urges to hit, throw, break, or smash things 5. I have no trouble making decisions 6. I experience repeated unpleasant thoughts 7. I get easily upset 8. I try not to talk or think about things that bother me 9. I never feel like yelling, swearing, or shouting 1. I feel mentally exhausted (R) 2. Right now, it would take a lot of effort for me to concentrate on something (R) 3. I need something pleasant to make me feel better (R) 4. I feel motivated 5. If I were given a difficult task right now, I would give up easily (R) 6. I feel drained (R) 7. I have lots of energy 8. I feel worn out (R) 9. If I were tempted by something right now, it would be very difficult to resist (R) Comparing psychometric properties of the SRF-18 and the SSCCS Reliability for the SRF-18: α = .81 • Emotional subscale: α = .75 • Cognitive subscale: α = .62 • Behavioral subscale46: α = .56 Reliability for the SSCCS: α = .94 45 Items in the SRF-18 and the SSCCS have been presented “as is” to showcase a comparison of the psychometric properties. However, slight adaptations have been made to the items in both the SRF-18 and the SSCCS that were shared earlier to ensure that the items capture respondent’s states and not their general tendencies of self-regulation. 46 Please note that for the SRF-18, I am not looking at the subscale level. 231 Table 32 (cont’d) 10. I handle stress well 11. I experience uncontrollable temper outbursts 12. I can easily keep up with my friendships and relationships 13. I cry easily 14. I have difficulties remembering things 15. I find it easy to stick to a healthy diet 16. I feel moody 17. I have urges to beat, injure, or harm someone 18. I rarely get frustrated 10. I would want to quit any difficult task I was given (R) 11. I feel calm and rational 12. I can’t absorb any more information (R) 13. I feel lazy (R) 14. Right now I would find it difficult to plan ahead (R) 15. I feel sharp and focused 16. I want to give up (R) 17. This would be a good time for me to make an important decision 18. I feel like my willpower is gone (R) 19. My mind feels unfocused right now (R) 20. I feel ready to concentrate 21. My mental energy is running low (R) 22. A new challenge would appeal to me right now 23. I wish I could just relax for a while (R) 24. I am having a hard time controlling my urges (R) 25. I feel discouraged (R) 232 APPENDIX E: MEASURES USED Measures used in the current research are presented in this appendix in the order in which they appear in each study. Also included are instructions and response scales, and if any adaptations to the original measures were made, these adaptations are also described here. Please note, while several measures repeat across different studies, I have chosen to include them again in each study to help the reader easily track the various measures. PILOT STUDY Manipulation Check Items. The first10-items are extracted from the PANAS (Watson, Clark, and Tellegen, 1988); I’ve added items from 11 - 13 Rating scale: (1 = very slightly or not at all, 2 = a little, 3 = moderately, 4 = quite a bit, 5 = extremely) Instructions: Do you currently feel: Dependent Variable Task: The Unsolvable Anagram Task (Smith, 2002). Instructions: Below is a list of words, the letters of which have been re-arranged. Please re- arrange the letters so that all of them are used when spelling a word. The words get progressively more difficult. Please continue until you have completed all of the words or until you are tired. 1. Distressed 2. Upset 3. Guilty 4. Scared 5. Hostile 6. Irritable 7. Ashamed 8. Nervous 9. Jittery 10. Afraid 11. Stigmatized 12. None of the above 13. [Open-ended item] 1. E N C Q S E U 2. S C W I N O S N 3. T N E R Y S U V I 4. S R C N E T E I E P 5. T R N E G E R I N W 6. S C E I Y Z H O H T P 7. S I T N L M I X I E S E 8. C R I T A I R L A S A H 9. N D E F I T A T F I D E R 233 Instructions: Rate the following on a scale of 1 = strongly disagree to 5 = strongly agree (Items below are extracted from Converse & DeShon, 2009) 1. I exerted a lot of effort during the word puzzle task 2. I felt frustrated during the word puzzle task 3. I had to exert self-control during the word puzzle task Control Variables State-Trait Anxiety Inventory, Form Y (Spielberger, 1983). Instructions: Rate the following to the extent that it is generally true for you Rating scale: 1 = strongly disagree to 4 = strongly agree 1. I am happy 2. I am content 3. I feel satisfied with myself 4. I feel pleasant 5. I feel secure 6. I lack self-confidence 7. I feel inadequate 8. I feel like a failure 9. I am a steady person 10. I wish I could be as happy as others seem to be 11. I make decisions easily 12. I am ‘calm, cool, and collected’ 13. I feel rested 14. Some unimportant thought runs through my mind and bothers me 15. I worry too much over something that really doesn’t matter 16. I get in a state of tension or turmoil as I think over my recent concerns and interests 17. I have disturbing thoughts 18. I take disappointments so keenly that I can’t put them out of my mind 19. I feel that difficulties are piling up so that I can’t overcome them 20. I feel nervous and restless Conscientiousness (Big Five Inventory; John, Donahue, & Kentel, 1991) Instructions: Here are a number of characteristics that may or may not apply to you. Please rate the following statements to indicate the extent to which you agree or disagree with that statement. Rating Scale: 1 = Disagree strongly to 5 = Agree strongly 234 Retrieved from John, Naumann, and Soto (2008) Note. Only Conscientiousness items will be used for the study; others are for exploratory purposes only. Demographic Variables Please report the following: 1. Age [open-ended item] 2. Gender [open-ended item] 3. How many years have you been employed in total (i.e., across all the organizations you have worked for) [open-ended item] 4. How many years have you worked in your current organization? [open-ended item] 5. What is your current job title? 6. Race-ethnicity: a. White b. Hispanic or Latino c. Black or African American d. Native American or American Indian e. Asian/Pacific Islander f. Other Cover story for performance discussion task. Participantss were told the following: “We want to understand how people interact with their virtual team members and how they achieve their goals to perform on shared team tasks. In the upcoming pages, we will tell you about the goal we want you to try and achieve as you work on a word-scrambling task with your virtual team. After informing you about your goal, we will then introduce you to your virtual team members 235 who will also be working on the same task online. As a team member, your job is to ensure that you follow the goal you’ve been assigned, and that you do your best to assist your team in solving the word-scrambling task. The instructions for the word scrambling task will be presented at a later time.” Piloting Varying Degrees of Manipulations for Study Two Manipulation Version A – The Original Version Performance Discussion Notes for Stephanie Capelli Participants: Connor (Team Lead), Michael, Zach, and Dave Dated: 06/12/2017 Connor: I am collecting feedback for Stephanie’s upcoming performance review. Feel free to be absolutely honest in discussing her strengths and weaknesses with me – don’t hold back, and say whatever is on your mind. I will obviously share a cleaned-up version of the notes with her, and the feedback will be shared anonymously. So, what do you think of Stephanie’s work so far? Michael: I think she is okay to work with, but not the brightest quantitatively speaking. Zach: Yes, but that’s typical – I mean, we know that women struggle with even the most basic math. I have to admit sometimes I find myself wishing we could just hire men for these roles. Dave: …We would be in legal trouble then Zach. Anyway, I thought she was okay after I coached her on the models. Zach: My biggest problem with her is her inability to take a joke – like seriously, we like goofing around and making jokes…I hated how she went crying to you Connor over that joke I shared with the whole team! I still cannot believe she called me sexist. Connor: Well you know the HR policy Zach, and we have been advised before to take it easy on the girls...but between us, I agree, Stephanie knows how to ruin fun. 236 Michael: We should keep her out of our client dinners too; she is literally the opposite of fun. If she wants to be a worker bee, perhaps Connor you can assign her the non-client facing tasks? Dave: I would like for her to get the feedback to lighten up a bit in life and to not be such a whiner about everything. I think she may be able to improve over time on her quantitative skills. Connor: Okay, I will share this feedback with Stephanie. Thanks guys! Manipulation Version B – Moderate strength Performance Discussion Notes for Stephanie Capelli Participants: Connor (Team Lead), Michael, Zach, and Dave Dated: 06/12/2017 Connor: I am collecting feedback for Stephanie’s upcoming performance review. Feel free to be absolutely honest in discussing her strengths and weaknesses with me – don’t hold back, and say whatever is on your mind. I will obviously share a cleaned-up version of the notes with her, and the feedback will be shared anonymously. So, what do you think of Stephanie’s work so far? Michael: I think she is okay to work with, but she very often struggles with quantitative work. Zach: Yes, but that’s typical – I mean, we know that women struggle with even the most basic math. Dave: …Yes, that is so very true, Zach. Anyway, I thought she was okay after I coached her on the models. Zach: My biggest problem with her is her inability to take a joke – like seriously, we like goofing around and making jokes…but she is so over-sensitive about everything! Connor: Well you know the HR policy Zach, we have to watch what we say around girls… but between us, I agree, Stephanie is particularly over-sensitive. 237 Michael: Could we keep her out of our client dinners; she is literally the opposite of fun. If she wants to be a worker bee, perhaps Connor you can assign her the non-client facing tasks? Dave: I would like for her to get the feedback to lighten up a bit in life and to not be so sensitive all the time. I think she may be able to improve over time on her quantitative skills. Connor: Okay, I will share this feedback with Stephanie. Thanks guys! Manipulation Version C – Subtle strength was created, but then not piloted Performance Discussion Notes for Stephanie Capelli Participants: Connor (Team Lead), Michael, Zach, and Dave Dated: 06/12/2017 Connor: I am collecting feedback for Stephanie’s upcoming performance review. Feel free to be absolutely honest in discussing her strengths and weaknesses with me – don’t hold back, and say whatever is on your mind. I will obviously share a cleaned-up version of the notes with her, and the feedback will be shared anonymously. So, what do you think of Stephanie’s work so far? Michael: I think she is okay to work with, but can sometimes struggle with quantitative work. Zach: Yes, but that’s gender differences at play – women do struggle with math. Dave: …Yes, that can be true sometimes. Anyway, I thought she was okay after I coached her on the models. Zach: My biggest problem with her is that I don’t feel comfortable goofing around and making jokes when she is around…she can be a bit sensitive at times. Connor: Well you know the HR policy Zach, we have to be professional at work...but between us, I agree, Stephanie is just a tad bit over-sensitive. 238 Michael: I would hate to keep her out of our client dinners, but she is literally the opposite of fun. Given her preference to be a worker bee, perhaps Connor you can assign her the non-client facing tasks? Dave: I would like for her to get the feedback to lighten up a bit in life. I really do think she can improve over time on her quantitative skills. Connor: Okay, I will share this feedback with Stephanie. Thanks guys! Attention Check for the Performance Discussion task Instructions: Please tell us about the goal you were assigned for this task: 1. My goal was to highlight my work contributions to my team 2. My goal was to assert my identity and to be true to who I am when interacting with my team 3. My goal was to make every effort to get along with my team 4. I was not assigned a goal Piloting how often women experience sexism 1. Have you experienced sexism on the job over the past two weeks? 2. If so, how many times have you experienced sexism on the job? 239 STUDY ONE Self-regulatory depletion: State self-control capacity scale (Ciarocco, Twenge, Muraven, & Tice, 2007). Rating scale: Not true (1) to very true (7) Likert-type scale Instructions: Given the sexist event you have just written about, please respond how true each of the following statements are for you at this moment. 1. I feel mentally exhausted right now (R) 2. It would take a lot of effort for me to concentrate on something right now (R) 3. Right now, I need something pleasant to make me feel better (R) 4. I feel motivated right now 5. If I were given a difficult task right now, I would give up easily (R) 6. I feel drained right now (R) 7. I have lots of energy right now 8. I feel worn out right now (R) 9. If I were tempted by something right now, it would be very difficult to resist (R) 10. I would want to quit any difficult task I was given right now (R) 11. I feel calm and rational right now 12. I can’t absorb any more information right now (R) 13. I feel lazy right now (R) 14. Right now I would find it difficult to plan ahead (R) 15. I feel sharp and focused right now 16. I want to give up right now (R) 17. This would be a good time for me to make an important decision 18. Right now, I feel like my willpower is gone (R) 19. My mind feels unfocused right now (R) 20. I feel ready to concentrate right now 21. My mental energy is running low right now (R) 22. A new challenge would appeal to me right now 23. Right now, I wish I could just relax for a while (R) 24. I am having a hard time controlling my urges right now (R) 25. I feel discouraged right now (R) 240 Accepting Identity Management Strategies (compiled across Moody-Ayers et al, 2005; Roberts, Settles, Jellison, 2008; and items that have been drafted for the current study) Instructions: Given the sexist event you have just written about, please share how you managed the sexist situation by indicating the extent to which you engaged in the following behaviors. Rating scale: (1) Strongly disagree to (5) Strongly Agree 1. Accepting sexism 2. Ignoring sexism 3. Keeping it to myself 4. Avoiding sexism 5. Trying to conduct myself in a manner inconsistent with the stereotypes of being a woman 6. Trying to bond with others like me 7. Trying to use self-deprecating humor to diffuse the situation Open-ended item. If you are unable to find behaviors that you engaged in, please share how you felt, thought, or acted in the situation [Open-ended item] Challenging Identity Management Strategies (compiled across Moody-Ayers et al, 2005; Roberts, Settles, Jellison, 2008; and items that have been drafted for the current study) Instructions: Given the sexist event you have just written about, please share how you managed the sexist situation by indicating the extent to which you engaged in the following behaviors. Rating scale: (1) Strongly disagree to (5) Strongly Agree 1. Speaking up 2. Trying to change things 3. Working harder to prove them wrong 4. Trying to educate my colleagues about the stereotypes of other women 5. Trying to communicate the inaccuracy of the stereotypes about my gender 6. Trying to make favorable social comparisons that make women look better than men 7. Denying that there is stigma on me because of being a woman Open-ended item. If you are unable to find behaviors that you engaged in, please share how you coped with the situation [Open-ended item] Self-regulatory foci. Promotion and prevention foci will be measured by the Work Regulatory Focus scale (Neubert, Kacmar, Carlson, Chonko, & Roberts, 2008). Response Scale: Strongly disagree (1) to Strongly agree (5) Instructions: Please rate the following statements that are true of you generally, across most situations. Prevention focus 1. I concentrate on completing my work tasks correctly to increase my job security 2. At work I focus my attention on completing my assigned responsibilities 3. Fulfilling my work duties is very important to me 4. At work, I strive to live up to the responsibilities and duties given to me by others 5. At work I am often focused on accomplishing tasks that will support my need for security 241 6. I do everything I can to avoid loss at work 7. Job security is an important factor for me in any job search 8. I focus my attention on avoiding failure at work 9. I am very careful to avoid exposing myself to potential losses at work Promotion focus 10. I take chances at work to maximize my goals for advancement 11. I tend to take risks at work in order to achieve success 12. If I had an opportunity to participate on a high-risk, high-reward project I would definitely take it 13. If my job did not allow for advancement, I would likely find a new one 14. A chance to grow is an important factor for me when looking for a job 15. I focus on accomplishing job tasks that will further my advancement 16. I spend a great deal of time envisioning how to fulfill my aspirations 17. My work priorities are impacted by a clear picture of what I aspire to be 18. At work, I am motivated by my hopes and aspirations. Identity Centrality (MIBI; Sellers et al., 1997) Response Scale: Strongly disagree (1) to Strongly agree (7) Instructions: Please rate the following statements that are true of you generally, across most situations. 1. Overall, being a woman has very little to do with how I feel about myself (R) 2. In general, being a woman is an important part of my self-image. 3. My destiny is tied to the destiny of other women. 4. Being a woman is unimportant to my sense of what kind of person I am. (R) 5. I have a strong sense of belonging with women. 6. I have a strong attachment to other women. 7. Being a woman is an important reflection of who I am. 8. Being a woman is not a major factor in my social relationships (R) Here is a comparison of the original and adapted items: Seller’s MIBI items (adaptations of the MIBI for gender centrality measure in the current study are italicized and bolded): 1. Overall, being Black has very little to do with how I feel about myself (R) was adapted as ‘Overall, being a woman has very little to do with how I feel about myself’ (R) 2. In general, being Black is an important part of my self-image was adapted as ‘In general, being a woman is an important part of my self-image’. 3. My destiny is tied to the destiny of other Black people was adapted as ‘My destiny is tied to the destiny of other women’ 4. Being Black is unimportant to my sense of what kind of person I am (R) was adapted as ‘Being a woman is unimportant to my sense of what kind of person I am (R)’ 5. I have a strong sense of belonging to Black people was adapted as ‘I have a strong sense 6. I have a strong attachment to other Black people was adapted as ‘I have a strong of belonging with women’ attachment to other women’ 242 7. Being Black is an important reflection of who I am was adapted as ‘Being a woman is an important reflection of who I am’ 8. Being Black is not a major factor in my social relationships (R) was adapted as ‘Being a woman is not a major factor in my social relationships (R)’ Coping self-efficacy scale (Chesney et al., 2006) Response scale: An 11-point scale; anchor points on the scale 0 (‘cannot do at all’), 5 (‘moderately certain can do’), and 10 (‘certain can do’) Instructions: When things aren’t going well for you, or when you’re having problems, how confident or certain are you that you can do the following: Use problem-focused coping 1. Break an upsetting problem down into smaller parts 2. Sort out what can be changed, and what cannot be changed 3. Make a plan of action and follow it when confronted with a problem 4. Leave options open when things get stressful 5. Think about one part of the problem at a time 6. Find solutions to your most difficult problems Stop unpleasant emotions and thoughts 1. Make unpleasant thoughts go away 2. Take your mind off unpleasant thoughts 3. Stop yourself from being upset by unpleasant thoughts 4. Keep from feeling sad Get support from friends and family 1. Get friends to help you with the things you need 2. Get emotional support from friends and family 3. Make new friends Diversity climate measure (Button, 1996; 2001) Instructions: Please answer the following by thinking about the climate of your organization where you faced the sexist event that you have just described. Rating scale: 7-point Likert-type scale ranging from 1 (strongly disagree) to 7 (strongly agree) 1. My employer is affirming toward female employees (R) 2. The leaders of this organization are committed to the equitable treatment of female employees (R) 3. The policies of this organization are fair and equitable to its female employees (R) 4. This organization does not treat female employees fairly. 5. This organization takes steps to ensure that females are treated just like males (R) 6. My co-workers are more likely to be supportive of females because of the training programs maintained by this organization (R) 7. This organization unfairly discriminates against female employees in the distribution of job-related opportunities (e.g., promotions, work assignments) 8. This organization unfairly discriminates against female employees in the distribution of benefits. 243 Role hierarchy. (King et al., 2017) Instructions: In the sexist event you just narrated, please tell us a bit about the perpetrator. Specifically, please tell us where in the organizational hierarchy was this person situated compared to you? For example, if the person was a supervisor, you will mark “someone at a higher level than me”, while if the person was someone you manage, then mark “someone at a junior level than me.” Note, if this event occurred in a group setting, and multiple people were involved, you are able to select multiple options below. Rating scale: 1. Someone at a higher level than me 2. Someone at the same level as me 3. Someone who was below my level 4. Client(s) Control Variables State-Trait Anxiety Inventory, Form Y (Spielberger, 1983). Instructions: Rate the following to the extent that it is generally true for you Rating scale: 1 = strongly disagree to 4 = strongly agree 1. I am happy 2. I am content 3. I feel satisfied with myself 4. I feel pleasant 5. I feel secure 6. I lack self-confidence 7. I feel inadequate 8. I feel like a failure 9. I am a steady person 10. I wish I could be as happy as others seem to be 11. I make decisions easily 12. I am ‘calm, cool, and collected’ 13. I feel rested 14. Some unimportant thought runs through my mind and bothers me 15. I worry too much over something that really doesn’t matter 16. I get in a state of tension or turmoil as I think over my recent concerns and interests 17. I have disturbing thoughts 18. I take disappointments so keenly that I can’t put them out of my mind 19. I feel that difficulties are piling up so that I can’t overcome them 20. I feel nervous and restless Conscientiousness (Big Five Inventory; John, Donahue, & Kentel, 1991) Instructions: Here are a number of characteristics that may or may not apply to you. Please rate the following statements to indicate the extent to which you agree or disagree with that statement. Rating Scale: 1 = Disagree strongly to 5 = Agree strongly 244 Retrieved from John, Naumann, and Soto (2008) Note. Only Conscientiousness items will be used for the study; others are for exploratory purposes only. Demographic Variables Please report the following: 1. Age [open-ended item] 2. Gender [open-ended item] 3. Race-ethnicity: a. White b. Hispanic or Latino c. Black or African American d. Native American or American Indian e. Asian/Pacific Islander f. Other 4. How long have you worked in the organization where you encountered sexism? 5. How many years have you been employed in total (i.e., across all the organizations you have worked for) [open-ended item] 6. How many years have you worked in your current organization? [open-ended item] 7. What is your current job title? [open-ended item] a. < 1 year b. 1 year – 5 years c. 6 years – 10 years d. 11 years – 15 years e. 16 years and up 245 STUDY TWO Manipulation check for the Performance Discussion task Rating Scale: 1 = Disagree strongly to 5 = Agree strongly Instructions: Please rate the following: 1. The team is sexist 2. It is appealing to work with this team Attention Check for the Performance Discussion task Instructions: Please tell us about the goal you were assigned for this task: 1. My goal was to highlight my work contributions to my team 2. My goal was to assert my identity and to be true to who I am when interacting with my team 3. My goal was to make every effort to get along with my team 4. I was not assigned a goal Self-regulatory depletion: State self-control capacity scale (Ciarocco, Twenge, Muraven, & Tice, 2007). Rating scale: Not true (1) to very true (7) Likert-type scale Instructions: Given the sexist event you have just written about, please respond how true each of the following statements are for you at this moment. 1. I feel mentally exhausted right now (R) 2. It would take a lot of effort for me to concentrate on something right now (R) 3. Right now, I need something pleasant to make me feel better (R) 4. I feel motivated right now 5. If I were given a difficult task right now, I would give up easily (R) 6. I feel drained right now (R) 7. I have lots of energy right now 8. I feel worn out right now (R) 9. If I were tempted by something right now, it would be very difficult to resist (R) 10. I would want to quit any difficult task I was given right now (R) 11. I feel calm and rational right now 12. I can’t absorb any more information right now (R) 13. I feel lazy right now (R) 14. Right now I would find it difficult to plan ahead (R) 15. I feel sharp and focused right now 16. I want to give up right now (R) 17. This would be a good time for me to make an important decision 18. Right now, I feel like my willpower is gone (R) 19. My mind feels unfocused right now (R) 20. I feel ready to concentrate right now 21. My mental energy is running low right now (R) 22. A new challenge would appeal to me right now 23. Right now, I wish I could just relax for a while (R) 24. I am having a hard time controlling my urges right now (R) 246 25. I feel discouraged right now (R) Accepting Identity Management Strategies (compiled across Moody-Ayers et al, 2005; Roberts, Settles, Jellison, 2008; and items that have been drafted for the current study) Instructions: Given the sexist event you have just written about, please share how you managed the sexist situation by indicating the extent to which you engaged in the following behaviors. Rating scale: (1) Strongly disagree to (5) Strongly Agree 1. Accepting sexism 2. Ignoring sexism 3. Keeping it to myself 4. Avoiding sexism 5. Trying to conduct myself in a manner inconsistent with the stereotypes of being a woman 6. Trying to bond with others like me 7. Trying to use self-deprecating humor to diffuse the situation Open-ended item. If you are unable to find behaviors that you engaged in, please share how you coped with the situation [Open-ended item] Challenging Identity Management Strategies (compiled across Moody-Ayers et al, 2005; Roberts, Settles, Jellison, 2008; and items that have been drafted for the current study) Instructions: Given the sexist event you have just written about, please share how you managed the sexist situation by indicating the extent to which you engaged in the following behaviors. Rating scale: (1) Strongly disagree to (5) Strongly Agree 1. Speaking up 2. Trying to change things 3. Working harder to prove them wrong 4. Trying to educate my colleagues about the accomplishments of other women 5. Trying to communicate the inaccuracy of the stereotypes about my gender 6. Trying to make favorable social comparisons that make women look better than men 7. Denying that there is stigma on me because of being a woman Open-ended item. If you are unable to find behaviors that you engaged in, please share how you coped with the situation [Open-ended item] Identity Centrality (MIBI; Sellers et al., 1997) Response Scale: Strongly disagree (1) to Strongly agree (7) Instructions: Please rate the following statements that are true of you generally, across most situations. 1. Overall, being a woman has very little to do with how I feel about myself (R) 2. In general, being a woman is an important part of my self-image. 3. My destiny is tied to the destiny of other women. 4. Being a woman is unimportant to my sense of what kind of person I am. (R) 5. I have a strong sense of belonging with women. 6. I have a strong attachment to other women. 247 7. Being a woman is an important reflection of who I am. 8. Being a woman is not a major factor in my social relationships, (R) General self-efficacy scale (Jerusalem & Schwarzer, 1992; Schwarzer & Jerusalem, 1995) Response scale: 1 = Not at all true, 2 = Hardly true, 3 = Almost true, and 4 = Very true Instructions: Please rate the following statements that are true of you generally, across most situations. 1. I can always manage to solve difficult problems if I try hard enough. 2. If someone opposes me, I can find means and ways to get what I want. 3. It is easy for me to stick to my aims and accomplish my goals 4. I am confident that I could deal efficiently with unexpected events 5. Thanks to my resourcefulness, I know how to handle unforeseen situations. 6. I can solve most problems if I invest the necessary effort 7. I can remain calm when facing difficulties because I can rely on my coping abilities 8. When I am confronted with a problem, I can usually find several solutions. 9. If I am in trouble, I can usually think of something to do 10. No matter what comes my way, I am usually able to handle it. Control Variables State-Trait Anxiety Inventory, Form Y (Spielberger, 1983). Instructions: Rate the following to the extent that it is generally true for you Rating scale: 1 = strongly disagree to 4 = strongly agree 1. I am happy 2. I am content 3. I feel satisfied with myself 4. I feel pleasant 5. I feel secure 6. I lack self-confidence 7. I feel inadequate 8. I feel like a failure 9. I am a steady person 10. I wish I could be as happy as others seem to be 11. I make decisions easily 12. I am ‘calm, cool, and collected’ 13. I feel rested 14. Some unimportant thought runs through my mind and bothers me 15. I worry too much over something that really doesn’t matter 16. I get in a state of tension or turmoil as I think over my recent concerns and interests 17. I have disturbing thoughts 18. I take disappointments so keenly that I can’t put them out of my mind 19. I feel that difficulties are piling up so that I can’t overcome them 20. I feel nervous and restless Conscientiousness (Big Five Inventory; John, Donahue, & Kentel, 1991) 248 Instructions: Here are a number of characteristics that may or may not apply to you. Please rate the following statements to indicate the extent to which you agree or disagree with that statement. Rating Scale: 1 = Disagree strongly to 5 = Agree strongly Retrieved from John, Naumann, and Soto (2008) Note. Only Conscientiousness items will be used for the study; others are for exploratory purposes only. Demographic Variables Please report the following: 1. Age [open-ended item] 2. Gender [open-ended item] 3. Race-ethnicity: a. White b. Hispanic or Latino c. Black or African American d. Native American or American Indian e. Asian/Pacific Islander f. Other 4. How many years have you been employed in total (i.e., across all the organizations you have worked for) [open-ended item] 5. How many years have you worked in your current organization? [open-ended item] 6. What is your current job title? [open-ended item] Anagram Task (Solvable anagrams for filler task). 249 Below is a list of words, the letters of which have been re-arranged. Please re-arrange the letters so that all of them are used when spelling a word. Please continue until you have completed all of the words or until you are tired. 1. T B N E 2. E R O T 3. Y R S T O 4. E N R S C E 5. E E D O N C 6. S C N R Y E E 7. L E S C A R D 8. R A S T E R G N 9. R U T E E A R C 250 STUDY THREE Psychosomatic Health Indices. Hopkins Checklist (The somatic subscale; Derogatis, Lipman, Rickels, & Uhlenhuth, 1974). Response Scale: (1 = Not at all, 2 = A little, 3 = Quite a bit, 4 = Extremely) Scoring: Average scores, with higher scores indicating higher XXX (in the explanation extract from body of study) Instructions: How have you felt during the past seven days including today? Use the following scale to describe how distressing you have found these things over this time: 1. Headaches (1) 2. Faintness or dizziness (4) 3. Pains in the heart or chest (12) 4. Feeling low in energy or slowed down (14) 5. Pains in the lower part of your back (27) 6. Soreness of your muscles (42) 7. Trouble getting your breath (48) 8. Hot or cold spells (49) 9. Numbness or tingling in parts of your body (52) 10. A lump in your throat (53) 11. Weakness in parts of your body (56) 12. Heavy feeling in your arms or legs (58) Psychological Strains. (General Health Questionnaire; Goldberg, 1972). Items extracted from Hyman (2001): Response Scale: (0 = Less than usual, 1 = No more than usual, 2 = Rather more than usual, and 3 = Much more than usual) Scoring: Summed score Instructions: I would like to know how your health has been today, in comparison to your usual self. Please refer only to present and recent complaints from the day, not those that you had in the past. It is important that you answer ALL the questions: Have you today: 1. Been able to concentrate on whatever you are doing? 2. Lost much sleep over worry? 3. Felt that you are playing a useful part in things? 4. Felt capable about making decisions about things? 5. Felt constantly under strain? 6. Felt that you couldn’t overcome your difficulties? 7. Been able to enjoy your normal day-to-day activities? 8. Been able to face up to your problems? 9. Been feeling unhappy and depressed? 10. Been losing confidence in yourself? 11. Been thinking of yourself as a worthless person? 251 12. Been feeling reasonably happy, all things considered? Self-regulatory depletion: 10-item short form State self-control capacity scale (Ciarocco, Twenge, Muraven, & Tice, 2007). Rating scale: Not true (1) to very true (7) Likert-type scale Instructions: Given the sexist event you have just written about, please respond how true each of the following statements are for you at this moment. 3. Right now, I need something pleasant to make me feel better (R) 6. I feel drained right now (R) 9. If I were tempted by something right now, it would be very difficult to resist (R) 10. I would want to quit any difficult task I was given right now (R) 11. I feel calm and rational right now 12. I can’t absorb any more information right now (R) 13. I feel lazy right now (R) 14. Right now I would find it difficult to plan ahead (R) 15. I feel sharp and focused right now 16. I want to give up right now (R) 18. Right now, I feel like my willpower is gone (R) Control Variables State-Trait Anxiety Inventory, Form Y (Spielberger, 1983). Instructions: Rate the following to the extent that it is generally true for you Rating scale: 1 = strongly disagree to 4 = strongly agree 1. I am happy 2. I am content 3. I feel satisfied with myself 4. I feel pleasant 5. I feel secure 6. I lack self-confidence 7. I feel inadequate 8. I feel like a failure 9. I am a steady person 10. I wish I could be as happy as others seem to be 11. I make decisions easily 12. I am ‘calm, cool, and collected’ 13. I feel rested 14. Some unimportant thought runs through my mind and bothers me 15. I worry too much over something that really doesn’t matter 16. I get in a state of tension or turmoil as I think over my recent concerns and interests 17. I have disturbing thoughts 18. I take disappointments so keenly that I can’t put them out of my mind 19. I feel that difficulties are piling up so that I can’t overcome them 20. I feel nervous and restless 252 Conscientiousness (Big Five Inventory; John, Donahue, & Kentel, 1991). Instructions: Here are a number of characteristics that may or may not apply to you. Please rate the following statements to indicate the extent to which you agree or disagree with that statement. Rating Scale: 1 = Disagree strongly to 5 = Agree strongly Retrieved from John, Naumann, and Soto (2008) Note. Only Conscientiousness items will be used for the study; others are for exploratory purposes only. Demographic Variables Please report the following: 1. Age [open-ended item] 2. Gender [open-ended item] 3. Race-ethnicity: a. White b. Hispanic or Latino c. Black or African American d. Native American or American Indian e. Asian/Pacific Islander f. Other Proximal goals (created for the current study) Instructions: For each type of goal below (definitions provided), please rate your most important goal that you were pursuing at work today. Rating scale: Choice 1. Self-enhancement goals 253 2. Self-verification goals 3. Belonging goals Identity Management Strategies (compiled across Moody-Ayers et al, 2005; Roberts, Settles, Jellison, 2008; and items that have been drafted for the current study); items 1 through 7 reference accepting identity management strategies, and items 8 through 14 reference challenging identity management strategies. Item 15 is to allow participants to enter open-ended responses if any. Instructions: Given the sexist event you have just written about, please share how you managed the sexist situation by indicating the extent to which you engaged in the following behaviors. Rating scale: Choice 1. Accepting it 2. Ignoring it 3. Keeping it to myself 4. Avoiding it 5. Trying to conduct myself in a manner inconsistent with the stereotypes of being a woman 6. Trying to bond with others like me 7. Trying to use self-deprecating humor to diffuse the situation 8. Speaking up 9. Trying to change things 10. Working harder to prove them wrong 11. Trying to educate my colleagues about the accomplishments of other women 12. Trying to communicate the inaccuracy of the stereotypes about my gender 13. Trying to make favorable social comparisons that make women look better than men 14. Denying that there is stigma on me because of being a woman 15. Open-ended item. If you are unable to find behaviors that you engaged in, please share how you coped with the situation [Open-ended item] 4. How many years have you been employed in total (i.e., across all the organizations you have worked for) [open-ended item] 5. How many years have you worked in your current organization? [open-ended item] 6. What is your current job title? [open-ended item] Demographic Variables Please report the following: 1. Age [open-ended item] 2. Gender [open-ended item] 3. Race-ethnicity: a. White b. Hispanic or Latino c. Black or African American d. Native American or American Indian e. Asian/Pacific Islander f. Other 254 APPENDIX F: VIVID RECALLS EXAMPLES The tables below showcase a selection of open-ended vivid recalls in both the experimental and control conditions for Pilot A. Table 33: Experimental condition responses Pilot A. Participant A 255 Examples of Open-ended responses in the experimental condition “I worked at Burger King and the younger men always wanted to feel on us woman and buy us drinks. I was tired of it so I went to the boss and he said if I wasn't so pretty it wouldn't of happen. So we all went out one night and a gentleman rapped [sic] me from work. I went to the police and they ended up shutting the place down / [sic]. I’m still messed up from it all.” “So, one time I was working. I was a waitress at the time. There was a man who was being extremely rude towers me. I tried to just push through and continue giving service to the table, but the guy just kept going. I took the situation to the manager. He didn't seem to care and just told me I was only acting upset because I was an emotional woman.” “It was about 3 years ago at work. I had lost about 20 pounds. One of my reps who [sic] I hadn't see for about a year had come in for a meeting. I was in the kitchen making coffee and looking out the window, my rep snuck up behind me, put his arms around, tried to kiss me on my lips. Told me I told [sic] so hot and skinny and that I had big tits, and slapped my butt - I was furious. I went to HR - all that he was told was to not ever do that to me again. The day he left he said again that I had a hot body and liked [sic] his lips.” Examples of Open-ended responses in the control condition “I was 15 years old and was able to go to Wisconsin for my summer vacation with my aunt and uncle. We went to stay with my aunt brother. Her brother lived on a big dairy farm. It was my first time in Wisconsin and it was so pretty. This one day we went to the back of the farm and there was a river there and we rode on inter tubes down the river. It was a great summer and I learned a lot about myself that year…” “My husband and I took a trip to Alexandria, Virginia. We went and saw George washing tons house at Mt.Vernon. My husband and I had so much fun. We enjoyed the time we had by ourselves without our 2 year old. My reactions to the experience was [sic] joy, because I was enjoying getting out of the routine and doing something we both enjoyed. Also, it was a stress reliever because we didn't have any responsibilies [sic] with our baby. We were both able to fully relax and enjoying our surroundings. When we came back to work, we were able to focus. Our home life was less stressful after the vacation. We were able to learn and enjoy a new place.” “I remember when I was little my mom used to take us to the movies every Friday after school if he did good throughout that week” Table 34: Control condition responses Pilot A. Participant D B C B C APPENDIX G: MPLUS SYNTAX MPlus Syntax A generic model of MPlus commands is extracted from quantpsy.org to showcase the logic of the analysis where x denotes independent variables, m denotes mediator, and y denotes outcomes (please note, when conducting the actual analysis x, m, and y will be replaced with the variable names from the current study): CENTERING IS GROUPMEAN (x m); ! group-mean center IVs and mediator CLUSTER IS id; ! Level-2 identifier WITHIN ARE x m; ! identify variables with within-person variance BETWEEN ARE xmean mmean ! identify variables with between-person variance ANALYSIS: TYPE IS TWOLEVEL RANDOM; MODEL: ! model specification follows %WITHIN% ! model for within-effects follows m ON x (aw); ! regress m on x, call the slope “aw” y ON m (bw); ! regress y on m, call the slope “bw” y ON x; ‘%BETWEEN% ! model for between-effects follows mmean y; ! estimate level-2 (residual) variances for mmean and y mmean ON xmean (ab); ! regress mmean on xmean, call the slope "ab" y ON mmean (bb); ! regress y on mmean, call the slope "bb" y ON xmean; ! regress y on xmean MODEL CONSTRAINT: ! section for computing indirect effects NEW(indb indw); ! name the indirect effects indw=aw*bw; ! compute the Within indirect effect indb=ab*bb; ! compute the Between indirect effect OUTPUT: TECH1 TECH8 CINTERVAL; ! request parameter specifications, starting values, optimization history, and confidence intervals (p.5) 256 APPENDIX H: TRAINING PROTOCOL Training Protocol for ESM Participants This protocol had three parts. First, I defined sexism for the participants using a dictionary definition and then defined hostile and benevolent sexism. Second, I presented examples of day-to-day sexist events that I had gathered from reading different blogs, class readings, and personal experience. The goal was to familiarize participants with how sexism can manifest itself in different types of behaviors and comments. Finally, I presented a measure by Klonoff and Landrine (1995) intended to have the participants reflect upon their own lives and how often they may have faced sexism and the forms that the sexist events took. The goal of these three pieces was to immerse the participants in discussions of sexism that move from impersonal to personalized, self-reflective mode so that when a sexist event happens, it is at the forefront of their cognition to make a note of in their diary. The definitions were presented as part of a welcome email to the participants who had opted in to the study. Email Hello, My name is Dia Chatterjee and I am writing you about a research project you recently agreed to participate in via Qualtrics. Dr. Ann Marie Ryan and I would like to thank you for opting to participate in our research on sexism and how it impacts women at work. By opting in for this research, you have partnered with us to take an important step towards understanding (a) the frequency with which women in the workforce face sexism, (b) how women in the workforce cope with sexism and manage their key goals at work, and (c) the impact of sexism on women’s psychological and physical health. By sharing your experiences over the next four weeks, you will help us shed light on these important questions. One of our key goals with this research is to 257 offer research based, actionable advice to organizations on creating safe and supportive systems for women, and we want you to know that the insights you will provide us are critical to achieving that goal. We will begin sending you the research surveys starting Monday, February 4th, 2018 and will end the surveys on March 4th, 2018. As mentioned earlier, you will be sent two surveys each day: (a) in the morning, we will ask you to fill out your goals for the day, and (b) in the evening, we will ask you to report on any events of sexism that you may have encountered. The surveys will have detailed instructions so you don’t have to worry about remembering these things. That said, we would like to take this opportunity to share some definitions of sexism and definitions of different goals with you. Please take a few minutes to read through these materials (see attached). These definitions will help you familiarize yourself with the language we use in the surveys, and this familiarity help you provide insightful answers to our questions. Over the course of the next four weeks, if you have any questions, please feel free to contact me at email.private@gmail.com (preferred email) or at email.private@msu.edu. We truly appreciate your interest and commitment in helping us understand how sexism impacts women in the work place. Thanks & Warm Regards, Dia Chatterjee Training Materials Defining sexism. Sexism can manifest at work in many different forms. We will begin with first presenting the definition of sexism, and then showcasing sexism in its different forms using examples. These examples are the “lived-experiences” of women in different workplaces as reported in various media, blogs and forums on sexism. Merriam-Webster dictionary defines 258 sexism as follows: “Prejudice or discrimination based on sex; especially: discrimination against women” and “behavior, conditions, or attitudes that foster stereotypes of social roles based on sex”. Scholars of sexism distinguish between hostile sexism – the ‘in your face’ kinds of behaviors that would on the surface be described as prejudicial against women, and benevolent sexism – the ‘subtle and indirect’ kinds of behaviors, that on the surface may look positive and kind, but may actively promote prejudice or discrimination against women. For example, hostile sexism is when a man tells a woman that she is bad at math because she is a woman. This is hostile because on the surface this type of speech considers the woman as incapable of doing math because she is a woman. In contrast, benevolent sexism is when a man assumes that his female classmate will need help on math and issues a self-invite to coach her after class. This is an example of benevolent sexism because while on the surface the man is trying to be helpful to his female peer, he is doing so out of the stereotypical belief that she needs help on math because she is a woman. Now that we have covered how sexism is typically defined, below we present a few examples of sexism at workplace. 259 Table 35: Day to day sexist situations to train participants to recognize sexist events. • Day to day examples of sexism at workplace shared by women I was told that I should not apply for the executive business training because I will not be able to keep the work hours that men can keep. • At med school, they told me that to be a strong surgeon I would have to think like a man and be less emotional • My boss told me that men are more capable at about everything than women, except when it comes to taking care of children. • • A female colleague told me that she hangs out with men because men don’t create drama at work like women do. I was told that males can use aggressive language at work, but that’s just men being men, but women should make it a point to be lady like and speak politely and respectfully to their colleagues I was told that I am bossy! • • Men at work try to do things for me all the time and it makes me feel like they see me as their inferior I am a tall blonde and men compliment me for my looks. They make it seem like I was hired for my looks, when I am a highly competent litigator. Situations Individual beliefs communicated to you by peers, supervisors, clients (note these individuals could be male or female) • 260 • My manager seeks men’s views more than women’s in meetings • When women at work voice their ideas in the meetings, the males on • Males on my team are promoted faster than women even though the team dismiss those ideas women work so much harder • My manager always asks the women on the team to plan and host parties but does not bother the men-folk • • My male supervisor comments on my appearance and/or my clothes • When I am focused on a task, I tend to scrunch up my face – my male colleagues ask me to smile. No one ever tells a man to smile though. I am expected to nurture people at work just because I am a woman – they call me the “work mom” (but the men are not called “work dads”) • At work, I am expected to be the peacekeeper because I am a woman • Men on my team are always assuming that I need help with technical stuff when I am an engineer just like them • The other day a guy from the sales team tried to explain my job to me even though he full well knows that I am an expert in the field • The guys don’t invite me to client dinners where deals are made because they say that they often discuss things that would be offensive to the female sensibilities • The men on my team are always making crude sexual jokes in my presence • When men get heated at work they call it “duking it out”, when women get heated at work they call it “being bitchy” • The men on the trading floor use words that reference the female anatomy to call another male “weak” • After a rough meeting, I was told to “man up!” 261 Table 35 (cont’d) Individual behaviors directed at you by peers, supervisors, and/or clients Table 35 (cont’d) Institutional policies that may create conditions for sexism • • I recently found out that even though I have comparable education and experience, my male colleague earns more than me. When I asked my boss about this pay differential, he stated that I should be thankful for what I earn and not be nosy about other peoples’ salaries. I am required to wear heels and be uncomfortable at work in the name of appearing ‘professional’ • Women are not granted promotion at the same rate as men in our firm because my organization believes that after I have a child I will take time off from work • Clients in my firm often ask for a male manager to head their casework and the management complies as a matter of policy • My female boss told me that I needed to wear make-up to look ‘presentable’ even though the amount of foundation on my face has nothing to do with my abilities as a consultant It is annoying how most of the client facing materials and presentations etcetera at work showcase male images in leadership roles • Self-reflection exercise. Once participants go over these examples, the Schedule of Sexist Events measure (Klonoff & Landrine, 1995) was also included in the email to get them to think about how they may have endured sexism throughout their lives. The measure was used as a training aid for women to reflect on their own experiences with sexism so that they are better able to report on sexist events as they occur once the ESM study is initiated. Participants were asked to “think about each of the items presented below and jot down if you have ever experienced similar incidents. Think about your recalled experiences in as vivid a manner as you can”: 1. Have you been treated unfairly by teachers or professors because you are a woman? 2. Have you been treated unfairly by your employer, boss, or supervisors because you are a woman? 3. Have you been treated unfairly by your coworkers, fellow students, or colleagues because you are a woman? 262 4. Have you been treated unfairly by people in service jobs (by store clerks, waiters, bartenders, waitresses, bank tellers, mechanics and others) because you are a woman? 5. Have you been treated unfairly by strangers because you are a woman? 6. Have you been treated unfairly by people in helping jobs because you are a woman? (Helping jobs include doctors, nurses, psychiatrists, case workers, dentists, school counselors, therapists, school principals etcetera) 7. Have you been treated unfairly by neighbors because you are a woman? 8. Have you been treated unfairly by your boyfriend, husband, or other important man in your life because you are a woman? 9. Were you denied a raise, a promotion, a tenure, a good assignment, a job, or other such thing at work that you deserved because you are a woman? 10. Have you been treated unfairly by your family because you are a woman? 11. Have people made inappropriate or unwanted sexual advances to you because you are a woman? 12. Have people failed to show you the respect that you deserve because you are a woman? 13. Have you wanted to tell someone off for being sexist? 14. Have you been really angry about something really sexist being done to you? 15. Have you ever been forced to take drastic steps (filing a grievance, filing a lawsuit, quitting your job, moving away, or other actions) to deal with something sexist that was done to you? 16. Have you been called a sexist name like bitch, cunt, chick, or other names? 17. Have you gotten into an argument or a fight about something sexist that was done or said to you or someone else? 263 18. Have you been made fun of, picked on, pushed, shoved, hit, or threatened with harm because you are a woman? 19. Have you heard people making sexist jokes, or degrading sexual jokes? 20. How different would your life be now if you HAD NOT BEEN treated in a sexist and unfair way? Defining goals. All of us pursue different kinds of goals each day, and some goals may be more important than others based on what each situation requires from us. In work settings, organizational scholars have identified three types of goals that occur most commonly. 1. ‘Self-enhancement goals’ – When we try to ensure that we perform to the best of our capabilities at work, or when we try to put our best foot forward in discussions and meetings with team-members, supervisors, and customers we are pursuing self- enhancement goals. For example, when at work we try to shine a light on our good performance as a way to help us move ahead at work via promotions and pay-raises, better client portfolios, or getting access to resources at work that can help facilitate our performance in future, we are engaging in self-enhancement. a. Reflection exercise: Think about a time when you engaged in self-enhancement goals and describe the situation. 2. Self-verification goals: When we try to ensure that we bring our whole self to our work, or when we try to assert our identity in discussions and meetings with team-members, supervisors, and customers we are pursuing self-verification goals. When a male highlights how he is balancing work and parenting a month-old baby at the same time, or when a team-member states that she does not mind working long hours because she still retains the immigrant work-ethic, or when a Caucasian woman describes how her 264 farming family raised her to be an early-riser, or when a religious minority individual proudly wears an identifiable symbol to work (e.g., a Christian woman wears a cross to work, or a Muslim woman wears a hijab) – these are all examples of verbal and non- verbal self-verification goals. People pursue self-verification goals in organizations because identity is closely tied to our sense of self and our self-definitions. To have our team members, supervisors, peers, and customers see us for who we are can often create a heightened sense of self-esteem. a. Reflection exercise: Think about a time when you engaged in self-verification goals and describe the situation. 3. Belonging goals: When we try to ensure that we get along with our team-members, supervisors, and customers we are pursuing belonging goals. People pursue belonging goals in organizations because humans, generally speaking, like to get along with those who are part of our immediate groups. In organizations, fostering cordial relationships with others can be one way to fulfill this fundamental need to belong. So when people try to foster friendships, cliques, and networks at work they are pursuing belonging goals. a. Reflection exercise: Think about a time when you engaged in belonging goals and describe the situation. 265 REFERENCES 266 REFERENCES Acker, J. (1973). Women and social stratification: A case of intellectual sexism. American journal of sociology, 78, 936–945. Ackerman, J. M., Goldstein, N. J., Shapiro, J. R., & Bargh, J. A. (2009). You wear me out: The vicarious depletion of self-control. Psychological Science, 20, 326–332. Aguinis, H., Gottfredson, R. K., & Joo, H. (2013). Best-practice recommendations for defining, identifying, and handling outliers. Organizational Research Methods, 16, 270–301. Arvey, R. D., & Cavanaugh, M. A. (1995). Using surveys to assess the prevalence of sexual harassment: Some methodological problems. Journal of Social Issues, 51, 39–52. Ashforth, B. E., & Humphrey, R. H. (1993). Emotional labor in service roles: The influence of identity. Academy of management review, 18, 88–115. 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. Asparouhov T. (2005). Sampling weights in latent variable modeling. Structural Equation Modeling, 12, 411–434. Austin, J. T., & Vancouver, J. B. (1996). Goal constructs in psychology: Structure, process, and content. Psychological bulletin, 120, 338–375. Bair, A. N., & Steele, J. R. (2010). Examining the consequences of exposure to racism for the executive functioning of Black students. Journal of Experimental Social Psychology, 46, 127– 132. Bakker, A. B., & Demerouti, E. (2007). The job demands-resources model: State of the art. Journal of managerial psychology, 22, 309–328. Bandura, A. (1977). Self-efficacy: toward a unifying theory of behavioral change. Psychological review, 84, 191–215. Barber, L. K., Barnes, C. M., & Carlson, K. D. (2013). Random and systematic error effects of insomnia on survey behavior. Organizational Research Methods, 16, 616–649. Barrick, M. R., Mount, M. K., & Li, N. (2013). The theory of purposeful work behavior: The role of personality, higher-order goals, and job characteristics. Academy of management review, 38, 132–153. 267 Baumeister, R. F. (1982). A self-presentational view of social phenomena. Psychological Bulletin, 91, 3–26. Baumeister, R. F., & Leary, M. R. (1995). The need to belong: desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117, 497–529. Baumeister, R. F., DeWall, C. N., Ciarocco, N. J., & Twenge, J. M. (2005). Social exclusion impairs self-regulation. Journal of Personality and Social Psychology, 88, 589–604. Baumeister, R. F., Gailliot, M., DeWall, C. N., & Oaten, M. (2006). Self‐regulation and personality: How interventions increase regulatory success, and how depletion moderates the effects of traits on behavior. Journal of Personality, 74, 1773–1802. Beal, D. J., & Weiss, H. M. (2003). Methods of ecological momentary assessment in organizational research. Organizational Research Methods, 6, 440–464. Becker, L. (2008). Can Positive Mood Replenish Resource Depletion Resulting in Less Compliance through a Heuristic? (Bachelor's thesis, University of Twente). Behrendt-Mihalski, J. L. (2017). An examination of the moderating role of gendered racial identity centrality in the relationship between gendered racism and depression (Doctoral dissertation, The University of North Carolina at Charlotte). Bentler, P. M., & Chou, C. P. (1987). Practical issues in structural modeling. Sociological Methods & Research, 16, 78–117. Berdahl, J. L., Magley, V. J., & Waldo, C. R. (1996). The sexual harassment of men?: Exploring the concept with theory and data. Psychology of Women Quarterly, 20, 527–547. Berinsky, A. J., Huber, G. A., & Lenz, G. S. (2012). Evaluating online labor markets for experimental research: Amazon. com's Mechanical Turk. Political Analysis, 20, 351–368. Berjot, S., & Gillet, N. (2011). Stress and Coping with Discrimination and Stigmatization. Frontiers in Psychology, 2, 2–33. Blanz, M., Mummendey, A., Mielke, R., & Klink, A. (1998). Responding to negative social identity: A taxonomy of identity management strategies. European Journal of Social Psychology, 28, 697–729. Block, J. H. (1973). Conceptions of sex role: Some cross-cultural and longitudinal perspectives. American psychologist, 28, 512–526. Bolino, M. C., & Turnley, W. H. (1999). Measuring impression management in organizations: A scale development based on the Jones and Pittman taxonomy. Organizational Research Methods, 2, 187–206. 268 Bombay, A., Matheson, K., & Anisman, H. (2014). Appraisals of discriminatory events among adult offspring of Indian residential school survivors: The influences of identity centrality and past perceptions of discrimination. Cultural diversity and ethnic minority psychology, 20, 75– 87. Bonner, S. E., Hastie, R., Sprinkle, G. B., & Young, S. M. (2000). A review of the effects of financial incentives on performance in laboratory tasks: Implications for management accounting. Journal of Management Accounting Research, 12, 19–64. Boom, A. S. (2009). Are you in for the long run?: the role of depletion and replenishment in the persuasion process (Master's thesis, University of Twente). Bosson, J. K., Vandello, J. A., Burnaford, R. M., Weaver, J. R., & Arzu Wasti, S. (2009). Precarious manhood and displays of physical aggression. Personality and Social Psychology Bulletin, 35, 623–634. Boyles, P. (2008). " Thank you for letting me be myself": Exploring the effects of identity management strategies on engagement levels of lesbian, gay and bisexual employees (Doctoral dissertation, Virginia Tech). Brewster, M. E., & Moradi, B. (2010). Perceived anti-bisexual prejudice experiences: Scale development and evaluation. Journal of Counseling Psychology, 57, 451–468. Brown, J. K. (1970). A note on the division of labor by sex. American Anthropologist, 72, 1073– 1078. Brown, R. P., & Pinel, E. C. (2003). Stigma on my mind: Individual differences in the experience of stereotype threat. Journal of experimental social psychology, 39, 626–633. 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, 3–5. Bureau of Labor Statistics (2017). Labor force statistics from the current population survey. Retrieved from: https://www.bls.gov/cps/cpsaat18.htm Burke, R. J., & Deszca, G. 1988. Career orientations, satisfaction and health among police officers: Some consequences of person-job misfit. Psychological Reports, 62, 639–649. Button, S. B. (1996). Organizational efforts to affirm sexual diversity: A multi-level examination. Unpublished doc- toral dissertation, Pennsylvania State University, University Park. Button, S. B. (2001). Organizational efforts to affirm sexual diversity: A cross-level examination. Journal of Applied Psychology, 86, 17–28. 269 Button, S. B. (2004). Identity management strategies utilized by Lesbian and gay employees a quantitative investigation. Group & Organization Management, 29, 470–494. Bryk, A. S., Raudenbush, S. W., & Congdon, R. T. (1994). HLM 2/3. Hierarchical Linear Modeling with the HLM/2L and HLM/3L Programs. Cameron, J. E. (2004). A three-factor model of social identity. Self and identity, 3, 239–262. Cammann, C., Fichman, M., Jenkins, G. D., and Klesh, J. (1983). The Michigan organizational assessment survey: Conceptualization and instrumentation. In S.E. Seashore, E.E. Lawler III, P. H. Mirvis, and C. Cammann (Eds.), Assessing organizational change: A guide to methods, measures and practices. New York: Wiley Interscience, 71–138. Carter, L. A. (2016). Rigorous methods for the analysis, reporting and evaluation of ESM style data (Doctoral dissertation, University of Manchester). Carver, C. S., & Scheier, M. F. (1981). Attention and Self-Regulation: A Control-Theory Approach to Human Behavior. New York: Springer. Carver, C. S., & Scheier, M. F. (1990). Origins and functions of positive and negative affect: a control-process view. Psychological Review, 97, 19–35. Carr, D., & Friedman, M. A. (2005). Is obesity stigmatizing? Body weight, perceived discrimination, and psychological well-being in the United States. Journal of Health and Social Behavior, 46, 244–259. Chatterjee, P. (1989). Colonialism, nationalism, and colonialized women: The contest in India. American ethnologist, 16, 622–633. Chaudoir, S. R., & Fisher, J. D. (2010). The disclosure processes model: understanding disclosure decision-making and post-disclosure outcomes among people living with a concealable stigmatized identity. Psychological Bulletin, 136, 236–256. Chesney, M. A., Neilands, T. B., Chambers, D. B., Taylor, J. M., & Folkman, S. (2006). A validity and reliability study of the coping self-efficacy scale. British Journal of Health Psychology, 11, 421–437. Chiaburu, D. S., & Baker, V. L. (2006). Extra-role behaviors challenging the status-quo: Validity and antecedents of taking charge behaviors. Journal of Managerial Psychology, 21, 620–637. Chisholm, S. (1970). The 51% minority. In Discrimination against women, Part II, Hearing before the Special Subcommittee on Education of the House Committee on Education and Labor. Washington, D.C.: U.S. Government Printing Office. Christensen, T. C., Barrett, L. F., Bliss-Moreau, E., Lebo, K., & Kaschub, C. (2003). A practical guide to experience-sampling procedures. Journal of Happiness Studies, 4, 53–78. 270 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. Crowe, E., & Higgins, E. T. (1997). Regulatory focus and strategic inclinations: Promotion and prevention in decision-making. Organizational behavior and human decision processes, 69, 117–132. Chou, C.-P., & Bentler, P. M. (1995). Estimates and tests in structural equation modeling. In R. Hoyle (Ed.), Structural equation modeling: Issues, concepts, and applications (pp. 37-55). Newbury Park, CA: Sage. Chung, J. (2013). Studies in Leadership Indira Nooyi [Interview Transcript]. Vidyow.com, December 20: http://vidyow.com/video/studies-in-leadership-indra-no/uxoknwRvbtj/ Ciarocco, N. J., Sommer, K. L., & Baumeister, R. F. (2001). Ostracism and ego depletion: The strains of silence. Personality and Social Psychology Bulletin, 27, 1156–1163. Ciarocco, N., Twenge, J. M., Muraven, M., & Tice, D. M. (2007). The state self-control capacity scale: Reliability, validity, and correlations with physical and psychological stress. Unpublished manuscript. 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, 78–95. Converse, P. D., & DeShon, R. P. (2009). A tale of two tasks: Reversing the self-regulatory resource depletion effect. Journal of Applied Psychology, 94, 1318–1324. Collinson, D., & Hearn, J. (1994). Naming men as men: Implications for work, organization and management. Gender, Work & Organization, 1, 2–22. Corrigan, P. W., & Penn, D. L. (1999). Lessons from social psychology on discrediting psychiatric stigma. American Psychologist, 54, 765–776. Crescioni, A. W., & Baumeister, R. F. (2009). Alone and aggressive: Social exclusion impairs self-control and empathy and increases hostile cognition and aggression. Bullying, rejection, and peer victimization: A social cognitive neuroscience perspective, 251–277. Crocker, J. (2002). The costs of seeking self–esteem. Journal of Social Issues, 58, 597–615. Crocker, J., & Major, B. (1989). Social stigma and self-esteem: The self-protective properties of stigma. Psychological review, 96, 608–630. Csikszentmihalyi, M., & Hunter, J. (2003). Happiness in everyday life: The uses of experience sampling. Journal of happiness studies, 4, 185–199. 271 DeJordy, R. (2008). Just passing through: Stigma, passing, and identity decoupling in the work place. Group & Organization Management, 33, 504–531. Derogatis, L. R. (1977). The SCL-90 manual: Scoring, administration, and procedures. Baltimore: Johns Hopkins University School of Medicine, Clinical Psychometrics Unit. Derogatis, L. R., Lipman, R. S., Rickels, K., Uhlenhuth, E. H., & Covi, L. (1974). The Hopkins Symptom Checklist (HSCL): A self‐report symptom inventory. Systems Research and Behavioral Science, 19, 1–15. Derogatis, L. R., & Melisaratos, N. (1983). The Brief Symptom Inventory: An introductory report. Psychological Medicine, 13, 596–605. DeShon, R. P., & Gillespie, J. Z. (2005). A motivated action theory account of goal orientation. Journal of Applied Psychology, 90, 1096–1127. DeWall, C. N. (2010). Forming a basis for acceptance: Excluded people form attitudes to agree with potential affiliates. Social Influence, 5, 245–260. DeWall, C. N., & Baumeister, R. F. (2006). Alone but feeling no pain: Effects of social exclusion on physical pain tolerance and pain threshold, affective forecasting, and interpersonal empathy. Journal of Personality and Social Psychology: Attitudes and Social Cognition, 91, 1– 15. DeWall, C. N., Baumeister, R. F., Mead, N. L., & Vohs, K. D. (2011). How leaders self-regulate their task performance: evidence that power promotes diligence, depletion, and disdain. Journal of Personality and Social Psychology, 100, 47–65. De Witt Huberts, J. C., Evers, C., & De Ridder, D. T. (2014). “Because I Am Worth It” a theoretical framework and empirical review of a justification-based account of self-regulation failure. Personality and Social Psychology Review, 18, 119–138. Diefendorff, J. M., Erickson, R. J., Grandey, A. A., & Dahling, J. J. (2011). Emotional display rules as work unit norms: a multilevel analysis of emotional labor among nurses. Journal of occupational health psychology, 16(2), 170–186. Dovidio, J. F., Major, B., & Crocker, J. (2000). Stigma: introduction and overview. In T. F. Heatherton, R. E. Neck,M. R. Hebl, & J. G. Hull (Eds.), The social psychology of stigma (pp. 1– 28). Guilford Press. Eagly, A. H., & Karau, S. J. (2002). Role congruity theory of prejudice toward female leaders. Psychological Review, 109, 573–598. Earnshaw, V. A., Lang, S. M., Lippitt, M., Jin, H., & Chaudoir, S. R. (2015). HIV stigma and physical health symptoms: Do social support, adaptive coping, and/or identity centrality act as resilience resources?. AIDS and Behavior, 19, 41–49. 272 Eccles, J. S., & Wigfield, A. (2002). Motivational beliefs, values, and goals. Annual Review of Psychology, 53, 109–132. Ely, R., & Padavic, I. (2007). A feminist analysis of organizational research on sex differences. Academy of Management Review, 32, 1121–1143. Fakra, E., Hyde, L. W., Gorka, A., Fisher, P. M., Muñoz, K. E., Kimak, M., Halder, I., Ferrell, R.E., Manuck, S.B., & Hariri, A. R. (2009). Effects of HTR1A C (− 1019) G on amygdala reactivity and trait anxiety. Archives of General Psychiatry, 66, 33–40. Ferraro, K. F. (1996). Women's fear of victimization: Shadow of sexual assault?. Social Forces, 75, 667–690. Fisher, C. D., & To, M. L. (2012). Using experience sampling methodology in organizational behavior. Journal of Organizational Behavior, 33, 865–877. Fitzgerald, L. F. (1993). Sexual harassment: Violence against women in the workplace. American Psychologist, 48, 1070–1076. Fitzgerald, L.F., Drasgow, F., Hulin, C.L., Gelfand, M.J. & Magley, V.J. (1997). The antecedents and consequences of sexual harassment in organizations: a test of an integrated model. Journal of Applied Psychology, 82, 578–589. Fitzgerald, L. F., Magley, V. J., Drasgow, F., & Waldo, C. R. (1999). Measuring sexual harassment in the military: the sexual experiences questionnaire (SEQ—DoD). Military Psychology, 11, 243–263. Folkman, S., & Lazarus, R. S. (1988b). Manual for the Ways of Coping Scale. Palo Alto, CA: Consulting Psychology Press. Ford, M. E. (1992). Motivating humans: Goals, emotions, and personal agency beliefs. Sage Publications. Gailliot, M. T., & Baumeister, R. F. (2007). The physiology of willpower: Linking blood glucose to self-control. Personality and Social Psychology Review, 11, 303–327. Gailliot, M. T., Baumeister, R. F., DeWall, C. N., Maner, J. K., Plant, E. A., Tice, D., Brewer, L.E., Schmeichel, B. J. (2007). Self- control relies on glucose as a limited energy source: Willpower is more than a metaphor. Journal of Personality and Social Psychology, 92, 325–336. Gibson, A. (2016). Stop With the Questions Already! The Effects of Questionnaire Length and Monetary Incentives on Insufficient Effort Responding (Doctoral dissertation, Wright State University). 273 Goffman, E. (1963). Stigma: Notes on the management of spoiled identity. Englewood Clifls, NJ: Prentice-Hall. Goffman, E. (1974). Frame analysis: An essay on the organization of experience. Cambridge, MA: Harvard University Press. Gouin, J.P., Glaser, R., Malarkey, W. B., Beversdorf, D., & Kiecolt-Glaser, J. (2012). Chronic stress, daily stressors, and circulating inflammatory markers. Health Psychology, 31, 264–268. Hagger, M. S., Wood, C., Stiff, C., & Chatzisarantis, N. L. (2010). Ego depletion and the strength model of self-control: a meta-analysis. Psychological Bulletin, 136, 495–525. Halbert, C. (1997). Tough enough and woman enough: Stereotypes, discrimination, and impression management among women professional boxers. Journal of Sport and Social Issues, 21, 7–36. Harper, S. R. (2004). The measure of a man: Conceptualizations of masculinity among high- achieving African American male college students. Berkeley Journal of Sociology, 89–107. Harris, S.M. (1995). Psychosocial development and Black male masculinity: Implications for counseling economically disadvantaged African American male adolescents. Journal of Counseling and Development, 73, 279–287. Harris, C., Daniels, K., & Briner, R. B. (2003). A daily diary study of goals and affective well‐ being at work. Journal of Occupational and Organizational Psychology, 76, 401–410. Hassan, E. (2006). Recall bias can be a threat to retrospective and prospective research designs. The Internet Journal of Epidemiology, 3, 339–412. Hayes, A. F., & Preacher, K. J. (2014). Statistical mediation analysis with a multicategorical independent variable. British Journal of Mathematical and Statistical Psychology, 67, 451–470. Heatherton, T. F., & Polivy, J. (1991). Development and validation of a scale for measuring state self-esteem. Journal of Personality and Social psychology, 60, 895–910. Heilman, M. E., & Haynes, M. C. (2005). No credit where credit is due: attributional rationalization of women's success in male-female teams. Journal of Applied Psychology, 90, 905–916. Higgins, E. T. (1997). Beyond pleasure and pain. American Psychologist, 52, 1280–1300. Higgins, E. T. (1998). Promotion and prevention: Regulatory focus as a motivational principle. Advances in Experimental Social Psychology, 30, 1–46. Hitchcock, J. M., & Wilson, H. S. (1992). Personal risking: lesbian self-disclosure of sexual orientation to professional health care providers. Nursing Research, 41, 178–183. 274 Hmieleski, K. M., & Baron, R. A. (2008). Regulatory focus and new venture performance: A study of entrepreneurial opportunity exploitation under conditions of risk versus uncertainty. Strategic Entrepreneurship Journal, 2, 285–299. Hofmann, W., Baumeister, R. F., Förster, G., & Vohs, K. D. (2012). Everyday temptations: An experience sampling study of desire, conflict, and self-control. Journal of Personality and Social Psychology, 102, 1318–1335. Holmes, J. (2000). Politeness, power and provocation: How humour functions in the workplace. Discourse Studies, 2, 159–185. Holland, J., Ramazanoglu, C., Sharpe, S., & Thomson, R. (2004). The male in the head: Young people, heterosexuality and power. London: Tufnell. Holmes, J., & Schnurr, S. (2005). Politeness, humor, and gender in the workplace: negotiating norms and identifying contestation. Hoobler, J. M., Wayne, S. J., & Lemmon, G. (2009). Bosses' perceptions of family-work conflict and women's promotability: Glass ceiling effects. Academy of Management Journal, 52, 939– 957. Horton, J. J., & Chilton, L. B. (2010). The labor economics of paid crowdsourcing. In Proceedings of the 11th ACM Conference on Electronic Commerce, 209–218. Hox, J. (1998). Multilevel modeling: When and why. In Classification, data analysis, and data highways (pp. 147-154). Springer, Berlin, Heidelberg. http://www.danielsoper.com/statcalc/calculator.aspx?id=89 https://familystudiescenter.byu.edu/Documents/Reports/Multilevel-Mediation.pdf http://www.experiencesampler.com https://www.lifedatacorp.com/experience-sampling- app/?gclid=CMiTvPi12dQCFYiNswodyssH_w https://www.merriam-webster.com/dictionary/sexism http://www.statmodel.com/discussion/messages/23/1542.html?1403622520 http://quantpsy.org/pubs/syntax_appendix_081311.pdf Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural equation modeling: a multidisciplinary journal, 6, 1–55. 275 Hyman, O. (2001). Perceived social support and symptoms of secondary traumatic stress in disaster workers (Order No. 9999235). Available from ProQuest Dissertations & Theses Global. (304749136). Ilies, R., & Judge, T. A. (2002). Understanding the dynamic relationships among personality, mood, and job satisfaction: A field experience sampling study. Organizational Behavior and Human Decision Processes, 89, 1119–1139. Inzlicht, M., & Kang, S. K. (2010). Stereotype threat spillover: how coping with threats to social identity affects aggression, eating, decision making, and attention. Journal of Personality and Social Psychology, 99, 467–481. Inzlicht, M., McKay, L., & Aronson, J. (2006). Stigma as ego depletion how being the target of prejudice affects self-control. Psychological Science, 17, 262–269. Janssen, L., Fennis, B. M., Pruyn, A. T. H., & Vohs, K. D. (2008). The path of least resistance: Regulatory resource depletion and the effectiveness of social influence techniques. Journal of Business Research, 61, 1041–1045. Jefferson, K., Neilands, T. B., & Sevelius, J. (2013). Transgender women of color: discrimination and depression symptoms. Ethnicity and Inequalities in Health and Social Care, 6, 121–136. Jensen-Campbell, L. A., Knack, J. M., Waldrip, A. M., & Campbell, S. D. (2007). Do Big Five personality traits associated with self-control influence the regulation of anger and aggression?. Journal of Research in Personality, 41, 403–424. Jerusalem, M. and Schwarzer, R. (1992). Self-efficacy as a resource factor in stress appraisal processes. In R. Schwarzer (ed.), Self-efficacy: Thought control of action. Washington, DC: Hemisphere, 195–213. Jex, S. M., Bliese, P. D., Buzzell, S., & Primeau, J. (2001). The impact of self-efficacy on stressor–strain relations: Coping style as an explanatory mechanism. Journal of Applied Psychology, 86, 401–409. John, O. P., Naumann, L. P., & Soto, C. J. (2008). Paradigm shift to the integrative big five trait taxonomy. Handbook of Personality: Theory and Research, 3, 114–158. Johnson, R. E., & Chang, C.-H. (2008b). Development and validation of a work-based regulatory focus scale. Paper presented at the 23rd Annual Society for Industrial and Organizational Psychology Conference, San Francisco. Jones, K. P., & King, E. B. (2014). Managing concealable stigmas at work: A review and multilevel model. Journal of Management, 40, 1466–1494. Judge, T. A., & Ilies, R. (2002). Relationship of personality to performance motivation: A meta- 276 analytic review. Journal of Applied Psychology, 87, 797–807. Kane, M. J., Brown, L. H., McVay, J. C., Silvia, P. J., Myin-Germeys, I., & Kwapil, T. R. (2007). For whom the mind wanders, and when: An experience-sampling study of working memory and executive control in daily life. Psychological Science, 18, 614–621. Kanfer, R. (1990). Motivation theory and industrial and organizational psychology. Handbook of industrial and organizational psychology, 1, 75–130. Karoly, P. (1993). Mechanisms of self-regulation: A systems view. Annual Review of Psychology, 44, 23–52. Kerber, L. K. (1988). Separate spheres, female worlds, woman's place: The rhetoric of women's history. The Journal of American History, 75, 9–39. King, E. B., Mohr, J. J., Peddie, C. I., Jones, K. P., & Kendra, M. (2017). Predictors of identity management: An exploratory experience-sampling study of lesbian, gay, and bisexual workers. Journal of Management, 43, 476–502. 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, 566– 601. Klar, M., & Kasser, T. (2009). Some Benefits of Being an Activist: Measuring Activism and Its Role in Psychological Well‐Being. Political Psychology, 30, 755–777. Knouse, L. E., Mitchell, J. T., Brown, L. H., Silvia, P. J., Kane, M. J., Myin-Germeys, I., & Kwapil, T. R. (2008). The expression of adult ADHD symptoms in daily life: an application of experience sampling methodology. Journal of Attention Disorders, 11, 652–663. Kool, W., McGuire, J. T., Rosen, Z. B., & Botvinick, M. M. (2010). Decision making and the avoidance of cognitive demand. Journal of Experimental Psychology: General, 139, 665–682. Koole, S. L., Tops, M., Strübin, S., Bouw, J., Schneider, I. K., & Jostmann, N. B. (2014). The ego fixation hypothesis: Involuntary persistence of self-control. The control within: Motivation and its Regulation, 95–112. Koss, M. P. (1990). Changed lives: The psychological impact of sexual harassment. Ivory Power: Sexual Harassment on Campus, 73–92. Koss, M. P. (2000). Blame, shame, and community: Justice responses to violence against women. American Psychologist, 55, 1332–1343. Klonoff, E. A., & Landrine, H. (1995). The schedule of sexist events. Psychology of Women Quarterly, 19, 439–470. 277 Kira, I. A., Hanaa, S., & Bujold-Bugeaud, M. (2015). Gender discrimination (GD): A conceptual trauma-based framework for GD and the development of Gender Discrimination Inventory. Psychology, 6, 2041–2070. Kitada, M. (2013). Code of behaviour at sea: Women seafarers’ shipboard identity management. WMU Journal of Maritime Affairs, 12, 213–227. Kulik, C. T., Bainbridge, H. T., & Cregan, C. (2008). Known by the company we keep: Stigma- by-association effects in the workplace. Academy of Management Review, 33, 216–230. Lach, D., Ingram, H., & Rayner, S. (2004). Maintaining the status quo: how institutional norms and practices create conservative water organizations. Tex L. Rev. 83, 2027–2053. LaPiere, R. T. (1934). Attitudes vs. actions. Social Forces, 13, 230–237. Leary, M. R., & Kowalski, R. M. (1990). Impression management: A literature review and two- component model. Psychological bulletin, 107, 34–47. Lee, A. Y., & Aaker, J. L. (2004). Bringing the frame into focus: the influence of regulatory fit on processing fluency and persuasion. Journal of personality and social psychology, 86(2), 205- 218. Lee, D. L., & Ahn, S. (2013). The relation of racial identity, ethnic identity, and racial socialization to discrimination–distress: A meta-analysis of Black Americans. Journal of Counseling Psychology, 60, 1–14 . Lee, K., Kim, E., Bhave, D. P., & Duffy, M. K. (2016). Why victims of undermining at work become perpetrators of undermining: An integrative model. Journal of Applied Psychology, 101, 915–924. Lian, H., Brown, D. J., Ferris, D. L., Liang, L. H., Keeping, L. M., & Morrison, R. (2014). Abusive supervision and retaliation: A self-control framework. Academy of Management Journal, 57, 116–139. Liberman, N., Idson, L. C., Camacho, C. J., & Higgins, E. T. (1999). Promotion and prevention choices between stability and change. Journal of Personality and Social Psychology, 77, 1135– 1145 Liberman, N., Molden, D. C., Idson, L. C., & Higgins, E. T. (2001). Promotion and prevention focus on alternative hypotheses: implications for attributional functions. Journal of Personality and Social Psychology, 80, 5–18. Lin, B. C. (2013). "Do Not Disturb": A Micro-Macro Examination of Intrusions at Work. Lin, W., Ma, J., Wang, L., & Wang, M. (2015). A double‐edged sword: The moderating role of 278 conscientiousness in the relationships between work stressors, psychological strain, and job performance. Journal of Organizational Behavior, 36, 94–111. Locke, E. A., & Latham, G. P. (1990). A theory of goal setting & task performance. Prentice- Hall, Inc. Lockwood, P., Marshall, T. C., & Sadler, P. (2005). Promoting success or preventing failure: Cultural differences in motivation by positive and negative role models. Personality and Social Psychology Bulletin, 31, 379–392. Lord, R. G., Diefendorff, J. M., Schmidt, A. M., & Hall, R. J. (2010). Self-regulation at work. Annual Review of Psychology, 61, 543–568. Loutsiou-Ladd, A., Panayiotou, G., & Kokkinos, C. M. (2008). A review of the factorial structure of the Brief Symptom Inventory (BSI): Greek evidence. International Journal of Testing, 8, 90–110. Lu, N., Han, Y., Chen, T., Gunzler, D. D., Xia, Y., Lin, J. Y., & Tu, X. M. (2013). Power analysis for cross-sectional and longitudinal study designs. Shanghai Archives of Psychiatry, 25, 259–262. 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, 678–704. Lyons, B. J., Wessel, J. L., Tai, Y. C., & Ryan, A. M. (2014). Strategies of job seekers related to age-related stereotypes. Journal of Managerial Psychology, 29, 1009–1027. MacCallum, R. C., Browne, M. W., & Sugawara, H. M. (1996). Power analysis and determination of sample size for covariance structure modeling. Psychological Methods, 1, 130– 149. Maier, G. W., & Brunstein, J. C. (2001). The role of personal work goals in newcomers' job satisfaction and organizational commitment: A longitudinal analysis. Journal of Applied Psychology, 86, 1034–1042. Major, V. S. (2004). Pregnancy in the workplace: Stigmatization and work identity management among pregnant employees (Doctoral dissertation). Mathews, A., & MacLeod, C. (1985). Selective processing of threat cues in anxiety states. Behaviour Research and Therapy, 23, 563–569. McAdams, D. P. (1996). Personality, modernity and the storied self: A contemporary framework for studying persons. Psychological Inquiry, 7, 295–321. 279 McAdams, D. P. (1999). Personal narratives and the life story. In L. Pervin & O. John (Eds.), Handbook of personality: Theory and research (2nd ed.): 478–500. New York: Guilford Press. 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, 35–62. Meisenbach, R. J. (2010). Stigma management communication: A theory and agenda for applied research on how individuals manage moments of stigmatized identity. Journal of Applied Communication Research, 38, 268–292. Mendoza-Denton, R., Downey, G., Purdie, V. J., Davis, A., & Pietrzak, J. (2002). Sensitivity to status-based rejection: implications for African American students' college experience. Journal of Personality and Social Psychology, 83, 896–918. Miller, C. C., Cardinal, L. B., & Glick, W. H. (1997). Retrospective reports in organizational research: A reexamination of recent evidence. Academy of Management Journal, 40, 189–204. Miller, C. T., & Major, B. (2000). Coping with stigma and prejudice. In: Heatherton TF, Kelck RE, Hebl MR, Hull JG, editors. The Social Psychology of Stigma. New York: Guilford Press; 243–272. Minikel-Lacocque, J. (2013). Racism, college, and the power of words: Racial microaggressions reconsidered. American Educational Research Journal, 50, 432–465. Mohr, J., & Fassinger, R. (2000). Measuring dimensions of lesbian and gay male experience. Measurement and Evaluation in Counseling and Development, 33, 66–66. Morgan, L. M. (2002). The nature, antecedents, and consequences of social identity-based impression management: Uncovering strategies for professional image construction in cases of negative stereotyping. Unpublished doctoral dissertation, University of Michigan, Ann Arbor, Michigan. Mosher, D. L. (1991). Macho men, machismo, and sexuality. Annual Review of Sex Research, 2, 199–247. Muraven, M., & Slessareva, E. (2003). Mechanisms of self-control failure: Motivation and limited resources. Personality and Social Psychology Bulletin, 29, 894–906. Muraven, M., Tice, D. M., & Baumeister, R. F. (1998). Self-control as a limited resource: Regulatory depletion patterns. Journal of Personality and Social Psychology, 74, 774–789. Mussweiler, T., Gabriel, S., & Bodenhausen, G. V. (2000). Shifting social identities as a strategy for deflecting threatening social comparisons. Journal of Personality and Social Psychology, 79, 398–409. 280 Napa Scollon, C., Prieto, C. K., & Diener, E. (2009). Experience sampling: promises and pitfalls, strength and weaknesses. Assessing well-being, 157–180. National Institutes of Health. Outreach notebook for the NIH guidelines on inclusion of women and minorities as subjects in clinical research. Bethesda, MD: Author. Nelson, A., Cooper, C. L., & Jackson, P. R. (1995). Uncertainty amidst change: The impact of privatization on employee job satisfaction and well‐being. Journal of Occupational and Organizational psychology, 68, 57–71. Nes, L. S., Carlson, C. R., Crofford, L. J., de Leeuw, R., & Segerstrom, S. C. (2011). Individual differences and self-regulatory fatigue: optimism, conscientiousness, and self-consciousness. Personality and Individual Differences, 50, 475–480. Neubert, M. J., Kacmar, K. M., Carlson, D. S., Chonko, L. B., & Roberts, J. A. (2008). Regulatory focus as a mediator of the influence of initiating structure and servant leadership on employee behavior. Journal of Applied Psychology, 93, 1220–1233. Nezlek, J. B. (2001). Multilevel random coefficient analyses of event-and interval-contingent data in social and personality psychology research. Personality and Social Psychology Bulletin, 27, 771–785. Ng, R. (1993). " A Woman out of Control": Deconstructing Sexism and Racism in the University. Canadian Journal of Education/Revue Canadienne de l'education, 189–205. Nielsen, K., & Cleal, B. (2011). Under which conditions do middle managers exhibit transformational leadership behaviors?—An experience sampling method study on the predictors of transformational leadership behaviors. The Leadership Quarterly, 22, 344–352. Nikolaou, I., Vakola, M., & Bourantas, D. (2008). Who speaks up at work? Dispositional influences on employees' voice behavior. Personnel Review, 37, 666–679. Nolen-Hoeksema, S., Larson, J., & Grayson, C. (1999). Explaining the gender difference in depressive symptoms. Journal of Personality and Social Psychology, 77, 1061–1072. Norris, J., Nurius, P. S., & Dimeff, L. A. (1996). Through her eyes: Factors Affecting Women’s Perception of and Resistance to Acquaintance Sexual Aggression Threat. Psychology of Women Quarterly, 20(1), 123–145. Padin, A. C. (2016). Implicit attitudes, physical activity and self-regulatory capacity (Doctoral dissertation, The Ohio State University). Pascoe, E. A., & Smart Richman, L. (2009). Perceived discrimination and health: a meta-analytic review. Psychological Bulletin, 135, 531–554. 281 Pateman, C. (1989). The disorder of women: Democracy, feminism, and political theory. Stanford University Press. Pelled, L. H., Eisenhardt, K. M., & Xin, K. R. (1999). Exploring the black box: An analysis of work group diversity, conflict and performance. Administrative Science Quarterly, 44: 1–28. Petriglieri, J. L. (2011). Under threat: Responses to and the consequences of threats to individuals' identities. Academy of Management Review, 36, 641–662. Phelan, Julie E., Corinne A. Moss‐Racusin, and Laurie A. Rudman (2008). "Competent yet out in the cold: Shifting criteria for hiring reflect backlash toward agentic women." Psychology of Women Quarterly, 32, 406–413. Phillips, K. W., Rothbard, N. P., & Dumas, T. L. (2009). To disclose or not to disclose? Status distance and self-disclosure in diverse environments. Academy of Management Review, 34, 710– 732. Pinel, E. C. (1999). Stigma consciousness: the psychological legacy of social stereotypes. Journal of Personality and Social Psychology, 76, 114–128. Podsakoff, N. P., LePine, J. A., & LePine, M. A. (2007). Differential challenge stressor- hindrance stressor relationships with job attitudes, turnover intentions, turnover, and withdrawal behavior: a meta-analysis. Journal of Applied Psychology, 92, 438–454. Porter, S. R., & Whitcomb, M. E. (2005). Non-response in student surveys: The role of demographics, engagement and personality. Research in Higher Education, 46, 127–152. Posner, J., Russell, J. A., & Peterson, B. S. (2005). The circumplex model of affect: An integrative approach to affective neuroscience, cognitive development, and psychopathology. Development and Psychopathology, 17, 715–734. Pringle, J. K. (2008). Gender in management: theorizing gender as heterogender. British Journal of Management, 19(s1), S110-S119. Pronin, E., Steele, C. M., & Ross, L. (2004). Identity bifurcation in response to stereotype threat: Women and mathematics. Journal of Experimental Social Psychology, 40(2), 152-168. Pryor, J. B., Reeder, G. D., Yeadon, C., & Hesson-McInnis, M. (2004). A dual-process model of reactions to perceived stigma. Journal of personality and social psychology, 87(4), 436 – 452. Quinn, D. M., & Chaudoir, S. R. (2009). Living With a Concealable Stigmatized Identity: The Impact of Anticipated Stigma, Centrality, Salience, and Cultural Stigma on Psychological Distress and Health. Journal of Personality and Social Psychology, 97, 634–651. Quintana, S. M., & Maxwell, S. E. (1999). Implications of recent developments in structural equation modeling for counseling psychology. The Counseling Psychologist, 27, 485–527. 282 Rafaeli, A., & Sutton, R. I. (1987). Expression of emotion as part of the work role. Academy of Management Review, 12, 23–37. Ragins, B. R. (2008). Disclosure disconnects: Antecedents and consequences of disclosing invisible stigmas across life domains. Academy of Management Review, 33, 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, 1244–1261. Raman, P. (2008). Achieving positive social identity: Women's coping strategies in response to status inequality in television portrayals. The University of Arizona. Reise, S. P., Ventura, J., Nuechterlein, K. H., & Kim, K. H. (2005). An illustration of multilevel factor analysis. Journal of Personality Assessment, 84, 126–136. Repetti, R.L. (1993). Short term effects of occupational stressors on daily mood and health complaints. Health Psychology, 12, 125–131. Reskin, B. F. (1988). Bringing the men back in: Sex differentiation and the devaluation of women's work. Gender & Society, 2, 58–81. Richman, L. S., Bennett, G. G., Pek, J., Siegler, I., & Williams, R. B. (2007). Discrimination, dispositions, and cardiovascular responses to stress. Health Psychology, 26, 675–683. Richman, L. S., & Leary, M. R. (2009). Reactions to Discrimination, Stigmatization, Ostracism, and Other Forms of Interpersonal Rejection: A Multimotive Model. Psychological Review, 116, 365–383. Ridgeway, C. L., & Correll, S. J. (2004). Unpacking the gender system: A theoretical perspective on gender beliefs and social relations. Gender & Society, 18, 510–531. Roberts, L. M. (2005). Changing faces: Professional image construction in diverse organizational settings. Academy of Management Review, 30, 695–711. Roberts, L. M., Cha, S. E., & Kim, S. S. (2014). Strategies for managing impressions of racial identity in the workplace. Cultural Diversity and Ethnic Minority Psychology, 20, 529–540. Roberts, L. M., Settles, I. H., & Jellison, W. A. (2008). Predicting the strategic identity management of gender and race. Identity: An International Journal of Theory and Research, 8, 269–306. Robnett, R. (2015). Girls’ and Women’s Preferred Methods of Coping with Gender Bias in STEM. International Journal of Gender, Science and Technology, 8, 19–37. 283 Rosette, A. S., & Tost, L. P. (2010). Agentic women and communal leadership: How role prescriptions confer advantage to top women leaders. Journal of Applied Psychology, 95, 221– 235. Rucker, L. S., West, L. M., & Roemer, L. (2010). Relationships among perceived racial stress, intolerance of uncertainty, and worry in a black sample. Behavior Therapy, 41, 245–253. Rudman, L. A., & Glick, P. (2001). Prescriptive gender stereotypes and backlash toward agentic women. Journal of Social Issues, 57, 743–762. Russell, J. A. (1980). A circumplex model of affect. Journal of Personality and Social Psychology, 39, 1161–1178. Ryan, W. (1976). Blaming the victim (Vol. 226). Vintage. Sable, M. R., Danis, F., Mauzy, D. L., & Gallagher, S. K. (2006). Barriers to reporting sexual assault for women and men: Perspectives of college students. Journal of American College Health, 55, 157–162. Salvatore, J., & Shelton, J. N. (2007). Cognitive costs of exposure to racial prejudice. Psychological science, 18(9), 810-815. Schmader, T., & Johns, M. (2003). Converging evidence that stereotype threat reduces working memory capacity. Journal of personality and social psychology, 85(3), 440- 452. Schmader, T., Johns, M., & Forbes, C. (2008). An integrated process model of stereotype threat effects on performance. Psychological review, 115(2), 336- 356. Schmidt, A. M., Dolis, C. M., & Tolli, A. P. (2009). A matter of time: individual differences, contextual dynamics, and goal progress effects on multiple-goal self-regulation. Journal of Applied Psychology, 94, 692–709. Schwartz, J. E., & Stone, A. A. (1998). Strategies for analyzing ecological momentary assessment data. Health Psychology, 17, 6–16. Schwarzer, R. and Jerusalem, M. (1995). Generalized Self-Efficacy scale. In Weinman J., Wright S. and Johnston M., Measures in health psychology: A user’s portfolio. Causal and control beliefs Windsor, UK: NFER-NELSON, 35–37. Schwarzer, R., Mueller, J., & Greenglass, E. (1999). Assessment of perceived general self- efficacy on the Internet: Data collection in cyberspace. Anxiety, Stress and Coping, 12, 145–161. Sedikides, C., Gaertner, L., & Toguchi, Y. (2003). Pancultural self-enhancement. Journal of Personality and Social Psychology, 84, 60–79. 284 Sellers, R. M., Rowley, S. A., Chavous, T. M., Shelton, J. N., & Smith, M. A. (1997). Multidimensional Inventory of Black Identity: A preliminary investigation of reliability and construct validity. Journal of Personality and Social Psychology, 73, 805–815. Sellers, R. M., & Shelton, J. N. (2003). The role of racial identity in perceived racial discrimination. Journal of Personality and Social Psychology, 84, 1079–1092. Sellers, R. M., Smith, M. A., Shelton, J. N., Rowley, S. A., & Chavous, T. M. (1998). Multidimensional model of racial identity: A reconceptualization of African American racial identity. Personality and Social Psychology Review, 2, 18–39. Settles, I. H. (2004). When multiple identities interfere: The role of identity centrality. Personality and Social Psychology Bulletin, 30, 487–500. Shaver, K. G. (1970). Defensive attribution: Effects of severity and relevance on the responsibility assigned for an accident. Journal of Personality and Social Psychology, 14, 101– 113. Shih, M. (2004). Positive stigma: Examining resilience and empowerment in overcoming stigma. The ANNALS of the American Academy of Political and Social Science, 591, 175–185. Shih, M., Young, M. J., & Bucher, A. (2013). Working to reduce the effects of discrimination: Identity management strategies in organizations. American Psychologist, 68, 145–157. Skrla, L., Reyes, P., & Scheurich, J. J. (2000). Sexism, silence, and solutions: Women superintendents speak up and speak out. Educational Administration Quarterly, 36, 44–75. Skrondal, A., & Rabe-Hesketh, S. (2003). Multilevel logistic regression for polytomous data and rankings. Psychometrika, 68, 267–287. Smith, R. A. (2007). Language of the lost: An explication of stigma communication. Communication Theory, 17, 462–485. Smith, R. W. (2002). Effects of relaxation on self -regulatory depletion. Dissertation. Smith-Rosenberg, C. (1972). The hysterical woman: Sex roles and role conflict in 19th-century America. Social Research, 652–678. Ståhl, T., Van Laar, C., & Ellemers, N. (2012). The role of prevention focus under stereotype threat: Initial cognitive mobilization is followed by depletion. Journal of Personality and Social Psychology, 102, 1239–1251. Steele, C.M., & Aronson, J. (1995). Stereotype threat and the intellectual test performance of African Americans. Journal of Personality and Social Psychology, 69, 797–811. 285 Steele, C. M., Spencer, S. J., & Aronson, J. (2002). Contending with group image: The psychology of stereotype and social identity threat. In Advances in Experimental Social Psychology, 34 (pp. 379-440). Academic Press. Steinel, W., Van Kleef, G. A., Van Knippenberg, D., Hogg, M. A., Homan, A. C., & Moffitt, G. (2010). How intragroup dynamics affect behavior in intergroup conflict: The role of group norms, prototypicality, and need to belong. Group Processes & Intergroup Relations, 13, 779– 794. Stryker, S., & Serpe, R. T. (1994). Identity salience and psychological centrality: Equivalent, overlapping, or complementary concepts?. Social Psychology Quarterly, 16–35. Stuber, J., Meyer, I., & Link, B. (2008). Stigma, prejudice, discrimination and health. Social science & medicine (1982), 67, 351–357. Suárez Riveiro, J. M., Cabanach, R. G., & Arias, A. V. (2001). Multiple‐goal pursuit and its relation to cognitive, self‐regulatory, and motivational strategies. British Journal of Educational Psychology, 71, 561–572. Sue, D. W., Capodilupo, C. M., Torino, G. C., Bucceri, J. M., Holder, A., Nadal, K. L., & Esquilin, M. (2007). Racial microaggressions in everyday life: implications for clinical practice. American psychologist, 62, 271–286. Sumer, N., Karanci, A. N., Berument, S. K., & Gunes, H. (2005). Personal resources, coping self‐efficacy, and quake exposure as predictors of psychological distress following the 1999 earthquake in Turkey. Journal of Traumatic Stress, 18, 331–342. Swann, W. B., & Read, S. J. (1981). Self-verification processes: How we sustain our self- conceptions. Journal of Experimental Social Psychology, 17, 351–372. Swann, W. B. (1983). Self-verification: Bringing social reality into harmony with the self. In J. Suls & A. G. Greenwald (Eds). Swann, W. B. (1987). Identity negotiation: Where two roads meet. Journal of Personality and Social Psychology, 53, 1038–1051. Swim, J. K., & Thomas, M. A. (2006). Responding to everyday discrimination: A synthesis of research on goal-directed, self-regulatory coping behaviors. Stigma and group inequality: Social psychological perspectives, 105–126. Szymanski, D. M., & Owens, G. P. (2009). Group-level coping as a moderator between heterosexism and sexism and psychological distress in sexual minority women. Psychology of Women Quarterly, 33, 197–205. 286 Tavakoli, S., Lumley, M. A., Hijazi, A. M., Slavin-Spenny, O. M., & Parris, G. P. (2009). Effects of Assertiveness Training and Expressive Writing on Acculturative Stress in International Students: A Randomized Trial. Journal of Counseling Psychology, 56, 590–596. Thau, S., Aquino, K., & Poortvliet, P. M. (2007). Self-defeating behaviors in organizations: The relationship between thwarted belonging and interpersonal work behaviors. Journal of Applied Psychology, 92: 840–847. Thomas, V., & Azmitia, M. (2016). Tapping into the app: Updating the experience sampling method for the 21st century. Emerging Adulthood, 4, 60–67. Thoits, P. A. (1991). On merging identity theory and stress research. Social Psychology Quarterly, 101–112. Thompson, S. C. (1999). Illusions of control: How we overestimate our personal influence. Current Directions in Psychological Science, 8, 187–190. Tofighi, D., & Thoemmes, F. (2014). Single-level and multilevel mediation analysis. The Journal of Early Adolescence, 34, 93–119. Trawalter, S., & Richeson, J. A. (2006). Regulatory focus and executive function after interracial interactions. Journal of Experimental Social Psychology, 42, 406–412. Turchik, J. A., McLean, C., Rafie, S., Hoyt, T., Rosen, C. S., & Kimerling, R. (2013). Perceived barriers to care and provider gender preferences among veteran men who have experienced military sexual trauma: A qualitative analysis. Psychological Services, 10, 213–222. Vadillo, M. A., Gold, N., & Osman, M. (2016). The bitter truth about sugar and willpower: The limited evidential value of the glucose model of ego depletion. Psychological Science, 27(9), 1207-1214. VanDellen, M. R., Hoyle, R. H., & Miller, R. (2012). The Regulatory Easy Street: Self- Regulation Below the Self-Control Threshold Does not Consume Regulatory Resources. Personality and Individual Differences, 52, 898–902. Velez, B. L., Moradi, B., & Brewster, M. E. (2013). Testing the tenets of minority stress theory in workplace contexts. Journal of Counseling Psychology, 60, 532–542. Wagner, D. D., & Heatherton, T. F. (2012). Self-regulatory depletion increases emotional reactivity in the amygdala. Social cognitive and affective neuroscience, 8(4), 410-417. Wagner, D. T., Barnes, C. M., Lim, V. K., & Ferris, D. L. (2012). Lost sleep and cyberloafing: Evidence from the laboratory and a daylight saving time quasi-experiment. Journal of Applied Psychology, 97, 1068–1076. Walker, R. (2001). Becoming the third wave. Identity Politics in the Women's Movement, 78–87. 287 Walster, E. (1966). Assignment of responsibility for an accident. Journal of Personality and Social Psychology, 3, 73–79. Wang, L., Tao, T., Fan, C., Gao, W., & Wei, C. (2015). The Influence of Chronic Ego Depletion on Goal Adherence: An Experience Sampling Study. PloS one, 10, e0142220. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: the PANAS scales. Journal of Personality and Social Psychology, 54, 1063–1070. Weaver, J. R., Vandello, J. A., Bosson, J. K., & Burnaford, R. M. (2010). The proof is in the punch: Gender differences in perceptions of action and aggression as components of manhood. Sex Roles, 62, 241–251. Weiss, K. G. (2010). Too ashamed to report: Deconstructing the shame of sexual victimization. Feminist Criminology, 5, 286–310. Welsh, D. T., Ellis, A. P., Christian, M. S., & Mai, K. M. (2014). Building a self-regulatory model of sleep deprivation and deception: The role of caffeine and social influence. Journal of Applied Psychology, 99, 1268–1277. Welsh, D. T. (2014). Extending the Self-Regulatory Model Linking High Goals and Unethical Behavior: The Moderating Effects of Goal Commitment and Subconscious Priming. (Doctoral Dissertation, The University of Arizona). West, C., & Zimmerman, D. H. (1987). Doing gender. Gender & society, 1, 125–151. Weston, R., & Gore Jr, P. A. (2006). A brief guide to structural equation modeling. The Counseling Psychologist, 34, 719–751. Wheeler, S.C. & Petty, R.E. (2001). The effects of stereotype activation on behavior: A review of possible mechanisms. Psychological Bulletin, 127, 797–826. William R., Shadish, Cook, T. D., & Campbell, D. T. (2002). Experimental and quasi- experimental designs for generalized causal inference. Wadsworth Cengage learning. Wright, M. O. D., Masten, A. S., & Narayan, A. J. (2013). Resilience processes in development: Four waves of research on positive adaptation in the context of adversity. In Handbook of resilience in children (pp. 15-37). Springer, Boston, MA. Yang, J., Mossholder, K. W., & Peng, T. K. (2007). Procedural justice climate and group power distance: An examination of cross-level interaction effects. Journal of Applied Psychology, 92, 681–692 288