AN INVESTIGATION OF INEFFECTIVE ALLY BEHAVIORS By Lauren A. Collier A THESIS Submitted to Michigan State University i n partial fulfillment of the requirement for the degree of Psychology M aster of Arts 201 9 ABSTRACT AN INVESTIGATION OF INEF F ECTIVE ALLY BEHAVIORS By Lauren A. Collier Previous investigations of allyship more broadly make the assumption that allies are unilaterally helpful. Through conducting three survey studies, I aimed to (1) examine the effects of effective and ineffective ally behavior on psychological outcomes for members of marginalized groups, (2) examine the perceptions marginalized group members had of effective and ineffective allies, (3) determine whether there were specific ally motivations that predict effective and ineffective allyship behavior and (4) determine whether there were specific individual differences that predict effective ally behavior. The results revealed that ineffectiv e ally behavior was negatively related to psychological safety and positive affect and positively related to anxiety and negative affect for marginalized group members. A n inverse set of relationships was revealed for effective ally behaviors. In addition, internal motivation to respond without prejudice and social dominance orientation emerged as robust predictors of ally behavior. Specifically, internal motivation to respond without prejudice was negatively related to ineffective ally behaviors while soci al dominance orientation was positively related. These variables exhibited an inverse relationship with effective ally behavior. Implications, limitations and future directions are discussed. Keywords: allyship, ally, social dominance orientation. iii This work is dedicated to my ally, Dr. Grey Gundaker who saw me as I was during my freshman year and , at the same time, s aw me as who I will be 20 years from now. Her encouragement and real talk ignited my passion for research and ultimately led me to pursue higher education. I am forever thankful for her. iv ACKNOWLEDGEMENT S I am grateful to so many people for their patience , guidance and support during the thesis process and during this segment of the graduate school adventure. First, I would like to thank my advisor, Ann Marie Ryan for her mentorship, guidance and extremely thoughtful feedback on this project and on a pletho ra of other projects in the works. Beyond incredible scholarship she is also an amazing person who continues to be an exemplar of mentorship to myself and others . I would also like to thank my thesis committee members, Fred Leong and Chris Nye, for their t houghtful feedback on my proposal and for helping make this project even better. I n addition, I would also like to thank the members of the Michigan State University Diversity Lab Group for their continued support and feedback on a significant portion of m y research endeavors. Specifically, I would like to thank my friends Courtney Bryant and Caitlin Briggs for always providing feedback on my projects in Diversity Lab . I would also like to thank Cindy Grammatico and Justin Duby for their excellent work as r esearch assistants this summer. Last, but certainly not least, I would like to thank my husband Rodwin Spruel for his continued love, friendship and support through out our decade together. v TABLE OF CONTENTS LIST OF TABLES ... ................................ ................................ ................................ ........ v i i i LIST OF FIGURES ................................ ................................ ................................ ............. x INTRODUCTION ................................ ................................ ................................ ............... 1 Allies and Allyship Defined ................................ ................................ ..................... 4 Comparison of Allyship to Other Related Constructs ................................ ............. 7 Ally Motivation ................................ ................................ ................................ ........ 9 Ally Individual Differences ................................ ................................ ................... 15 Factors Affecting Marginalized Group Members Perceptions of Ally Behaviors 23 Outcomes of Allyship ................................ ................................ ............................ 26 The Ramifications of Ineffective Allyship ................................ ............................ 31 Other Measures of Allyship ................................ ................................ ................... 33 The Present Investigation ................................ ................................ ....................... 36 Study Rationale and Development of Hypotheses ................................ ................. 3 8 STUDY 1 METHOD ................................ ................................ ................................ ......... 4 6 Study Purpose ................................ ................................ ................................ ........ 4 6 Participan ts ................................ ................................ ................................ ............. 4 6 Measures ................................ ................................ ................................ ................ 4 8 Procedure ................................ ................................ ................................ ............... 5 6 STUDY 1 RESULTS ................................ ................................ ................................ ......... 5 7 Data Cleaning ................................ ................................ ................................ ........ 5 7 Descriptive Information ................................ ................................ ......................... 5 8 Qualitative Data Analyses ................................ ................................ ...................... 61 Quantitative Data Analyses ................................ ................................ .................... 63 STUDY 1 DISCUSSION ................................ ................................ ................................ ... 7 9 STUDY 2 METHOD ................................ ................................ ................................ ......... 8 1 Study Purpose ................................ ................................ ................................ ....... 8 1 Item Generation ................................ ................................ ................................ ..... 8 1 Participants ................................ ................................ ................................ ............. 8 4 Measures ................................ ................................ ................................ ................ 8 5 Procedure ................................ ................................ ................................ ............... 8 6 STUDY 2 RESULTS ................................ ................................ ................................ ......... 8 7 Data Cleaning ................................ ................................ ................................ ......... 8 7 Qualitative Data Analyses ................................ ................................ ...................... 8 7 Quantitative Data Analyses ................................ ................................ .................... 8 8 Item Selection for the Final Scale ................................ ................................ .......... 8 9 vi STUDY 2 DISCUSSION ................................ ................................ ................................ ... 9 4 STUDY 3 METHOD ................................ ................................ ................................ ......... 9 5 Study Purpose ................................ ................................ ................................ ........ 9 5 Participants ................................ ................................ ................................ ............. 9 5 Measures ................................ ................................ ................................ ................ 9 6 Procedure ................................ ................................ ................................ ............. 10 6 STUDY 3 RESULTS ................................ ................................ ................................ ....... 10 7 Data Cleaning ................................ ................................ ................................ ...... 10 7 Descriptive Information ................................ ................................ ....................... 10 7 Qualitative Coding Information ................................ ................................ ........... 1 11 Quantitative Data Analyses ................................ ................................ .................. 1 1 3 STUDY 3 DISCUSSION ................................ ................................ ................................ . 1 2 8 GENERAL DISCUSSION ................................ ................................ .............................. 12 9 Effects Allyship Behavior on Marginalized Group Members ............................. 1 30 Characteristics Related to Effective and Ineffective Ally Behaviors .................. 1 31 Findings from Preliminary Scale Development ................................ ................... 1 3 4 Limitations ................................ ................................ ................................ ........... 1 3 5 Future Research ................................ ................................ ................................ ... 13 9 CONCLUSION ................................ ................................ ................................ ................ 1 4 5 APPENDICES ................................ ................................ ................................ ................. 1 4 6 APPENDIX A: Ally Manipulations and Questions for Study 1 .......................... 1 4 7 APPENDIX B: Perception of Ally Characteristics Scale (PACS) ...................... 1 4 8 APPENDIX C: Trust Scale ................................ ................................ ................. .1 4 9 APPENDIX D: Positive and Negative Affect Schedule (PANAS) .................... .1 50 APPENDIX E: Voice Questions ................................ ................................ ......... .1 51 APPENDIX F: Stigma Consciousness ................................ ................................ .1 52 APPENDIX G: Identity Centrality ................................ ................................ ..... .1 53 APPENDIX H: Perceived Discrimination ................................ .......................... .1 5 4 APPENDIX I: Demographic Questions ................................ .............................. .1 5 5 APPENDIX J : Consent Forms for All Studies ................................ ................... .1 64 APPENDIX K : Debriefing Forms for All Studies ................................ .............. .1 60 APPENDIX L: Items Presented to Participants in Study 2 ................................ .16 6 APPENDIX M: Majority Group Perceptions of Ally Behaviors ........................ .1 6 9 APPENDIX N: Social Dominance Orientation ................................ .................. .1 70 APPENDIX O: Motivation to Respond Without Prejudice ................................ .1 71 APPENDIX P: Prosocial Motivations Scale ................................ ....................... .1 72 APPENDIX Q: The Selfishness Questionnaire ................................ .................. .1 73 APPENDIX R: Short Form HEXACO (Honesty - Humility and Openness) ....... .1 7 4 APPENDIX S: Political Orientation Assessment ................................ ............... .1 7 5 APPENDIX T: Interpersonal Reactivity Index (IRI) ................................ .......... .1 7 6 vii APPENDIX U: Self - Importance of Moral Identity Scale (SIMI) ...................... .1 7 7 APPENDIX V: Moral Re lativism Scale (MRS) ................................ ................. .1 7 8 APPENDIX W: Willingness to Learn About Marginalized Groups .................. .1 7 9 REFERENCES ................................ ................................ ................................ ................ 1 80 viii LIST OF TABLES T able 1 : Conceptual Allyship Typology ................................ ................................ ............ 11 T able 2 : Examples of Effective and Ineffective Ally Behaviors ................................ ....... 32 Table 3 : Previous Empirical Measures of Allyship ................................ ........................... 34 Table 4 : Summary of Thesis Studies ................................ ................................ ................. 45 Table 5 : Descriptive Statistics for All Categorical Measures in Study 1 .......................... 5 8 Table 6 : Descriptive Statistics for All Continuous Measures in Study 1 .......................... 58 Tabl e 7 : Bivariate Correlations Between all Measures in Study 1 ................................ .... 60 Table 8 : Qualitative Response Examples from Participants in Study 1 ............................ 61 Table 9 : Linear Regression Testing the Effect of Ally Type on All Outcomes ................ 64 Table 10 : MANCOVA Testing Effect of Ally Type on Psychological Outcomes ............ 6 5 Table 1 1 : MANCOVA Testing Effect of Ally Type on Perceptions of Ally .................... 6 7 Table 1 2 : Moderated Regression of Stigma Consciousness Predicting Outcomes ........... 6 8 Table 1 3 : Moderated Regression of Identit y Centrality Predicting Outcomes .................. 7 2 Table 1 4 : Moderated Regression of Perceived Discrimination Predicting Outcomes ...... 7 6 Table 1 5 : ABQ Top 10 Most Frequently Occurring Codes from Study 1 Data ................ 8 1 Table 1 6 : Factor Structure of Final Selected Items ................................ ........................... 9 0 Table 1 7 : Study 2 Correlations between PACS and ABQ Scores ................................ ..... 9 2 Table 1 8 : T - tests to Examine Mean Differences Between Sample Types ......................... 9 3 Table 1 9 : Study 3 Scenarios and Response Options ................................ .......................... 9 7 Table 20 : Descriptive Statistics for All Measures in Study 3 ................................ .......... 10 8 Table 2 1 : Bivariate Correlations Between all Measures in Study 3 ................................ 1 10 Table 2 2 : Summary of Int ra class Correlation Results for Scenario Behavioral Coding . 1 12 ix Table 2 3 : MR of Motivations Predicting Allyship Behaviors ................................ ......... 1 1 4 Table 2 4 : MR of Individual Differences Predicting Effective Coded Behaviors ............ 1 1 6 Table 2 5 : MR of Individual Differences Predicting Ineffective Coded Behaviors ......... 1 1 7 Table 2 6 : MR of Individual Differences Predicting Self - Report Effective Behaviors .... 1 1 8 Table 2 7 : MR of Individual Differen ces Predicting Self - Report Ineffective Behaviors . 1 20 Table 2 8 : Hierarchal Regression Predicting Effective Coded Behaviors ........................ 1 22 Table 2 9 : Hierarchal Regression Predicting Ineffective Coded Behaviors ..................... 1 2 3 Table 30 : Hierarchal Regression Predicting Self - Reporte d Effective Behaviors ............ 1 2 4 Table 3 1 : Hierarchal Regression Predicting Self - Reported Ineffective Behaviors ......... 1 2 5 Table 32 : Relative Weights Analyses of Predictors in Study 3 ................................ ....... 12 6 x LIST OF FIGURES Figure 1 : Conceptual Thesis Model ................................ ................................ ................... 3 8 Figure 2 : Effect of Ally Type on Stigma Consciousness / Psychological Safety ............... 6 9 Figure 3 : Effect of Ally Type on Stigma Consciousness / Anxiety ................................ ..... 70 Figure 4 : Effect of Ally Type on Stigma Consciousness/Positive Affect ......................... 7 1 Figure 5 : Effect of Ally Type on Identity Centrality/Psychological Safet y ...................... 7 3 Figure 6 : Effect of Ally Type on Identity Centrality/Anxiety ................................ ........... 7 4 Figure 7 : Effect of Ally Type on Identity Centrality/Positive Affec t ................................ 7 5 Figure 8 : Effect of Ally Type on Perceived Discrimination/Anxiety ................................ 7 7 Figure 9 : Effect of Ally Type on Perceived Discrimination/Positive Affect ..................... 7 8 1 INTRODUCTION Many organizations seek to hire employees from diverse backgrounds as a means of improving their overall quality. Indeed, diverse work teams tend to be more innovative, make better business decisions and ultimately allow companies to make more money ( Dickey, 2018; Larson, 2017; Hewlett, Marshall, & Sherbin, 2013). While many organizations do hire individuals from diverse and historically excluded backgrounds, some struggle to retain them. One reason for this lack of retention may involve a failure to fully include individuals from these backgrounds within the organization. At its core, an inclusive work environment is one where all employe es are valued, respected, supported, treated fairly and are included in decision making processes (Nishii, 2013; Ferdman, 2014; Winters, 2014). When there is a lack of inclusion, employees are likely to leave their positions. Individuals from marginalized backgrounds are more likely to experience more negative social interactions in the workplace and are offered fewer opportunities for career advancement (Hofhuis, Van der Zee, & Otten, 2014). Both of these factors were found to contribute to actual turnover decisions made by employees from marginalized backgrounds (Hofhuis et al., 2014) . Other research suggests that the inclusion of employees from marginalized backgrounds is important at all levels of the o rganization. Research by Davidson (2012) found that Black managers are 40% more likely to leave their positions than their White counterparts. When asked why they decided to leave, these managers mentioned that they were not considered for stretch assignme nts which are beneficial for continued advancement. In addition, they reported that they had been given inaccurate performance feedback (Davidson, 2012). 2 Although there are some legal protections for individuals from certain diverse backgrounds in the wo rkplace (e.g., the Civil Rights A ct of 1964) there are no legal protections or mandates regarding the inclusion of employees from various backgrounds. Thus, it is often up to the organization to engage in these initiatives at their discretion. While organi zations may or may not choose to implement inclusive practices, individual employees may choose to act independently and exhibit behaviors that may increase feelings of inclusion among employees from m arginalized groups. Indeed, they may serve as allies fo r their fellow employees from said groups. While individuals may wish to be allies on behalf of members of marginalized groups and some act accordingly to that end, not all allies engage in behaviors that are actually helpful for members of marginalized gr oups. For example, allies who claim to do work they are not doing, focus on their own experience as an ally rather than focusing on those they claim to help, and allies who become defensive when others tell them that their actions are not helpful may indee d be less effective allies (Smith, 2013). Even well - intentioned allies can be ineffective by attempting to implement changes or interventions without considering marginalized voices and by asking individuals from marginalized groups to teach them about opp ression rather than seeking out information on their own (Charles, 2016). I am interested in investigating factors that make ally behavior ineffective and determining whether this ineffective behavior results in negative psychological outcomes for members of marginalized groups. Th is project contributes to the literature in several ways. First, the present investigation will acknowledge that ally behaviors can be ineffective as well as effective. Most research and discussion on this topic assume that ally ship behavior will be effective and helpful for marginalized group members. However, this study will examine allyship from a more nuanced viewpoint in that such an assumption will not be made. This study will examine the behaviors 3 that are considered to be effective and the ones that are ineffective. Second, this study will characterize effective ally behavior in a novel way. To date, few studies have conceptualized the effectiveness of allies by considering the perspectives and ratings of those i n marginalized groups ( Brooks & Edwards, 2009; Brown & Ostrove, 2013). In this study , effective behavior will be determined by having minority group members rate the effectiveness of the allyship behaviors they personally experienced. Third, this effort ma y provide insights into who tends to be more effective as an ally. I will investigate the degree to which several motivational and individual difference variables relate to effective ally behavior to that end. This introduction is organized into seven mai n parts. First, the concepts of allies and allyship are defined and research on these concepts will be reviewed. In addition, theories relevant to allyship behavior are discussed. Second, I discuss the various motivations one may have for engaging in allys hip behaviors. Third, I discuss individual differences that may be related to engaging in allyship behavior. Fourth, I discuss outcomes of effective ally behavior and present allyship as a mechanism for identity safety, a construct that may well be related to inclusion. Fifth , I discuss the ramifications of ineffective ally behavior and the factors that may contribute to such ineffectiveness. Sixth, I briefly discuss other measures of allyship. Lastly, I present my hypotheses and discuss a series of studies designed to investigate the effects and characteristics of allyship behaviors. Specifically, I examine ally behaviors are effective and the behaviors that are ineffective , examine the characteristics of allies who are more likely to engage in effective an d ineffective behaviors and examine the ramifications of allyship behaviors on marginalized group members . Results of each study will be discussed in turn and a general discussion will follow. 4 Allies and Allyship Defined Previous research has who works to end oppression in his or her personal and professional life through support of, and . Other research has bro adened this definition to include members of other minority groups advocating on behalf of each other (Brown & Ostrove, 2013). One can be an ally to individuals from a wide range of backgrounds. Often these groups include racial minorities, gender minoriti es (usually women), and LGB individuals (Griffin, 1997). However, one can also be an ally for others based on their age, religion, and psychological or developmental disabilities as well (Griffin, 1997). Ideal allies would step in on behalf of those in mar ginalized groups and act on their behalf during negative incidents or incidents of non - inclusion. The act of allies stepping in may increase feelings of inclusion through identity safety. Ally development is thought to be a dynamic process that occurs ove r a period of time. Bishop (2002) outlines six stages of ally development in her model. These stages include: existence, recognizing the intersectional nature of o ppression, understanding and managing the marginalization and begin working towards changing them, actively looking for ways to advocate for change, supporting tho se in minority groups and changing the minds of others through word and deed, and looking past the difficult experiences that may come with being an ally and focusing on the positive aspects of the role. Collins and Chlup ( 2014) further refine this model a nd conceptualize ally development as a cyclical process rather than a series of sequenced steps. They argue that individuals are motivated to become allies through awareness of bias and 5 social injustices. The ally then moves continually through the cyclica l process of - ableism, heterosexism), deciphering personal commitment to advocacy, being an ally and telling others, recognizing areas of (dis)comfort with the all y role, assessing environments and facing discomfort(s), and advocating for change and seeking additional education (Collins & Chlup, 2014) . To date, there is no known empirical work done using this model. There are many potential factors that may influen act on behalf of those in marginalized groups in a given situation. The bystander intervention model ( Darley & Latané , 1970) offers one lens through which one can consider how such action may occur. In th is model, th e bystander must progress through five steps in a prescribed order for action to be taken. These steps include: (a) noticing the event, (b) interpreting the event as an emergency that requires assistance, (c) accepting responsibility for intervening in sai d event, (d) knowing how to help in the situation, and (e) implementing intervention decisions. Drawing influences from the bystander intervention model, the Confronting Prejudiced Responses (CPR) model by Ashburn - Nardo, Morris and Goodwin (2008) may also provide a more specific framework through which ally action occurs when a perceived discriminatory event happens to a person nearby. Although their model is step - wise, they assert that steps may or may not be followed in order and that some steps may be sk ipped in route to the completion of the intervening action. In the CPR model, a discriminatory event occurs and the steps occur as follows: (a) the event must be interpreted as discrimination, (b) the event must be interpreted as an emergency, (c) the obse rver must take responsibility for engaging in an action, (d) the observer must identify the correct response to take, and (e) the observer must act. This series of steps is theoretically thought to result in confronting discrimination. When considered toge ther, 6 bystander intervention theory and CPR theory serve to explain the mechanism through which allyship behavior occurs. They may also highlight areas of potential ally shortcomings. For example, an ally may be motivated to assist when a person from a mar ginalized background is experiencing negative treatment; however, if the ally does not know how to help, cannot identify the correct response or otherwise does not act, their response as an ally may be considered ineffective. Thus, considering the stages t hrough which helpful action occurs may inform thinking on how these processes may break down in the context of allyship behaviors. - cultural competence model of psychotherapy also provides a lens through which the effectiveness of ally behavior can be considered. In this theory, competence is broadly defined as being prepared to complete a task or the ability to perform said task and the ability to be culturally competent (i.e. being able to provide effective psychotherapy treatment to i ndividuals from different racial and cultural backgrounds) . Competence (a) personal characteristics, (b) skills or intervention tactics and the (c) interpersonal processes that are involved (Sue 2009). To succeed in effective treatment , the psychotherapist must be aware of their biases and how this may influence their treatment of the client and must have considered to be culturally sensitive and re levant to the client (Sue, 2009). Similarly, when engaging in allyship behaviors, allies from majority groups may differ in the degree to which they are culturally competent based on the three categories of characteristics outlined by Sue (2009). T his view aligns with allyship theories previously described (Bishop, 2002; Collins & Chlup, 2014). 7 Comparison of Allyship to Other Related Constructs There are other constructs that may be related to allyship more broadly . To compare and differentiate allyship from potentially related factors, social support and advocacy will be discussed in turn. Social support hen & McKay, 1984). Social support has long established roots in the occupational health literature. Specifically, social support has been considered to be a theoretical mechanism for dealing with occupational stressors although empirical results are mixed in terms of whether or not this construct is related to employee wellbeing (see Beehr, King & King, 1990 and Nahum - Shani , Bamberger & Bacharach, 2015 ). Although both allyship and social support appear to serve the same purpose of buffering coworkers from stressful or event threatening situations, there are also at least t wo important differences between these constructs. First, while allyship can involve providing social support, it also involves taking action on behalf of marginalized group s (Brown & Ostrove, 2013). Conversely, social support has often been operationalized as something that takes place during an interpersonal conversation between employees or between employees and supervisors or helping with general work or non - work tasks ra ther than intervening in a situation ( Beehr et al. , 1990 ; Glanz, Riner and Visawanth, 2008 ) . Second , the positive effects on the recipient of allyship may have little to do with reciprocity while there is evidence that guilt and indebtedness may affect the recipient when receiving workplace social support (Bowling, Beehr, & Swader, 2005). For workplace social support, it is arguable that this indebtedness may occur because the behaviors completed by the giver (e.g., providing a listening ear, helping clean u p a co - context of allyship , effective behaviors cannot be returned in a similar fashion . Indeed, recipients 8 from marginalized groups lack the privilege required to int ervene on behalf of majority group members and it is questionable whether expecting something in return would be something that characterizes a good ally. For these reasons, social support may be one category of allyship behavior but allyship includes othe r behaviors as well. Advocacy and allyship are also closely related constructs that appear to be quite similar upon first glance. Both advocates and allies use their position to take action on behalf of individuals from marginalized groups. However, when looking at exactly which actions allies and advocates take part in and considering the cost , risk and level of effect involved in engaging in these actions, one may be able to conceptually separate these constructs. For example, Anicha, Bilen - Green and Bu rnett (2018) explain in their program for men working with and on behalf of women in the professoriate that there were some differences between allies and advocates. Specifically, allies were expected to take action at a more localized level (e.g., within their own department) by speaking up at meetings on behalf of their colleagues, inviting women to collaborate with them on research projects and serving on committees in place of women faculty members. Comparatively, advocates were expected to be committed to learning the effect of gender bias on the academic careers of women more broadly and to educate themselves and the allies about larger issues surrounding gender inequity. Advocates were also expected to encourage the hiring and promotion of women in to higher level positions and to be sure that women are treated fairly and equitably within their own institution. Using the organization of their program as an example, allies could be conceptualized as working at the local and interpersonal level to provide support to marginalized group members while advocates could be conceptualized as working at the systemic level to support marginalized group members by making the system or environment more equitable and just. Often allies and advocates are taking 9 action at a cost or risk. For allies the cost and risk may be more marginal than for the advocate who must put in more time and sustained effort to make system - level changes. Ally Motivation Allies may act on behalf of marginalized groups due to motivational factors. A theoretical classification of helping behaviors in the workplace by Chou and Stauffer (2016) may also inform thinking regarding the motivation of allies more broadly. In their conceptualization, helping behavior is categorized into three types: (a) unsolicited proactive helping behavior, (b) unsolicited reactive helping behavior (c) solicited reactive helping behavior. The first two types may be highly relevant to understanding the motivations behind allyship behavior 1 . Unsolicited p - Stauffer, 2016) 2 . Those who engage in such behaviors are thought to do so without consideration - being and being constantly aware of other organizational members current and future tasks in order to anticipate needs (Chou & Stauffer, 2016). The authors posit that the c ause or motivation of unsolicited proactive helping behaviors is not external to the helper (e.g., the situation or rewards they may obtain for their I wo uld argue that this would also include values. Such unsolicited proactive helping may be the basis of effective allyship due to the genuine caring for well - being and constantly anticipating 1 Solicited reactive helping behavior involves voluntary helping after one is asked for help. While it is possible that on e can engage in ally - like behavior after being specifically asked for help, it is arguable that allies would be likely to help without being asked. Because of this, the antecedents of both types of unsolicited helping are thought to be more relevant to con ceptualizing allyship motivation. 2 I note that this definition may appear similar to organizational citizenship behaviors. I discuss differences between allyship behaviors and organizational citizenship behaviors later on in the manuscript. 10 the needs of others in order to take effective action (Brown & Ost rove, 2013). As such, this framework suggests that considering individual differences may be important to understanding the motivation behind this type of helping behavior. ions exhibited before being asked by coworkers that help the coworkers with work - related issues Stauffer, 2016). This type of helping behaviors aligns more with the byst ander intervention theory and CPR theory in that perception of the need of help is highly important to the determination of whether an actor may choose to act. Indeed, the authors ascribe that one of the primary motivations for unsolicited reactive helping behavior is the perception that another person is in need of help and that this help would be of value to said other (Chou & Stauffer, - benefit analysis. The authors p osit that in providing unsolicited help to a party in need, the helper is able to establish a relationship in which the person that is helped would be obligated to repay the helper at some unspecified time in the future. Indeed, they cite social exchange t heory (Blau, 1964) as an explanatory mechanism for this type of helping in that helper is using this behavior as a way to receive rewards and benefits from the one they are helping. If the desire for social exchange is the primary motivation for unsolicite d reactive helping behavior, that behavior may be related to ineffective allyship behavior. The theoretical classification outlined by Chou and Stauffer (2016) implies that there are at least two broad motivational factors that may be related to engaging in helping behaviors (see Table 1). 11 Table 1 : Conceptual Allyship Typology Ally Type Behavio r Type Motivation Type Description Ideal Ally Effective Other - Oriented An ally who engages in effective behaviors and is primarily motivated to act on behalf of marginalized individuals for altruistic reasons. Egocentric Ally Effective Self - Oriented An ally who engages in effective behaviors and is primarily motivated to act on behalf of marginalized individuals for self - related reasons. Good Intentions Ally Ineffective Other - Oriented An ally who engages in ineffective behaviors and is primarily motivated to act on behalf of marginalized individuals for altruistic reasons. Bad Ally Ineffective Self - Oriented An ally who engages in ineffective behaviors and is primarily motivated to act on behalf of marginalized individ uals for self - related reasons. First, there is other - oriented motivation. Other - oriented motivation for helping behaviors is best characterized as engaging in helping others due to a genuine interest in the wellbeing of those being helped without regard to what one could gain or lose fr om the interaction. This type unsolicited proactive helping behavior. Allies may be motivated to help those in marginalized groups because they have empathy for them and a re willing to help without regard to the cost to themselves. This motivation is also highly consistent with the empathy - altruism hypothesis. The empathy - altruism hypothesis (Batson, 1987; Batson, Lishner & Stocks, 2015) states that one may 12 be motivated to help another person due to caring about their wellbeing. Indeed, previous research has also linked empathic concern to engaging in helping behaviors (see Batson, Lishner & Stocks, 2015 for review). Based on previous theory, it is possible that allies may b e motivated to help those in marginalized groups because they are concerned about their well - being. Taken together, previous theories suggest that ally motivation can be other - oriented. Second, there is self - oriented motivation. Self - oriented motivation f or helping behaviors is best characterized as engaging in helping others due to factors related to oneself. This can include thinking of what one can gain or lose by helping and thinking about how one will be evaluated for engaging in helping behaviors. Th is type of motivation is highlighted in Chou and motivated to act on behalf of marginalized groups because they are considering what they are receiving in return and d etermining that the benefits of helping outweigh the costs. As mentioned previously, this motivation may be best explained by social exchange theory (Blau, 1964). When the costs for standing up for marginalized groups in the organizational context is low, people may be more likely to step in as allies. Similarly, if the benefits to the ally are high, they may be more likely to step in as well. However, if cost to the ally is high or if the benefits are low, they may be less likely to intervene on behalf of their marginalized coworkers. In addition, allies may be motivated to help because they view themselves as good people and they wish to verify this view with their actions. Self - verification theory presents the idea that people want others to view them in the same way they view themselves (Swann, 2012). Many people tend to view themselves and those they know as being moral or good people (Cohen, Painter, Turan, Morse & Kim, 2013). Therefore, it is possible that one reason allies may engage in allyship behav iors is 13 because they view themselves as good people and wish to be viewed in a similar way by others. Taken together, previous theories suggest that ally motivation can be self - oriented. Previous research has identified that people from marginalized group s seek allies who are supportive, committed to equality, willing to create an identity safe environment, affirming and willing to take informed action on their behalf (Brown & Ostrove, 2013; Brooks & Edwards, 2009 ). When an ally is focused more on what the y can receive as a result of the interaction, it is a possibility that they will be (1) less likely to engag e in effective behaviors, (2) engage in more ineffective behaviors and (3) may be perceived more negatively in the community of people they are atte mpting to help. As such, when exploring effective (and ineffective) ally behaviors it is important to consider the motivations behind why an ally chooses to engage in helping people from marginalized groups. This study examine s several variables that coinc ide with other - and self - oriented motivations. Specifically, I examine other - oriented motivations such as prosocial motivation and wanting to appear non - prejudiced for internal reasons and examine the self - oriented motivation of wanting to appear non - preju diced for external reasons. I review each of these motivational factors in turn in the section to follow . Prosocial Motivation . Engaging in helping with the goal of increasing the welfare of others (Grant, 2008) may be a relevant motivation for allyship be havior. Allies who engage in allyship behaviors because they genuinely care about the people being helped may thus be more likely to focus on the needs of those in marginalized groups. Furthermore, they may be less focused on what they stand to gain or los e when engaging in helping behaviors. Because of this, allies who engage in helping behavior due to prosocial motivation may be more likely to engage in effective allyship behavior and less likely to engage in ineffective behavior . 14 Internal and externa l reasons for wanting to appear non - prejudiced. In general, it may be socially desirable to appear non - prejudiced in modern society. However, the reasons why people want to appear this way differ in the general population and may well differ among the popu lation of individuals who identify as allies (Plant & Devine, 1998). Some allies are motivated to appear non - prejudiced for internal reasons. They may wish to appear in this way because it is intrinsically important to them and may align with their values. They may also wish to appear this way because they genuinely care about members of marginalized groups and do not want to negatively impact them. Taken together, it is likely that individuals who wish to appear non - prejudiced for internal reasons may also engage in effective ally behaviors. There are also external reasons to appear non - prejudiced. For example, some people may wish to appear non - prejudiced due to societal pressures or because they wish to be viewed positively b y others . Allies who are motiv ated to appear non - prejudiced for external reasons may engage in ally behaviors in order to be viewed positively by others or for some other form of social capital. Because their focus is on how other perceive them, less of their attention is on helping th ose in marginalized groups. Thus, allies who are motivated to appear non - prejudiced for external reasons may be more likely to engage in ineffective behaviors. In sum, motivations may be an important determinant of allyship behavior (see Table 1 for propos ed theoretical ally typology). When an ally is motivated for other - oriented reasons, it is possible that they will be more focused on the person or people they are trying to help than what they stand to gain or lose for their actions. This may ultimately c ontribute to them being more likely to engage in effective ally behaviors overall and less likely to engage in ineffective behaviors (e.g., the ideal ally). Conversely, when an ally is motivated for self - oriented reasons, it is likely that they will be mor e focused on themselves and what they can gain or lose from 15 engaging in allyship behavior. This may unfortunately contribute to them being less likely to engage in effective ally behaviors ally ship behaviors overall and more likely to engage in ineffective ally behaviors (e.g., the bad ally). However, it is also possible that allies can be motivated for self - oriented reasons and still be effective (e.g., the egocentric ally). Similarly, an ally could also be motivated for other oriented reasons and actually engage in ineffective behaviors (e.g., the good intentions ally). However, it is likely that the allies who would be most effective in terms of increasing positive outcomes for members of marginalized groups would be the ideal allies while those who would be least effective would be the bad allies. Egocentric and good intentions allies may have moderate effects in terms increasing positive outcomes for marginalized groups, but they are less likely to be as effective as the ideal ally. While these ally type s will not be directly tested during the course of these proposed studies, the typology highlights the importance of considering both behaviors and motivations when examining ally effectiveness. Ally Individual Differences Allies who are members of domin ant groups are thought to have a wide variety of qualities. In addition to being low on prejudice and being willing to understand their own position of privilege (Curtin, Kende & Kende, 2016), allies are also thought to affiliate with marginalized group me mbers and stand against their societal oppression (Wijeyesinghe, Griffin, & Love, 1997 , Broido, 2000 ) . Qualitative research by Brown and Ostrove (2013) found that when asking people of color to describe their white allies, allies were described as being affirming and taking informed action. Affirmation includes qualities such as caring about, being respec tful of, and communicating liking for people of color. Taking informed action involves being willing to be active and take a stand on issues of race or race bias in favor of the minority 16 groups. When asking LGBTQ+ employees what they sought from allies, th ey mentioned wanting to be accepted, supported, included, and treated fairly. In addition, they wanted to receive the same benefits and rights as all other employees and wanted to know that their identity would not result in losing their job or hinder them from being successful in their job role (Brooks & Edwards, 2009) . Taken together, these findings suggest that individuals from marginalized groups seek allies who will care about them and support them as people but that will also stand up for their rights as equal members of the organization. This suggests that marginalized group members may look to allies as a mechanism through which identity safety can be achieved. Indeed, there may be individual differences that are related to engaging in effective al lyship behaviors. These individual differences may include personality, empathy, perspective to endorse moral relativism), moral identity, lower degree of endorsement for social dominance orientation, being less selfish, being willing to learn about marginalized groups and having close relationships with people from marginalized groups. I will briefly discuss each in turn. Personality. General personality is a very widely considered antecedent to behaviors. Its prevalence is predicated on the idea that certain people engage in certain behaviors while other people do not as a result of semi - fixed characteristics. It is reasonable to explore the possibility t effective allyship behavior. Specifically, looking to the six - factor HEXACO model (Ashton & Lee, 2009), the traits of honesty - humility and openness to new experience may be particularly related to engaging in allyship behavior. The honesty - humility dimension is comprised of sincerity, fairness, greed avoidance and modesty. Allies high o n this trait may be less likely to engage in allyship behavior for self - oriented re asons and may be more effective as an ally 17 because they tend not to focus on the m sel ves . The openness to new experience dimension is comprised of aesthetic appreciation, inquisitiveness, creativity and unconventionality. Allies high o n this trait may be mo re likely to seek out information about people from marginalized groups and may also be more likely to challenge the status - quo through action. This may allow them to be more likely to engage in effective ally behavior and less likely to engage in ineffect ive ally behavior . Empathy . Feeling warmth, sympathy and having concern for others may also allow allies to engage in more effective behaviors (Davis, 1980). Previous research has indicated that there is a strong link between empathy and helping behavior more broadly ( Dovidio , Piliavin, Schroeder & Penner, 2006). Empathic concern may allow allies to be more attuned to the experiences of marginalized people and to the systematic disadvantages they face. In addition, empathic concern may allow allies to car e deeply for those they help and thus may motivate them to help. Both of these factors may explain why those higher in empathic concern may be more likely to engage in effective allyship behaviors and less likely to engage in ineffective behavior. Perspective Taking. may be highly relevant to engaging in effective allyship behaviors (Davis, 1980). Perspective taking may allow allies to place themselves in the position of people from m arginalized groups. Considering which actions may be best from the perspective of those in marginalized groups may then allow allies to engage in more effective behavior. Conversely, individuals who are lower on perspective taking may be more likely to eng age in ineffective behavior. Political Orientation engages in effective allyship behavior s for members of marginalized groups. Those who hold liberal beliefs tend to support policies that grant rights to people from marginalized groups (e.g., 18 civil rights and marriage equality). In addition, they tend to be more likely to believe that things are unequal between groups in America and that changes must be made. In a recent Pew Research Center (2017) survey, liberals were more likely to believe that the country still needed to make changes to give Black Americans equal rights and that women continue to face significant obstacles that make it more difficult for them to succeed compared to men. L iberals also report ed having positive views of immigrants and were more accepting of LGBTQ+ individuals (Pew Research Center, 2017). Because liberals tend to support policies and legislation that allow for rights to be extended to individuals from marginal ized groups, hold positive attitudes towards these groups, and recognize that these groups still face inequality, it is likely that liberals are more likely to hold the belief that these groups deserve said rights. Thus, those who are more liberal may be m ore likely to be effective allies to members of marginalized groups. Conversely, those who are more conservative may be more likely to engage in ineffective ally behavior. Moral Relativism. Viewing morality as being relative to the perspective of the ind ividual or culture one is from may be related to being ineffective as an ally. Previous research on moral relativism has found that the construct is inversely related to self - reported moral character (Collier, 2017 ; Collier - Spruel, Hawkins, Jayawickreme, F urr & Fleeson, 2019 ). People who believe that there is no objectively true morality may be less likely to be effective allies for people from marginalized groups. Part of allyship is holding the belief that systematic injustices are wrong and must be overt urned through acting on behalf of the oppressed ( Collins & Chlup , 2014) . This would necessitate believing that this oppression is objectively wrong which is directly at odds with the belief that there is no objective morality. Thus, allies who are lower in 19 moral relativism may be more likely to engage in more effective allyship behavior and less likely to engage in ineffective behavior . Moral Identity. influence their effectiveness as an ally. Those who view morality as being important to their sense of self are thought to be more willing to engage in moral actions and have more concern f or members of their outgroups (Aquino & Reed, 2002). Moral identity is often characterized as having two dimensions: internalization and symbolization. Internalization refers to the degree to which a person can quickly, efficiently and consistently access their internalized schema of morally relevant knowledge such as moral traits, moral goals and moral behaviors (Aquino, Freeman, Reed, Felps & Lim, 2009). Individuals with higher internalized moral identity are thought to be able to access this morally rele vant knowledge much faster than those who are lower in this trait. Because allies who are higher in internalized moral identity may be able to access the network of morally relevant knowledge more quickly, they may be able to act more quickly and effective ly on behalf of marginalized group members. Thus, allies higher in internalized moral identity may be more likely to engage in effective allyship behaviors. While internalization focuses on the private element of moral identity, symbolization refers to the degree to which an individual expresses their moral identity externally (Aquino et al., 2009; Aquino & Reed, 2002). In reference to the measure by Aquino and Reed (2002) , symbolization occurs through the action said person takes. Indeed, being a member of certain organizations, and wearing certain clothing would be ways an individual could identify themselves as having moral characteristics. While allies high in symbolized moral identity may be more visible, this visibility may not necessarily be linked to being effective. Although those higher in symbolized moral identity may externally project the appearance of an ally, they may actually be more 20 focused on the cues they send to others than their effectiveness towards members of marginalized groups. This m ay ultimately result in allies who are higher in symbolized moral identity engaging in more ineffective ally behaviors. Thus, allies who engage in a greater degree of effective behaviors may be more likely to have higher levels of internalized moral identi ty and lower levels of symbolized moral identity with a projected inverse relationship between these constructs and ineffective behaviors . Social Dominance Orientation . The degree to which one prefers hierarchy within society and desires that low - st effectiveness as an ally (Pratto, Sidanius, Stallworth & Malle, 1994). Those who are higher in social dominance orientation not only prefer inequality but may actively work to keep these inequ alities in place. Because of this, higher levels of social dominance orientation are more likely to be in direct conflict with effective allyship behaviors. Instead higher levels of this characteristic may be linked to engaging in ineffective ally behavior s. Selfishness . Individual differences in the degree to which one focuses on their own well - (2018) outline three types of selfishness. Egotistical selfishness is t he degree to which an individual is single - mindedly focused on the self without consideration towards others. Adaptive selfishness is the degree to which one is selfish to survive and help others that are close to them and can help them. Pathological selfi shness is the degree to which an individual is selfish to the point of sacrificing others to achieve their own ends. Allies who engage in ineffective behaviors may tend to be higher on the egotistical and adaptive selfishness dimensions. This is primarily due to the idea that these individuals may be motivated to be allies because they see others as a mean to an end. Those higher in egotistical selfishness may see helping marginalized groups as a 21 source of supply for their own self - esteem. Those higher in a daptive selfishness may see helping marginalized groups as an opportunity to eventually ask for something in return later. While allies who engage in ineffective behaviors may also be higher in pathological selfishness, it is arguable that people who are p athologically selfish are: (a) rare in the general population, (b) less likely to be allies at all. Those who are higher in egotistical and adaptive selfishness however, may be more prominent and more likely to identify as allies. Unfortunately, being high er on these dimensions may result in a greater likelihood of engaging in ineffective behaviors and a decreased likelihood of engaging in effective behaviors . Willingness to learn about marginalized groups. Empirical research and popular press sources have mentioned the importance of obtaining knowledge about the experiences of individuals from marginalized groups (Jones, Brewster & Jones; Charles, 2016; Sebastian, 2016). Being willing to seek this information may allow allies to better understand how individuals from marginalized groups experience systematic injustices. It may also allow allies to be more accepting and less dismissive when individuals from marginalized backgrounds share their experiences. Because of this, it is probable that being will ing to learn about marginalized groups will ultimately allow allies to be more likely to engage in effective behavior and less likely to engage in ineffective behavior . Close relationships with people from marginalized groups. Positive intergroup contact has long been thought to help decrease bias between groups (Pettigrew & Tropp, 2005). Friendship and other close relationships with members of marginalized groups may be characterized as being among the highest levels of intergroup contact. When allies ha ve friends from marginalized groups, they are more likely to learn about marginalized experiences. In addition, they may be more empathetic to the cause of the marginalized as they see those close to 22 them deal with larger systematic issues. Lastly, allies who have close relationships with marginalized group members may be more likely to notice systematic injustices because they are affecting those closest to them. All of these potential factors may allow allies who have close relationships with marginalized group members to be more likely to engage in effective ally behavior and less likely to engage in ineffective behavior . In sum, honesty - humility, openness to new experiences, empathy, perspective taking, internalized moral identity, willingness to learn about marginalized group and close relationships with marginalized groups are all thought to be related to engaging in more effective ally behavior, while moral relativism, being conservative, symbolized moral identity, social dominance orientation, and s elfishness are all thought to be related to ineffective allyship behaviors. Some may argue that research on allyship may benefit from considering organizational citizenship behavior (OCBs) frameworks. While it is reasonable to liken allyship to OCBs mo re broadly, there are at least two reasons why alternative frameworks may better conceptualize allyship in the organizational context. First, allyship behavior is generally directed towards a marginalized individual or marginalized group without much consi deration to work or organization - related reasons while OCBs can be related to individuals or groups for specifically work - related or organization - related reasons (Ehrhart & Naumann, 2004). While it is possible that allies would step in to confront structur al biases for work - related reasons, it is highly likely that many instances of effective allyship behaviors would occur independently, and perhaps, in spite of work tasks and work roles. Second, allyship behaviors tend to result in individual level outcome s for members of marginalized groups while OCBs can and often do result in individual and organizational level benefits for the organization (Podsakoff & MacKenzie, 1997). In a 23 review of organizational citizenship behaviors and organizational effectiveness , Podsakoff and MacKenzie (1997) present several organizational benefits to OCBs, including: enhancement of productivity, freeing up resources for effective use, enhanced stability of organizational lity to adapt to changes in the environment. While it is possible that allyship behaviors may allow for individuals from marginalized groups to have increased feelings of belonging and psychological safety in the workplace, allowing said individuals to per form better, it is unlikely that allyship will result in or fashion. However, allyship is likely to increase attraction to and retention of individuals from ma rginalized backgrounds into organizations. Thus, while allyship is not projected to benefit the organization in the way that OCBs tend to, it may result in other positive outcomes such as increased diversity , increased inclusion and increased retention fro m these groups . Factors Affecting Marginalized Perceptions of Ally Behaviors own individual differences may result in different perceptions of the eff ectiveness of ally behavior. In previous research, marginalized group members were asked to evaluate specific dimensions of allyship, namely affirmation and informed action (Brown & Ostrove, 2013). Examining interpersonal evaluations more broadly, Hartley et al (2016) used a set of items designed to assess general evaluative interpersonal dimensions. These dimensions included: liking, respecting, wanting to spend time with, admiring, competence, sociability and morality. It is possible that the way marginal ized group members view their allies may also be related to the In the present study, looking at various evaluative dimensions outlined in Hartley et al. (2016) separately could reveal a dis tinct 24 pattern of results for allies who engage in effective behaviors compared to allies who engage in ineffective behaviors. For example, perhaps individuals from marginalized groups may differ in the degree to which they respect, like and admire allies w ho engage in effective versus ineffective behavior but do not evaluate the two differently in terms of sociability. It is also possible that individuals from marginalized groups have a more global evaluation of the ally in which they view the allies who en gage in effective behaviors as being higher on all positive traits and lower on all negative ones compared to allies who engage in effective behaviors. Separate from interpersonal evaluations , the individual differences of marginalized group members may also be differentially related to perceptions of ally behavior. It is important to consider the ways in which individuals differ within their identity group as a means of determining these perceptions and their overall eff ects 3 . In the present study , there may be at least three psychological individual difference variables that may result in differential relationships between the projected predictors and outcomes. I will discuss stigma consciousness, identity centrality and perceived discrimination in turn as potential exploratory m oderators that may affect the degree to which effective and ineffective allyship behaviors affect outcomes for members of marginalized groups . Although I discuss potential directionality of relati onships in this section, examination of the psychological underlying mechanisms will be exploratory in the present study. Stigma Consciousness . I ndividuals from marginalized groups may differ from one another in the degree to which they expect to be stere otyped by others (Pinel, 1999). Stigma consciousness has been linked to being more likely to attribute negative comments to 3 Drawn from w ork in preparation by Dr. Chris Nye and Dr. Fred Leong 25 discrimination (Pinel, 2002). Other research linked stigma consciousness to turnover intentions and eventual turnover behavior with feeling disrespected as the mediator between these variables (Pinel & Paulin, 2005). Taken together, it is possible that individuals from marginalized groups who have a heightened degree of stigma consciousness will be particularly sensitive to ineffective ally behaviors and may in turn be more negatively impacted by them. Because of this, individuals who are higher in stigma consciousness may be more negatively impacted when engage in ineffective behavior compared to individuals who are lower in stigma con sciousness. Identity Centrality . Individuals from marginalized groups may differ in the degree to which their demographic characteristics are central to their sense of self (Brittan et al., 2013). Ethnic affirmation has been thought to be related to ad aptive functioning for members of racial and ethnic minorities including decreased drug and alcohol use, higher self - esteem and better academic achievement (Brittan et al., 2013). Because of this, it is possible that individuals who are higher in identity centrality may be more likely to be resilient when faced with discriminatory or biased behavior. When allies intervene on their behalf, it is possible that those higher in identity centrality may report better psychological outcomes and report negligible e ffects on psychological outcomes when allies engag e in ineffective behaviors due to the resilience buffer. Conversely, individuals lower in identity centrality may rate allies engaging in effective behaviors in a similar manner to those higher in identity centrality but allies engaging in ineffective behaviors more harshly than their counterparts. Perceived Discrimination . Individuals from marginalized groups may differ in the degree to which they experience discrimination in their daily life. Perceived discrimination has been long linked to negative health outcomes, decreased wellbeing and psychological distress (see Pascoe & Richman, 2009 for a metanalytic review) . Because of th e negative effects of perceived 26 wellbeing , allies engaging in effective behaviors may be particularly well received when an individual is experiencing discrimination frequently. Thus, it is possible that individuals who are higher in perceived discrimination may be particularly positivel y affected by allies who engage in effective ally behaviors on their behalf and may evaluate them more positively than individuals who experience lower degrees of perceived discrimination. In addition to potential predictors of differential perce ptions o f allyship behavior, it is also important to consider potential covariates that may remove error variance that is not related to the hypothesized relationships. Specifically , trust , stands out as a potential covariate . The degree to which individuals trust others in general ( Yamagishi & Yamagishi, 1994 ) may influence the degree to which they evaluate allies broadly and in terms of the degree to which they believe that the ally is engaging in helping behaviors for altruistic reasons. Because trust may influe nce the relationship between predictors and outcomes, it is important to consider it as a covariate. Outcomes of Allyship Although outcomes of allyship behavior will not be directly measured in the present investigation, it is important to discuss them due to the fact that they are often considered the ultimate goal of these behaviors. I will discuss identity safety , persuasion of similar others , and organizational structure change as potential outcomes of effective allyship. While allyship has been tho ught to be helpful in a diversity related context more broadly, it may also yield positive ramifications for the identity safety of marginalized group members in organizations. Many say that allyship efforts are thought to decrease prejudice from other 27 maj ority group members. Previous research indicates that when majority group individuals advocated on behalf of marginalized individuals, others in the majority group are more receptive to the message (Gardner, 2018). Among other benefits, allyship is also th ough to increase identity safety and decrease stereotype threat. Social identity threat may influence how well individuals from stigmatized groups will perform in performance related settings. It involves attention to situational cues in the environment that suggest that an individual may not be valued due to being a member of a marginalized or stigmatized group (Murphy and Taylor, 2012, Purdie - Vaughns et al., 2008). Situational cues are thought to influence expectations about what can be achieved in the environment. Early research on situational cues suggest that individuals use these cues to determine the degree to which the environment is safe (Goffman, 1971). Indeed, individuals have been known to make attributions regarding discriminatory behavior by attending to prejudice - related cues put forth by their surroundings (Major, Quinton, & Schmader, 2003). When cues suggest that individuals from minority groups will be marginalized or stigmatized (e.g., numeric underrepresentation, lack of inclusion statem ent, stereotypical objects or visual depictions) this results in perceiving the environment to contain identity threat, whether or not there is actual prejudice present (Murphy & Walton, 2013). The results of experiencing threatening cues are overwhelmingl y negative and include feeling like one does not belong (Walton & Cohen, 2007) and experiencing a decrease in executive functioning (Johns, Inzlicht & Schmader 2008). In workplace settings, individuals who experience a higher degree of identity threat also have higher levels of turnover and absenteeism, are considered to be poorer performers and are less likely to be considered for promotion (Avery, McKay & Wilson, 2007; Ilgen & Youtz, 1986; James, 2000; Landau, 1995). 28 In addition to the broader concept of identity threat, stereotype threat has been long theorized to negatively impact the performance of individuals from marginalized groups (Steele & Aronson, 1995). It is considered to occur when individuals from marginalized or stigmatized groups must attend to the possibility of being evaluated negatively due to stereotypes made about their group (Steele & Aronson, 1995; Walton, Murphy & Ryan, 2015). There are four conditions that are thought to contribute to stereotype threat: (1) individuals from stereotyp ed groups must be aware of the consistent, negative stereotype, (2) the task being completed is thought to assess an ability that is negatively stereotyped, (3) the task is difficult, and (4) the individual wants to perform well and also identifies with th e stigmatized group (Steele, 2010; Walton et al., 2015). When individuals from marginalized groups enter performance environments, such as the workplace, they may be motivated not to conform to the expectations of the negative stereotype. To avoid conformi ng to the stereotype, these individuals must engage in impression management which is thought to be cognitively taxing. Having to impression manage may then influence performance. Indeed, when individuals from marginalized groups experience stereotype thre at in these performance settings, they tend to perform worse than their actual ability would suggest (Inzlicht & Schmader 2012; Walton et al., 2015). When individual performance suffers, organizational success as a whole may be mitigated. While there have been several attempts to find ways to reduce stereotype threat and identity threat, one of the most notable ways is through increasing identity safety. Identity safe environments project the idea that being a member of a marginalized group will not impede advancement and success in performance settings (Davies, Spencer & Steele, 2005; Steele & Cohn - Vargas, 2013). There are several strategies that are thought to increase identity safety includ ing increasing the number of individuals from underrepresented gro ups, promoting 29 individuals from these groups, the use of value affirmation interventions, the use of social - belongingness interventions and broader diversity training (see Walton et al. 2015 for review). In identity safe environments, individuals from marg inalized groups may be more likely to perform to the best of their ability as they will not need to expend energy attending to threatening environmental cues (Markus, Steele & Steele, 2000). It is highly likely that effective allyship behavior can serve t o decrease feelings of identity threat and increase feelings of identity safety. Indeed, standing up against the oppression of marginalized group members in the workplace may be effective in three ways. First, effective allyship behaviors may mitigate thre atening cues. When the environment or other majority group out against these harmful actions and biases. Engaging in this immediate and interpersonal individua l level form of allyship this may increase the likelihood that marginalized group members feel as though their identity is safe in their workplace environment and that they will have a chance to thrive and advance. Effective allies can also increase iden tity safety more distally through helping change the organizational environment through persuading others and challenging pre - existing organizational structures to the extent that this is possible. With regard to persuading similar others, effective allies are thought to be more persuasive when standing up for minority group members themselves (Gardner, 2018). One of the reasons this is thought to occur is due to the idea that these allies have nothing to gain personally when speaking up for marginalized gr oups and promoting initiatives. Conversely, when individuals from marginalized group s attempt to persuade others in the majority group to consider diversity initiatives, marginalized group members are seen as requesting something for their benefit. More re cent research has found that 30 when black employees advocate for themselves in workplace contexts, they are seen as being more self - serving than white counterparts who advocate on their behalf (Gardner, 2018). Effective allyship behaviors from majority group members may be one way to navigate these biases from majority group coworkers. If allies are effective at helping majority group members confront their biases and change their behaviors, this may ultimately create an identity safe work environment for min ority group members. Lastly, just as effective allies can advocate at the individual and singular event level, they may also be able to affect changes at higher levels in organizations. Ideal allies would minimize the extent to which they benefit from str uctural biases by noticing them, pointing them out to those in authority and refusing to participate in these systems to the extent that they can. Examples would include speaking up about wage disparities, suggesting qualified people from marginalized grou ps when they hear about stretch projects, and speaking highly of people from marginalized groups to those in management or those who give employee evaluations. As previously mentioned, when discussing allyship and advocacy, changing systems may take more t ime and result in higher risks or costs to the person engaging in the behaviors. Thus, these behaviors may overlap with the behaviors of advocates. While some may argue that it is important to consider less subjective ways in which allies can engage in effective behaviors (i.e., persuading similar others, creating actual structural change), as opposed to more subjective metrics via self - report, the present study will focus exclusively on the interpersonal allyship beha viors that may increase identity safety for employees from marginalized groups. I have decided to do this because persuading similar others and creating structural change are outcomes that individual employees have relatively little control over. In the ca 31 behaviors to result in increased identity safety for their fellow employees. In the case of the latter, the organization as an entire system must be amenable to structural change for the behaviors to result in increased identity safety. Such structural change efforts may require the coordinated effort of many employees and those in management ranks or higher over a period of several years before incremental changes can occur. Compar atively, when individual employees from majority groups witness discriminatory or biased actions taken against their coworkers from marginalized groups, there may be actions they could take in the moment or soon after that may result in increased identity safety. Of the three outcomes mentioned, interpersonal actions at to focus on these for the present study. The Ramifications of Ineffective Allyship In g eneral, previous research focuses on the positive effects of allyship. This focus ultimately assumes that all allyship behaviors are effective and helpful for the person receiving ally aid. However, when considering the positive effects of effective allysh ip, it is also important to consider what can occur when allies engage in ineffective behaviors. While no empirical research has yet considered the adverse effects of ineffective or harmful allyship attempts, many popular press citations have made specu lations about the impact of these behaviors. Factors that are speculated to decrease the likelihood that positive outcomes are achieved include behaviors that either do not directly help people in marginalized groups or behaviors that place the burden of t he work on marginalized individuals. Examples of these ineffective behaviors include being patronizing, being self - serving, asking marginalized group members to teach them about their experience rather than seeking out information themselves, engaging in t one policing (e.g., controlling how minority thoughts are expressed) or threatening 32 to withdraw assistance if members in marginalized groups do not behave in a way that the ally deems to be appropriate (Ariel, 2017; Charles, 2016; Muniz, 2016; Zivad; 2017 M orrison , 201 3 ; Patton , 201 7 ; Threads of Solidarity, 201 7; see Table 2 for examples). Ta b le 2 : Examples of Effective and Ineffective Ally Behaviors Behavior Citation Effective Behaviors Taking informed action to address biases and oppression Brown and Ostrove (2013) Being affirming towards marginalized group members Brown and Ostrove (2013) Jones, Brewster and Jones (2014) Amplifying minority voices Sebastian (2017) Accepts constructive criticism Sebastian (2017) Reaching out to marginalized groups and asking what they need Sebastian (2017) Ineffective Behaviors Being patronizing or condescending Ariel (2017) Failing to listen to members of marginalized group members Charles (2016) Failing to speak up Charles (2016) Expecting marginalized group members to teach them about oppression rather than researching it for themselves Charles (2016) Responding negatively when being held accountable by marginalized group members Charles (2016) Giving advice to marginalized groups from a place of privilege Charles (2016) Offering conditional or contingent support Muniz (2016) Being an ally in name only Smith (2013) 33 Table 2. Focusing on their own experiences as an ally rather than on experiences of marginalized groups Smith (2013) Engaging in Tone Policing Ziyad (2017) Despite the fact that some people become allies with the best of intentions, these intentions may not necessarily result in effective behaviors on behalf of the group they are trying to help. Ultimately, behaviors that are not helpful may not only fail to mitigate identity - thr eat related cues, but these behaviors may also increase identity threat. It is also highly possible that when helpful behaviors occur unhelpful or harmful behaviors may occur during the same instance. These unhelpful or harmful behaviors may release threat ening or ambiguous cues that decrease the identity safety of those from marginalized groups. Thus, these threatening (unhelpful or ambiguous) cues may result in marginalized group members feeling a lack of inclusion in the organizational context. Other M easures of Allyship At this time, there are two known measures of allyship towards members of marginalized groups (see Table 3). 34 Table 3 : Previous Empirical Measures of Allyship Measure Citation Purpose Content Coverage Ally Identification Measure Jones, Brewster & Jones (2007) Designed to assess skills to support LBGTQ+ persons. Also assesses degree to which allies seek out information and have knowledge of LGBTQ+ persons Ally Cognitions Awareness of Bias Knowledge About Res ources Knowledge of Oppression Willingness to Learn About Ally Behaviors Seeking Information About LGBT persons Political Action Perception of Ally Scale Brown & Ostrove (2013) Designed to assess the degree to which allies engage in informed action and affirmation (emotional support) for members of minority racial groups Ally Behaviors Affirmation Informed Action Note: Content coverage does not refer to factor structure. The first one, known as the Ally Identification Measure (AIM) was crea ted by Jones, Brewster and Jones (2007) to assess the degree to which allies had skills and knowledge to support LGBTQ+ individuals. Item content included three factors: having knowledge about resources for LGBTQ+ individuals and being willing to support t hem (e.g., knowledge and skills factor), openness to learning more about LGBTQ+ individuals and being willing to take action on their behalf (e.g., openness and support factor), and awareness of oppression towards LGBTQ+ individuals (e.g., oppression aware ness factor). While this scale was validated, it is 35 arguable that some of the item content does not align with the overall factor. For example, there is one item in the knowledge and skills factor that would be better served in the openness and support fac have developed the skills necessary to provide support if a sexual minority person needs my help nversations with sexual conversations with LGBTQ+ allows one to be an ally as it is arguable that non - allies also engage in this behavior. The second measure, kn own as the Perception of Ally Characteristics (PACS) was created by Brown and Ostrove (2013) to assess the degree to which allies took informed action on behalf of members of racial and ethnic minority groups and the degree to which they provided affirmati on and emotional support to members of said groups. One of the strengths of this measure is that it was created with an eye towards what racial and ethnic minorities were seeking in an ally. Considering this perspective is quite rare in the larger allyship literature. In addition, this measure also demonstrates a distinction between taking action or advocating for the rights of the marginalized and providing emotional support for them. These measures present the idea that there may be more than one way to be an effective ally. This is particularly true when considering the wide variety of elements that are included in both previous measures. While these measures do include behavioral items, they also include having knowledge and skills, being aware of oppre as an ally. In addition, these measures make no distinctions between effective and ineffective behaviors. The present investigation aims to create a behavioral measure that conceptualizes effective allyship behav ior and ineffective allyship behavior. In addition, I plan to consider 36 several motivational factors and individual differences that may lead to those sets of behaviors. Focusing specifically on behaviors may allow for a more parsimonious measure. The Present Investigation The purpose of the present study is to examine the motivations, qualities and characteristics that contribute to effective and ineffective ally behaviors . While many effective allyship behaviors have already been identified in pre vious research (Brown & Ostrove, 2013), little attention has been paid to the role of ineffective allyship behavior and its ramifications. Thus, for the present study, I have elected to focus on the motivations and individual differences that may be relate d to ineffective allyship as well. This study characterize s allyship effectiveness in a novel way. Effectiveness was determined by having marginalized group members rate the effectiveness of the allyship actions they witnessed. Actions identified by margi nalized group members were then presented to majority group members who were asked to self - report the degree to which they complete said actions. In the larger literature, the few empirical studies that exist consider allyship through the viewpoint of the majority group member ally (Broido, 2000; Reason, Millar & Scales, 2005). Research has failed to take marginalized experiences into account with Brown and Ostrove (2013) and Brooks and Edwards (2009) being two of the primary exceptions. One of the main cr iticisms of allies in the popular press is acting without considering the wishes and voices of those in the marginalized group they are trying to help (Charles, 2016). Without operationalizing allyship effectiveness from the lens of the marginalized perspe ctive, research may be culpable of the same criticism. This study will result in the creation of a new quantitative measure in the larger ally literature. At this time, there appear to be only two other quantitative measures pertaining to 37 allyship: The Perception of Ally Characteristics Scale (PACS; Brown & Ostrove, 2013) and the Ally Identity Measure (AIM; Jones, Brewster & Jones, 2007). While these measures largely assess ally cognitions and actions, these measures do not consider the degree to which t hese actions are effective and fail to consider actions that are blatantly ineffective. Simply gaining knowledge about resources or being emotionally supportive may not translate into actions that increase inclusion and identity safety for marginalized gro up members. In addition, these measures were created for use in only one demographic domain (e.g., race, sexual orientation; see Table 3) despite the fact that members of many marginalized groups often seek similar outcomes (e.g., support and identity safe ty). The measure created in the present study is designed to be more general and is intended to be used to assess a variety of behaviors (both effective and ineffective) taken by allies on behalf of members of all marginalized groups. In addition, creation of a new measure may spark interest in the topic of allyship more broadly and allow for more empirical research to occur in this burgeoning area. Lastly, this study will consider several new qualities that may be related to effective and ineffective ally ship behaviors. Some will be drawn from more general personality research (e.g., HEXACO) and many of them will be drawn from moral psychological research (e.g., moral relativism, moral identity, empathy and perspective taking). This is due to the fact that the actions that comprise allyship behavior would also be characterized as moral actions more broadly (Hartley et al., 2016). 38 Study Rationale and Development of Hypotheses In order to achieve the goals outlined previously, several studies were conduct ed to develop an assessment of ally behavior and to see whether these behaviors influence feelings of identity safety for members of marginalized groups. The first two studies focus on the creation of the ally behaviors measure. The first study also examin e s whether ally behaviors (effective and ineffective) affect the degree to which individuals from marginalized groups feel identity safety. Lastly, the third study allow s me to examine the motivations and individual differences related to being an effectiv e ally and those that are related to being an ineffective ally. For a conceptual model of these hypotheses, please see Figure 1. Figure 1. Conceptual Thesis Model An ally may engage in allyship behaviors on behalf of members of marginalized groups for self - and other - oriented reasons. While these motivations may ultimately result in ally 39 behavior, the degree to which these behaviors are effective may be dependent on whether one is motivated to help for self - versus other - oriented reasons. Previous resea rch on general helping behaviors found that intrinsic motivation for helping mediated the link between empathy and helping behaviors while this relationship did not occur for extrinsic motivation (Pavey, Greitemeyer & Sparks, 2011). Similarly, other resear ch suggests that when one engages in helping behaviors for internal reasons (as opposed to external reasons), this resulted in a greater degree of wellbeing for not only the recipient of the help, but also for the helper (Weinstein & Ryan, 2010). Taken tog ether, this research suggests that when one helps for self - oriented reasons, specifically external self - oriented reasons, they may be less effective at actual helping as it relates to the perspective of the person receiving the help. A review on motivation s for helping behaviors suggests that there is a great deal of empirical support for the idea that a person is motivated to help altruistically due to feeling empathy for the person in need (Batson, Lange, Ahmad, & Lishner, 2003). While it is possible that allies could engage in effective behaviors and be motivated by self - oriented reasons (e.g., the egocentric ally) or that they could engage in ineffective behavior and be motivated by other - oriented reasons (the good intentions ally), allies in these categ ories will most likely be rarer and ultimately, they will be less effective than the ally who engages in effective behavior for other oriented reasons (see Table 1 for ally typology). Based on these previous findings, it is hypothesized that: Hypothesis 1a : Individuals who are higher in other oriented - motivations will be more likely to engage in effective ally behavior than those lower in other - oriented motivations. Specifically, individuals higher in ( a) prosocial motivation and ( b) motivation to appear no n - prejudiced for 40 internal reasons will be more likely to engage in effective ally behavior than those lower in these motivations. Hypothesis 1b: Individuals who are higher in self - oriented motivations will be more likely to engage in ineffective allyship b ehavior than those lower in these motivations. Specifically, individuals higher in motivation to appear non - prejudiced for external reasons will be more likely to engage in ineffective allyship behavior than those lower in this motivation. In addition to motivations, there may be individual differences that are related to engaging in allyship behavior more broadly. Empathy and perspective taking have been linked to helping behavior in much of the previous research (Batson, Lange, Ahmad, & Lishner, 2003; D avis & Maitner, 2010). It is indeed possible that these individual differences may be related to engaging in effective allyship behaviors as well. Feeling empathy for members of marginalized groups may allow one to be more emotionally supportive. Empathy m ay also allow these allies to feel motivated to act on behalf of marginalized group members. In addition, being willing and able to consider the perspectives of marginalized group members may lead to more effective helping behaviors. Furthermore, when peop le perceive that one is engaging in taking their own perspective, they tend to like the helper more and receive more help from them (Goldstein, perspective taking, the more likely they will engage in effective behavior. The degree to which one is liberal and the degree to one does not endorse the domination of lower status groups by higher status groups may influence their effectiveness as allies of members of marginal ized groups. Overall, liberal political candidates in recent years have tended 41 to support legislation that allows marginalized groups increased rights and freedoms in society. Therefore, people who support these candidates and their policies may support th e idea that marginalized groups do deserve said rights. Although I do not argue that liberalness causes one to engage in more effective behaviors, I do believe these variables may be correlated. Similarly related, having a lower degree of social dominance orientation may also be related to engaging in effective ally behaviors . Individuals low in social dominance orientation do not prefer arbitrary hierarchies based on demographic group status and instead prefer egalitarianism among groups (Pratto et al., 19 94). Individuals low in social dominance orientation may indeed view all people as equal and deserving of rights. Taken together, more effective allies are likely to be more liberal and less likely to endorse social dominance orientation. The degree to which one thinks about morality more broadly and in relation to the self rules as to what is right or wrong may run counter to the beliefs and actions of effec tive allies (Collier - Spruel et al. , 201 9 ). If one does not believe that systematic injustice is wrong, there is no reason to act and attempt to overturn it. Because of this, effective allies are likely to exhibit lower levels of moral relativism. The degre internalized moral identity did find that the construct was predictive of most types of prosocial behaviors and p rosocial action including : volunteering, social involvement, helping when others are experiencing an emergency, helping when others are experiencing negative emotions, helping privately without anyone knowing and helping without expectation or reward (Patr ick, Bodine, Gibbs, & Basinger, 2018). Because moral identity has been linked to a variety of prosocial behaviors more broadly, it is likely that moral identity, specifically internalized moral 42 identity will allow allies to engage in more effective allyshi p behaviors. Conversely, the positive relationship between symbolized moral identity and prosocial behaviors depends on whether or not the behavior results in positive recognition for the actor (Patrick et al., 2008). Thus, those higher in symbolized moral identity may be more interested in the rewards they reap as a result of their behavior than the degree to which they help others. Taken together, effective allies are projected to be less likely to be moral relativists and have higher levels of internaliz ed moral identity and lower levels of symbolized moral identity. Willingness to learn about marginalized groups and the degree to which one has close relationships with individuals from marginalized groups may also affect the degree to which one engages i n effective ally behavior . Knowledge about individuals from marginalized groups has been considered to be an important factor for allies (Jones, Brewster & Jones, 2007; Charles, 2016; Sebastian, 2016). Being willing to seek out information about marginaliz ed groups may help them learn more about the people being helped and may potentially allow allies to provide those being helped with better resources. In addition, close relationships with marginalized group members may prompt allies to take effective acti on to help those they are closest to. Taken together, allies who engage in effective behaviors may be more willing to learn about members of marginalized groups and have more close relationships with members of marginalized groups. ay also influence the degree to which an ally is effective. Specifically, the degree to which one is higher in honesty - humility and openness to new experiences (Ashton & Lee, 2009) may allow them to be more effective allies. Allies higher in the honesty - hu mility domain may be more modest and less focused on themselves and what they can gain from engaging in allyship behaviors. This may then allow them to be more focused on the needs of marginalized group members and thus be better allies to them. In additio n, being 43 open to new experiences may allow allies to be more interested in seeking out information about people in marginalized groups and more interested in learning about their experiences. These factors may allow allies who are high in openness to bette r understand those they are trying to help and therefore engage in more effective allyship behaviors. Thus, allies who are higher in the honesty - humility and openness to new experiences domains may be more likely to engage in effective allyship behaviors. It is therefore hypothesized that: Hypothesis 2: Effective allies will be (a) more empathetic, (b) be more likely to engage in perspective taking, (c) be more liberal, ( d ) have lower social dominance orientation ( e ) exhibit a lower degree of moral relativism (f) have a greater degree of internalized mora l identity (g) have a lower degree of symbolized moral identity (h) be more willing to learn about marginalized groups (i) have more close relationships with people in marginalized groups (j) will have a higher degree of honesty/humility and openness to ne w experiences compared to ineffective allies. Lastly, allies may differentially influence the degree to which people in marginalized groups experience identity safety. Allies who engage in effective behaviors may be viewed positively by those they are tryi ng to help. Specifically, t hey may be viewed as being as competent and capable but also interpersonally likable . Indeed, an ally who engages in effective allyship behavior may make individuals from marginalized groups feel a greater degree of psychological safety. Allies who engage in effective behaviors may be more successful at thwarting cues that are threatening to the identities of marginalized group members. Allies who engage in ineffective behaviors may be less likely to increase feelings of identity safety as they may be less likely to actively thwart threatening cues. Therefore, it is hypothesized that: 44 Hypothesis 3 a : Allies who engage in effective behaviors are more likely be perceived as ( a) increasing psychological safety , ( b) decreasing anxiet y ( c) increasing positive affect, ( d) decreasing negative affect, and e) increasing voice perceptions for marginalized group members compared to allies who engage in ineffective behaviors. Hypothesis 3b: Allies who engage in effective behaviors are more li kely to be evaluated positively by the individuals they help. Specifically, allies who engage in effective behaviors are ( a) more likely to be perceived as engaging in ally behaviors for other - oriented motivations, ( b) less likely to be engaging in ally be haviors for self - oriented motivations, more likely to be reported as engaging in ( c) informed action and ( d) affirmation ( e) more likely to be evaluated positively on all interpersonal evaluative dimensions and (f) more likely to have an improved relations hip with the individuals they help compared to allies who engage in ineffective behaviors. To test the proposed model, I examine perceptions of effective and ineffective all y behaviors and the effect of these perceptions on psychological outcomes for ma rginalized group members and subsequent interpersonal ally perceptions . Study 1 test s hypothes e s 3a and 3b in the model. Studies 1 and 2 also focus on the creation of a measure of general ally behaviors. Lastly, I examine whether individual differences are related to ally behaviors to test the other side of the proposed model. Study 3 directly test s H ypotheses 1a , 1 b and 2 in the proposed model. For a summary of all studies, please see Table 4. 45 Table 4 : Summary of Thesis Studies Study Number Study Goal Sample Type Sample Size 1 Primary goal is to determine whether effective and ineffective ally behaviors affect the degree to which marginalized group members experience identity safety. Secondary goal is to collect item content that will help with the development of a general measure of ally behaviors. Working adults who are racial/ethnic minorit ies or LGBTQ+ 2 11 2 Goal is to create and assess items reflecting typical ally behaviors using themes from Study 1 and literature from Table 1. Working adults who are racial/ethnic minorities or LGBTQ+ 2 25 3 Goal is to determine the relationship motivations and individual differences have with allyship behavior for majority group members. Working adults who are majority group members 300 46 STUDY 1 METHOD Study Purpose The purpose of the first study was to collect critical incidents of experiences with allyship from members of marginalized groups and to establish how the effectiveness of ally behavior relates to psychological safety , anxiety, positive affect, negative affect and voice for marginalized group members and their evaluations of their allies . Items for the new measure , which were tested in Study 2, were based on these critica l incidents and scenarios for study 3 were also drawn from these incidents . This study tested H ypothes e s 3 a and 3b . Participants F or the first study, participants were recruited from Amazon Mechanical Turk (Mturk). All participants were Americans who were racial and ethnic minorities or Americans who were sexual or gender minorities (e.g., LGBTQ+). Purposive sampling was used to be s ure that participants from these two specific marginalized groups were sampled. All participants were at least 18 years of age, were American citizens who had been living in the U.S. for the past five years and were fluent in English. Participants also rep orted working more than 30 hours per week and that they had experienced allyship behaviors from a member of their respective majority group in the past 12 months in order to qualify. Participants who met all of these criteria were selected to participate i n this study and received $1.00 for their participation in this study. The 2 11 p articipants in this study were all members of marginalized racial, ethnic, sexual and gender identity groups. With consideration to race and ethnicity, 32.2% of participants indicated that they were White - American, 27.1% indicated that they were Black - American, 15.6% indicated that they were Asian - American, 14.6% indicated that they were Hispanic - or 47 Latino - American 8% indicated that they belong to two or more rac ial/ethnic identity groups and 2.5% indicated that they were American - Indian. With consideration to sexual orientation, 44.7 % of participants indicated that they were heterosexual, 24.4% identified as homosexual, 23.4% identified as bisexual, 5% identifie d as pansexual and 2.5% identified as asexual. With consideration to gender identity, 49.5% of participants identified as male, 43% of participants identified as female, 4% of participants identified as non - binary and 2.5% of participants identified as tra nsgender with 1% of participants not indicating their gender identity. On average, participants were 32 years old ( SD = 8.65) , made 50,000 - 59,000 dollars annually (min = > $10,000 ; max = $150,000 <) erms of educational attainment (62.5%). When asked to describe their relationship to their ally , 2 6 . 6 % indicated that their ally was a close friend, 7 % indicated that their ally was a friend, 8 % indicated that their ally was their romantic interest or par tner, 9. 5 % indicated that their ally was an acquaintance, 0.5% of participants mentioned that their ally was a family member, 11. 6 % indicate d that their ally was a coworker, 2 1 . 6 % indicated that their ally was a stranger and 15 . 1 % indicated that their ally was someone else (e.g., colleague, fellow patient, mentor, professor, classmate, professional connection, supervisor). Post - hoc evaluation of the count data revealed that close friends and strangers were both found to be effective (30 close friends, 26 st rangers) and ineffective (23 close friends, 17 strangers). A chi - square test of independence determined that there were no significant differences between close friends and strangers in being an effective or ineffective al ly , X 2 (2, N = 96 ) = .15 , p = . 70. 48 Measures To test H ypothes es 3a and 3b , several categories of measures were implemented. First, I will discuss measures related to the evaluation of the ally - related experience. Then, I will discuss the covariate measure that was considered. Then, I will discuss the outcome measures implemented . Lastly, I will discuss the individual difference measures implemented . Ally Experience Related Measures Open ended responses regarding previous allyship experiences . All participants were purposely assigned ( within marginalized group) to describe either a positive (when the ally was effective) or negative (when the ally was ineffective) previous experience with allyship depending on their answer to the allyship experience screening question. Participants who i ndicated that they had a positive experience with allyship behavior over the past 12 months were assigned to describe a positive allyship experience. Participants who indicated that they had a negative experience or that indicated that they had both a posi tive and negative experience were assigned to describe a negative allyship experience. They were first asked to share 2 - 3 sentences summarizing the general experience. Then, they were asked to list specific behaviors that the ally was engaging in. Lastly, they were asked questions about their perceptions of the motivation of the ally described in their experience (see Appendix A for questions). Perceived Self and Other Motivation Questions. for engaging in allyship behavior, I asked them to respond to six self - wante - 49 averaged together within each motivation type. Higher scores indicated that participants attributed a greater degree of self - oriented motivation and other - oriented motivation to their The self - oriented motivation items did not demonstrate sufficient reliability ( =.63) while the other - oriented motiva tion items did demonstrate sufficient reliability ( =.7 3 ). Although these subscales were found to be moderately correlated in the expected direction ( r = - . 29 ) these subscales are conceptually distinct. Perceived Ally Characteristics Scale (PACS). To assess the degree to which allies have the qualities outlined by Brown and Ostrove (2013), participants were asked to report on the degree to which their ally was affirming and engaged in informed action (see Appendix B). N ot at A ll C harac V ery C haracteristic) scale participants were asked to respond to 6 informed action items and 4 affirmation items. Examples of informed action 4 was 5 knowledgeable about racial/ethnic (sexual minority) communities ot items were averaged together within each of their respective subscales. Higher scores on each of the subscales are indicative of engaging in a greater degree of affirming behavior and a greater degree of informed action respectively. Brown and Ostrove (2013) subjected this measure to an exploratory factor analysis and a confirmatory factor analysis and the two - factor structure exhibited good model fit (CFI =.94 and SRMR = .08). I also conducted a confirmatory factor 4 5 Items were changed to past tense as participants would be recounting a previous event rather than an ongoing one in the proposed study 50 analysis on this me asure using data collected for the present study and found that the two - factor structure exhibited good model fit (CFI =.95, TLI =.94, RMSEA = .08, SRMR = .05). The affirmation sub - scale exhibited sufficient reliability ( = . 90 ) as did the informed action subscale ( = . 77 ). These subscales were highly correlated with one another ( r = .6 4 ) ; however , there was not complete overlap between these subscales. Thus, each subscale may explain additional variance that the other subscale does not. Because of this a nd because the authors demonstrated model fit for the two - factor measure, retain ed the two factors for use in the present study despite the high correlation. Relationship with Ally Questions. with th eir ally and how their relationship may have changed following the incident described by the participant , I created two questions. For the first question, I asked participants to indicate the relationship they had with their ally at the time of the incident. Participants could choose from the following options: Close Friend, Friend, Romantic Partner/Spouse, Acquaintance, Family Member, Co worker and Stranger. If the relationship was not listed among the potential options, participants were free to specify the relationship using a fill - in - the - blank option. For the second question, I asked participants to indicate whether their relationship w ith the ally changed following the incident they described. Participants could choose from the following options: (coded as 3 ) , the relationship remained the same , and 1 ) . Thus, higher scores on this measure indicate that the improved . Evaluation Questions. N ot at A l V ery M uch) scale, participants rated the degree to which they like, respect, would enjoy spending time with and 51 admire the ally they described in their critical incident. They also rated the degree to which they found the ally to be moral (fair, honest, compassionate, not selfish), competent (capable, independent, determined, not incompetent) and sociable (extroverted, talkative, not reserved, not boring). I created additional items in which participants were asked to rate the degree to which the y found the ir ally to be effective (helpful, supportive, not harmful, not hurtful). These items N ot at A V ery M uch) scale. This measure demonstrated sufficient reliability ( = . 90). While general evaluation items were co nsidered as separate entities, items for morality, sociability, competence were averaged within each subscale to compute total scores. Higher scores indicate a greater degree of that evaluative element of the ally. These questions were created by the autho rs of Hartley et al. (2016) and there is no evidence that these questions were psychometrically tested. However, the authors treated each of these indices as distinct. These subscales exhibited sufficient reliability ( morality = . 93; competence = . 95; sociability = . 78) . Looking at the bivariate correlations between these dimensions, morality, competence, liking , respecting , and the desire to spend time with an ally were all highly related ( r = .7 2 - .8 5 ). Correlations between these dimensions and sociability were much lower ( r = .3 2 - .4 6 ). Because of this, it is arguable that there may be a superordinate category of positive evaluation from which sociability is a separate judgement . Upon conducting follow up confirmatory factor ana lys e s (CFA s ) using the lavaan package in R, it was revealed that a factor structure in which all positive evaluations were kept as separate factors did exhibit decent fit (CLI = .92, TLI = .89, RMSEA = .13, SRMR = .05), In addition it was revealed that a f actor structure in which all positive evaluations were joined together as one factor exhibited relatively decent fit ( CLI = .86, TLI = .83, RMSEA = .17, SRMR = .05) but not as good fit as when all evaluations were kept as separate factors. Because of the h igh correlations between the 52 positive evaluations, I decided to combine them and leave sociability separate for the analyses despite the CFA results . Covariate Trust. In order to assess the degree to which participants feel trust in general, I implemented the general trust scale (Yamagishi & Yamagishi, 1994; see Appendix C). Using a Responses to each item were averaged together to indicate the degree to which participants exhibit trust in general. Higher scores are indicative of a greater degree of trust. This measure was considered for use as a covariate to control for the degree to which participants trust others in general. This measure exhibited sufficient reliability ( =. 88). Outcomes Psychological Safety. In order to assess the degree to which participants feel psychologically safe, an edited version the three - item measure created by May, Gilson and Harter (2004) was used. Items were edited to reflect psychological safety around the ally in the , participants indicated the degree to which they felt psychologically Responses to each item were averaged together to indicate the degree to which participants feel psychologically safe. Higher scores are indicative of a greater degree of psychological safety. This measure exhibited sufficient reliability ( =. 7 2 ). 53 Posi tive and Negative Affect. The Positive and Negative Affect Schedule (PANAS ; Watson, Clark & Tellengen, 1988 ; see Appendix D ) was implemented as an index of the N scale, participants were asked to respond to 20 items indicated the degree to which they felt each of the affective responses as a result of the behavior of the ally they described during the study. Example item content includ Responses to each item were averaged together within each subscale to indicate the degree to scores are indicative of a greater degree of positive and negative affect respectively. These subscales exhibited sufficient reliability ( positive = . 92; negative = . 94). Although these subscales were found to be moderately correlated in the expected direction ( r = - .26) these subscale s are conceptually distinct. Anxiety. In order to assess the degree to which participants felt state anxiety as a result of their interaction with an ally, the state items from the short form of the State Trait Anxiety Inventory (STAI) w as implemented (Spielberger, Gorsuch, Lushene, Vagg, & Jacobs, 1983 ; Marteau & Bekker, 1992 responded to three reflecting state anxiety. I tem conten t include s , and Responses to each item were averaged together to indicate the degree to which the participant experienced anxiety during their interaction with their ally. Higher scores are indicative of a greater degree of anxiety. This measure exhibited sufficient reliability ( =. 86). Voice Questions. In order to assess the degree to which participants felt comfortable several voice questions were created (see Appendix E ) 54 Agree) scale, participants indicated the degree to which they felt comfortable providing feedback Responses to each item were averaged together to create a single score representing the e behaviors with their ally. These items were created by me and were not subjected to psychometric evaluation. This measure exhibited sufficient reliability ( =. 9 2 ). Individual Difference Measures Stigma Consciousness. To assess the degree to which partic ipants anticipate being negatively stereotyped by majority group members, the Stigma Consciousness Questionnaire (SCQ; Pinel 1999 ; s ee Appendix F ) was implemented. Participants responded to ten items with White (Heterosexual or Cisgen dered) people, I feel like they interpret all my behaviors in terms of the fact that I am a racial minority (sexual minority group member or gender minority representing the degree to which participants were stigma conscious. Higher scores indicated a greater degree of stigma consciousness. This measure exhibited sufficient convergent and discriminant validity in past research (Pinel, 1999) . This measure exhibited suffici ent reliability in the present study ( =. 8 2 ). Identity Centrality. To measure the degree to which participants found their identity group to be central to their identity, the Multi - group Ethnic Identity Measure (MEIM ; Phinney, 55 1992 ; see Appendix G ) was implemented. Participants responded to twelve items reflecting identi what my ethnicity (sexual orientation or gender identity) means to m were averaged together to create a single index of identity centrality with higher scores This measure exhibited sufficient relia bility ( =. 88). Perceived Discrimination. To assess the degree to which participants perceive that they experience discrimination in their daily life, the Everyday Discrimination Scale (Williams, Yu, Jackson and Anderson, 1997 ; see Appendix H ) was impleme nted. Participants responded to nine create a single index of perceived discrimination with higher scores indicating a higher degree of perceived discrimination. T he scale was determined to have excellent psychometric fit (CFI = 0.935; RMSEA = 0.080; Krieger, Smith, Naishadham, Hartman & Barbeau, 2005; Taylor, Kamarck & Shiffman 2004). This measure exhibited sufficient reliability ( =. 92). Demographic Questionnaire. Lastly, a demographic questionnaire was administered to all par gender, and sexual orientation (see Appendix I ). 56 Procedure Participants completed a series of prescreening questions in order to determine whether they qualifi ed for the study. Qualifying participants were then forwarded on to the informed consent (see Appendix J ). Consenting participants were asked to provide critical incidents in which they experienced either effective (helpful) or ineffective (not helpful) ally behaviors. Participants were assigned to the critical incident condition based on their responses during the prescreen. Specific ally, participants who indicated that they had a positive experience with allyship behavior over the past 12 months were assigned to describe a positive allyship experience while participants who indicated that they had a negative experience or had both a positive and negative experience were assigned to describe a negative allyship experience. After providing information about the allyship experience, participants were then asked to qualitatively describe the behaviors and motivations of the ally in their critical incident. They were then asked to respond to close - Participants were then asked to evaluate the ally generally and about their relationship with the ally. They were then aske d the degree to which they would feel psychologically safe around that they could voice their concerns to their ally regarding their behavior and the deg ree to which they felt the ally they described exhibited affirmation and informed action. Participants were also asked several individual difference questions designed to assess stigma consciousness, identity centrality, perceived discrimination and trust. Lastly, participants were asked to complete demographic questions before receiving their payment code and being debriefed (see Appendix K ). 57 STUDY 1 RESULTS Data Cleaning A total of 2132 attempted entries ( 1123 in the racial and ethnic minority version and 1009 in the sexual minority version into the study) occurred. In order to ensure the quality of the data, some participants were excluded from the analyses. Participants who did not meet the study qualifications or f ailed the pre - screening questions ( 78 8 participants in the racial and ethnic minority sample and 620 in the sexual and gender minority sample ) or provided unusable qualitative responses (190 participants in the racial and ethnic minority sample and 246 in the sexual and gender minority sample ) were excluded from analyses . These filters and the qualification check at the end of the survey which was used to verify that a participant was indeed a member of the marginalized group resulted in 97 8 participation a ttempts being screened out from the racial and ethnic minority sample and 866 participation attempts being screened out from the sexual and gender minority sample before data cleaning. Once these participants were properly excluded, additional participants were screened out for other reasons. This included a participant who described an ally that was not a majority group member (One participant from the racial and ethnic minority sample). In addition, participants who were found to have completed multiple s urveys were compensated for both submissions but only their first entry was included for analyses ( 17 participants) . Following these exclusions, a final sample of 200 participants remained (95 racial and ethnic minority participants and 105 sexual and gender minority participants ; see Table 5 ). 58 Table 5 : Descriptive Statistics for All Categorical Measures in Study 1 Racial/Ethnic Sample Sexuality/Gender Sample Total Sample Effective Ally N = 47 N = 57 N =104 Ineffective Ally N = 48 N = 48 N = 96 Total N =95 N = 105 N = 200 Descriptive Information For descriptive information for all scale items, see Tabl e 6 and for bivariate correlations between variables see Table 7 . Overall , it appears that means exhibited a tendency towards the middle on most measures and standard deviations were close to 1.00. With regard to the descriptive statistics there were no known exc eptional data abnormalities. The planned covariate, Trust, did not correlate highly with any of the outcomes and thus I decided not to enter it into the model as a covariate during this study. Table 6 : Descriptive Statistics for All Continuous Measures in Study 1 Measure Mean Standard Deviation Minimum Maximum Reliability Perceived Self - Oriented Motivation 2.93 1.03 1.00 5.00 .63 Perceived Other - Oriented Motivation 3.95 .93 1.00 5.00 .7 3 Subsequent Relationship with Ally 2.29 .68 1.00 3.00 N/A 59 Psychological Safety 3.87 1.02 1.00 5.00 .7 2 Anxiety 2.34 1.02 1.00 4.00 .86 Positive Affect 2.96 1.08 1.00 5.00 .92 Negative Affect 2.00 .99 1.00 4.80 .94 Voice 3.44 .97 1.00 5.00 .9 2 Informed Action 3.45 .86 1.00 5.00 . 77 Affirmation 3.80 1.11 1.00 5.00 . 90 Liking Evaluation 5.44 1.59 1.00 7.00 N/A Respecting Evaluation 5.40 1.80 1.00 7.00 N/A Spend Time with Evaluation 5.13 1.82 1.00 7.00 N/A Moral Evaluation 5.31 1.61 1.00 7.00 .93 Competence Evaluation 5.44 1.50 1.00 7.00 .95 Sociability Evaluation 5.03 1.33 1.00 7.00 .78 Helpful Evaluation 5.34 1.58 1.00 7.00 .90 Global Positive Evaluation 5.34 1.49 1.00 7.00 .96 Trust 3.42 .85 1.00 5.00 .88 Stigma Consciousness 3.26 .69 1.30 5.00 .8 2 Identity Centrality 3.76 .67 2.08 5.00 .88 Perceived Discrimination 2.93 1.07 1.00 6.00 .92 60 Table 7 : Bivariate Correlations Between all Measures in Study 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 1. Perceived Self - Motivation - 2. Perceived Other Motivation - .29** - 3. Psychological Safety - .43** .54** - 4. Anxiety .39** - .43** - .55** - 5. Positive Affect - .13 .27** .30** - .64 ** - 6. Negative Affect .34** - .33** - .58** .62 ** - .26 ** - 7. Voice .08 .20** .21** - .09 .24 ** - .09 - 8. Informed Action - .16* .43** .45** - .58 ** .55 ** - .31 ** .24 ** - 9. Affirmation - .33** .64** .62** - .59 ** .46 ** - .41 ** .24 ** .64 ** - 10. Liking - .39** .64** .58** - .54** .43** - .32** .23** .59** .76** - 11. Respecting - .40** .71** .63** - .54** .38** - .40** .22** .61** .75** .81** - 12. Spend Time With - .37** .64** .53** - .55** .44** - .36** .28** .61** .75** .75** .80** - 13. Moral Evaluation - .40** .73** .67** - .64** .45** - .46** .23** .61** .80** .81** .85** .81** - 14. Competence Evaluation - .35** .66** .61** - .53** .41** - .41** .24** .55** .72** .72** .82** .74** .84** - 15. Sociability Evaluation - .14* .28** .35** - .27** .31** - .21** .23** .24** .29** .37** .39** .32** .40** .46** - 16. Helpfulness Evaluation - .36** .63** .66** - .59** .34** - .55** .18* .57** .75** .69** .79** .71** .82** .79** .27** - 17. Trust .14* .16* .06 - .18* .28** - .03 .28** .26** .16* .15* .21** .22** .21** .16* .06 .16* - 18. Stigma Consciousness - .06 .04 .01 .19** - .24** - .01 - .09 - .10 - .05 - .11 - .01 - .06 - .01 .02 .03 .00 - .25** - 19. Identity Centrality .06 .37** .27** - .22** .30** - .24** .29** .33** .29** .26** .32** .25** .32** .35** .33** .32** .26** .18** - 20. Perceived Discrimination .22** - .14 - .25** .23** - .07 .41** - .02 - .04 - .23** - .16* - .22** - .16* - .17* - .21** - .12 - .29** - .06 .19** - .04 - 61 Qualitative Data Analyses Qualitative responses were solicited to help generate items for the second study and scenarios for the third study. Although there was no coding conducted on the types of incidents during Study 1, Table 8 provides a few examples of the qualitative descript ions provided by the participants in this study. Table 8 : Qualitative Response Examples from Participants in Study 1 Racial/Ethnic Sample Sexuality/Gender Sample Effective Ally Example 1 When I got off the ferry from Seattle, a group of angry white guys tried heckling me. An older gentleman, who was Caucasian , stepped in and started talking to me like I knew him. We pretended to know each other and talk about a project we were working on a t work and walked me the rest of the way to my next bus. After the guys left, he gave me a hug and told me "no one should be treated that way. You're a nice young gal - you don't deserve that." I thanked the man kindly for helping me and we went on our way . When I got engaged, I had a few naysayers on FB start cracking jokes about how I couldn't possibly be getting married because I'm bi so I can't stay with one person. I defended myself but of course they paid no mind. However, when one of my heterosexual , cis male friends stepped in in my defense they listened to him. I'm still annoyed at that attitude but this friend has helped me out a number of times. Effective Ally Example 2 I was in a discord chat with my girlfriend (who is white) and some of her f riends. We were playing a squad game of Fortnite. One of our squad mates, her friend, was killed. He was upset that I did not come to his aid in time and went on a racial epithet laced tirade. My girlfriend immediately came to my defense. Her response was so rapid, I barely even had time to process what was going on. When a person at a restaurant was saying how i was disgusting for talking about dating a woman. My son actually stepped up and said it was rude of them and that there was nothing disgusting or wrong about it and that the guy was just narrow minded and stupid. I was shocked but also proud of my son in that instance. 62 Table 8. Effective Ally Example 3 I was in a work meeting between the general manager, the product owner, and me. The product owner made a conscious effort to make sure I was recognized for the work I had accomplished. This actually had nothing to do with the meeting, but it was a gesture she provided. I didn't ask for it and it was very shocking. I was with my wife and a friend at a chain restaurant having dinner. We were minding our own business and only talking amongst ourselves. One guy at the table next to us was eaves dropping and the n started to make derogatory remarks towards us, calling us disgusting, immoral and other disgusting and offensive terms to refer to homosexuals. The waitress overheard some of this behavior and at first, simply asked the man to stop. He continued and she eventually gave him a speech about how disgusting his behavior was and had the manager ask him to leave. Ineffective Ally Example 1 I experienced a situation in a grocery store parking lot when a white person called me a racial slur over a parking space. The lady waited until I got out of the car and then started following me and my child. She used profanity and seemed to feel very ent itled. An employee (young white man) in the parking lot told her it was unnecessary. However, the lady also told the employee that white people need to stick together and he smiled and nodded. I was with my girlfriend's family and they were making anti - lg bt comments. I was upset but uncomfortable speaking up because they scared me, and I wasn't out to them yet. My girlfriend thought it would be helpful for them to know I'm non binary so I could explain why their comments are hurtful, and she outed me witho ut my permission. Ineffective Ally Example 2 It happened while in class. The discussion was about racial injustice and he was trying to say how people are responsible for their actions and how the system was not on the side of people of color but ended up arguing that today we have justice and that the responsibility is on everyone. Failing to see how the system is still corrupt. I was at the bar I usually go to, a straight sports bar. I frequent it because it's the closest to my house, and there aren't a lot of nightlife options. I was being called a fag by a drunk patron. A buddy of mine got into a verbal fight, which increased to physical fight, and escalated the situation even more and brought the attention of the entire bar to us. 63 Table 8. Ineffective Ally Example 3 I was at a bar and there was a guy who was talking about how he detested Japanese people because of WW2. There was a girl at the bar who tried to intervene on my behalf, but she only made things worse. She told him that she was 1/8 Korean and that Asian people are completely different from everyone else and that no one else understands what we go through. It only escalated from there since I did not know her, nor did I consider her Asian and the guy just got more belligerent. It was awful. We were at a friend's house party and every one was discussing sexual identity. I did not want to bring up my orientation. Some of the people at the party understood that I was bi, I never discussed the fact that I am pan with them because I felt as if they really would not be able to wrap their heads around the fact that I will date a person of any gender identity. The one straight friend who I did bring with me felt the need to out me to the group because she finds my se xual orientation really interesting and wanted everyone to be able to understand what a pan - sexual is. Quantitative Data Analyses The goal of this study was to test H ypothes es 3a and 3b which specified: E ffective ally behaviors are more likely to be related to higher levels of ( a) psychological safety , ( b) higher levels of voice, ( c) a higher perception of affirmation and informed action from the ally, ( d) a greater degree of positive affect and a ( e) lower degree of negative affect for marginalized group members compared to allies who engage in ineffective behaviors and allies who engage in effective behaviors are more likely to be evaluated positively by the individuals they help. Specifically, allies who engage in e ffective behaviors are ( a) more likely to be perceived as engaging in ally behaviors for other - oriented motivations, ( b) less likely to be engaging in ally behaviors for self - oriented motivations, more likely to be reported as engaging in ( c) informed acti on and ( d) affirmation ( e) more likely to be evaluated positively on all interpersonal 64 evaluative dimensions and (f) more likely to have an improved relationship with the individuals they help compared to allies who engage in ineffective behaviors. Before testing the hypotheses, I conducted preliminary linear regression analyses to determine the relationship between ally type (e.g., effective and ineffective) and the psychological and evaluative outcomes (see Table 9) . These analyses revealed that ally typ e significantly predicted outcomes in the expected directions with the exception of voice. Thus, being in the effective ally condition predicted significantly higher psychological safety, positive affect, perceived other - oriented motivation, improved subse quent relationships with allies, higher informed action, higher affirmation, higher global positive evaluation and higher sociability compared to being in the ineffective ally condition. Being in the effective ally condition significantly predicted lower a nxiety, lower negative affect, lower perceived self - orientation. Table 9 : Linear Regression Testing the Effect of Ally Type on All Outcomes Dependent Variable B Standard Error t sig Psychological Safety .91 .45 .13 7.07 <.01 Anxiety - 1.46 - .72 . 10 - 14.53 <.01 Positive Affect 1.04 .48 .13 7.71 <.01 Negative Affect - .91 - .46 .12 - 7.37 <.01 Voice .11 .06 .14 .78 .43 Perceived Self - Oriented Motivation - .75 - .36 .14 - 5.50 <.01 Perceived Other - Oriented Motivation .67 .36 .12 5.41 <.01 Subsequent Relationship .70 .52 .08 8.49 <.01 Informed Action .77 .45 .11 7.07 <.01 Affirmation 1.19 .54 .13 8.99 <.01 Global Positive Evaluation 1.60 .54 .18 9.00 <.01 Sociability Evaluation .75 .28 .18 4.16 <.01 Note: For ally type ineffective Allies were coded as 0 and effective allies were coded as 1. 65 In order to test Hypothesis 3a , a multivariate analysis of covariance (MANCOVA) was conducted to determine whether there were differences in the effect of the two ally types (e.g., effective and ineffective) on the dependent variables (e.g., psychological safety, anxiety, positive and negative affect, voice). T he type of marginalized group (African American or LGBTQ) was used as a covariate in this model. Because trust d id not correlate highly with any of the outcomes in the present study, it was not used as a covariate. The overall MAN C OVA was significant signifying that there was a significant effect of 48 ; F ( 6 ,1 9 2 ) = 34 . 89 , p < .01). When direction (see Table 10 ). On all dependent variables with the exception of voice, participants in the effective ally group rated their allies more positively than those in the ineffective ally group. While this relationship is not causal, these results indicate that there is a positive relationship between effective allies and well - being related outcomes for marginalized group memb ers. The covariate was non - significant ( p = .54) , suggesting that ally effectiveness did not differ as a result of the marginalized group type in the present study. Table 10 : MANCOVA Testing Effect of Ally Type on Psychological Outcomes Effective Ally Ineffective Ally F sig Dependent Variable Mean Standard Deviation Mean Standard Deviation Psychological Safety 4.31 .84 3.40 .98 49.92 <.01 Anxiety 1.64 .64 3.09 .77 208.32 <.01 Positive Affect 3.46 .85 2.42 1.05 58.48 <.01 Negative Affect 1.56 .80 2.47 .95 54.30 <.01 Voice 3.49 .94 3.38 .99 .63 .43 66 In order to test H ypothesis 3 b , a multivariate analysis of covariance (MANCOVA) was conducted to determine whether there were differences in the effect of the two ally types (e.g., effective and ineffective) on the evaluation - related variables (e.g., perceptions of ally motivations, re ported subsequent relationship with ally, perceptions of affirmation, informed action from the PACS, global positive evaluation and sociability evaluation ). T he type of marginalized group (African American or LGBTQ) was used as a covariate in this model. B ecause trust did not correlate highly with any of the outcomes in the present study, it was not used as a covariate. The overall MAN C OVA was significant signifying that there was a significant effect of the type of ally on ies F (16,182) = 16.19, p all results matched the hypothesized direction (see Table 1 1 ). P articipants in the effective ally group were more likely to report that their alli - oriented motivations, more likely to report that their relationship with their ally subsequently improved following the event , reported that their allies engaged in a greater degree of informed action and affirmation, evaluate d their allies more positively (globally and in terms of sociability) and were behavior was due to self - motivated motivations compared to those in the ineffective ally group. While this relationship is not causal, t hese results indicate that there is a positive relationship between effective allies and evaluations and perceptions of said allies for marginalized group members. The covariate was non - significant ( p = .54), suggesting that the evaluation of allies did not differ as a result of the marginalized group type in the present study. 67 Table 1 1 : MANCOVA Testing Effect of Ally Type on Perceptions of Ally Effective Ally Ineffective Ally F sig Dependent Variable Mean Standard Deviation Mean Standard Deviation Perceived Self - Oriented Motivation 2.57 .94 3.32 .98 30.22 <.01 Perceived Other - Oriented Motivation 4.28 .69 3.61 1.03 29.31 <.01 Subsequent Relationship 2.63 .51 1.93 .65 70.72 <.01 Informed Action 3.82 .74 3.04 .81 49.94 <.01 Affirmation 4.37 .72 3.18 1.12 80.96 <.01 Global Positive Evaluation 6.11 .98 4.51 1.51 80.26 <.01 Sociability Evaluation 5.39 1.25 4.64 1.30 17.33 <.01 In order to test whether certain individual differences (e.g., stigma consciousness, identity centrality and perceived discrimination) are related to differential reporting of experiencing each of th e outcome variables (e.g., psychological safety, anxiety, positive affect, negative affect and voice), a series of moderated regression analys es were conducted. To view the results of the final models, see Table s 1 2 - 1 4 . The first series of m oderated regression models tested the interactive effect of stigma consciousness and ally type to determine whether stigma consciousness moderated the effec t of ally type on each of the dependent variable s (see Table 1 2 ) . Marginalized group type was entered into the first step of the model as a control variable. Individuals with higher levels of stigma consciousness in the effective ally condition were more l ikely to report significantly higher levels psychological safety, lower levels of anxiety and higher levels of positive affect compared to those who were lower in stigma consciousness (see Figures 2 - 4 ). Conversely , individuals with 68 higher levels of stigma consciousness in the ineffective ally condition were more likely to report significantly lower levels of psychological safety, lower positive affect and greater anxiety. Table 1 2 : Moderated Regression of Stigma Consciousness Predicting Outcomes Psychological Safety Anxiety Positive Affect Negative Affect Voice Step 1 Marginalized Group Type - .03 (.13) - .04 (.10) .07 (.13) .16* (.13) .02 (.14) R 2 <.01 .01 .02 .02 <.01 p of R 2 Change .74 .14 .06 .05 .52 Step 2 Mean Centered Stigma Consciousness .00 (.07) .19** (.05) - .23** (.07) .03 (.06) - .08 (.07) R 2 <.01 .04 .06 .02 <.01 p of R 2 Change .93 .01 <.01** .77 .29 Step 3 Ally Type .45** (.06) - .72** (.05) .48** (.06) - .47** (.06) . 05 (.07) R 2 .20** .56** .29** .24 .01 p of R 2 Change <.01** <.01** <.01** <.01** .45 Step 4 Ally Type x Stigma Consciousness .14* (.06) - .14** (.05) .14* (. 0 6) - .03 ( . 06) - .06 (.07) R 2 .22** .58** .31** .24 .01 p of R 2 Change .03* <.01** <.01** .65 .39 Note: * p < 0.05 ; ** p < 0.0 1 . Values in parentheses are the standard errors of the regression estimates. Step 2 is a regression in which stigma consciousness is predicting outcome variables. 69 Figure 2. Effect of Ally Type on Stigma Consciousness / Psychological Safety Interaction 70 Figure 3. Effect of Ally Type on Stigma Consciousness / Anxiety Interac tion 71 Figure 4. Effect of Ally Type on Stigma Consciousness / P ositive Affect Interaction 72 The second series of moderated regression models tested the interactive effect of identity centrality and ally type to determine whether identity centrality moderated the effect of ally type on each of the dependent variables (see Table 1 3 ) . Marginalized group type was entered into the first step of the model as a control variable. Individuals with higher levels of identity centrality in the effective ally condition were more likely to report significantly higher levels psychological safety, lower levels of anxiety and higher levels of positive affect compared to those who were lowe r in identity centralit y (see Figures 5 - 7) . Conversely, identity centrality appeared to be unrelated to outcomes in the ineffective ally condition. Table 1 3 : Moderated Regression of Identity Centrality Predicting Outcomes Psychological Safety Anxiety Positive Affect Negative Affect Voice Step 1 Marginalized Group Type - .0 4 (.1 3 ) - .0 8 (. 09 ) . 12 (.13) .1 5 * (.1 2 ) .0 7 (.1 3 ) R 2 <.01 .01 .02 .02 <.01 p of R 2 Change .74 .14 .06 .05 .52 Step 2 Mean Centered Identity Centrality . 18** (.0 6 ) - .10 (.05) .2 3 ** (.0 6 ) - .15* (.06) .29** (.07) R 2 .04 .0 2 .0 7** .0 4* .09** p of R 2 Change .93 . 14 <.01** . 03* <.01** Step 3 Ally Type .45** (.06) - .72** (.05) .4 7 ** (.06) - .47** (.06) . 05 (.07) R 2 .2 4 ** .5 3 ** .29** .2 6** .0 9** p of R 2 Change <.01** <.01** <.01** <.01** .45 Step 4 Ally Type x Identity Centrality .1 7 * * (.06) - .1 0 * (.05) .14* (. 0 6) - .0 4 ( . 06) .03 (.07) R 2 .2 7 ** .5 4 ** . 31 ** .2 6** .0 9** p of R 2 Change <.01** .03* 0 2 * . 51 . 68 Note: * p < 0.05 ; ** p < 0.0 1 . Values in parentheses are the standard errors of the regression estimates. Step 2 is a regression in which identity centrality is predicting outcome variables. 73 Figure 5. Effect of Ally Type on Identity Centrality / Psychological Safety Interaction 74 Figure 6. Effect of Ally Type on Identity Centrality / Anxiety Interaction 75 Figure 7. Effect of Ally Type on Identity Centrality / P ositive Affect Interaction The final series of moderated regression models tested the interactive effect of perceived discrimination and ally type to determine whether perceived discrimination moderated the effect of ally type on each of the dependent variables (see Table 1 4 ) . Marginalized group type was entered into the first step of the m odel as a control variable. Individuals with higher levels of perceived discrimination in the effective ally condition were more likely to report significantly higher levels of anxiety and lower of positive affect compared to those who were lower in perceived discrimination (see Figures 8 and 9 ). Conversely, perceived discrimination had no 76 relation to anxiety for individuals in the ineffective ally condition, but, paradoxically, it appeared to be positively associated with positive affect for indiv iduals in the ineffective ally condition. Table 1 4 : Moderated Regression of Perceived Discrimination Predicting Outcomes Psychological Safety Anxiety Positive Affect Negative Affect Voice Step 1 Marginalized Group Type - .0 6 (.1 3 ) - .0 6 (. 10 ) . 10 (.13) .1 8* * (.1 1 ) .0 3 (.1 4 ) R 2 <.01 .01 .02 .02 <.01 p of R 2 Change .74 .14 .06 .05 .52 Step 2 Mean Centered Perceived Discrimination - . 19** (.0 6 ) .11* (.05) . 02 ( .07 ) .36** (.06) .00 (.07) R 2 .04* .0 2 .0 2 . 15** <.01 p of R 2 Change <.01** . 11 .88 <.01** .94 Step 3 Ally Type .45** (.06) - .72** (.05) .4 8 ** (.0 7 ) - .47** (.06) . 05 (.07) R 2 .2 4 ** .5 3 ** .2 4 ** . 37** <.01 p of R 2 Change <.01** <.01** <.01** <.01** .45 Step 4 Ally Type x Perceived Discrimination - .03 (.06) . 1 1 * (.05) - .1 9 * * (. 07 ) .03 ( . 06) - .14 (.07) R 2 .2 4 ** .5 5 ** . 28 ** . 37** .02 p of R 2 Change .69 .03* <.01** . 65 . 05 Note: * p < 0.05 ; ** p < 0.0 1 . Values in parentheses are the standard errors of the regression estimates. Step 2 is a regression in which perceived discrimination is predicting outcome variables. 77 Figure 8. Effect of Ally Type on Perceived Discrimination / Anxiety Interaction 78 Figure 9. Effect of Ally Type on Perceived Discrimination / Positive Affect Interaction Because the control variable, marginalized group type, was found to be a significant predictor of negative affect in all of the final models described above, I decided to look at the mean differences between mar ginalized group types on this outcome. Through examination of the means, it was revealed that individuals from sexual - and gender - identity marginalized groups ( M = 2.13, SD = 1.02) reported higher negative affect compared to individuals from racial or ethn ic marginalized groups ( M = 1.86, SD = .93). A follow up independent samples T test determined that these mean differences were not significant; t (198 )= - .195 , p = .053 . 79 STUDY 1 DISCUSSION Overall, the results of the first study provide strong support for H ypothesis 3 a and 3b . Individuals who were asked to describe an event in which they experienced an ineffective ally were more likely to also report a significantly greater degree of anxiety and negative affect and a significantly lower degree of positive affect and psychological safety than individuals who described an event in which they experienced an effective all y . In addition, individuals in the ineffective ally condition evaluated their ally more negatively than those in the effective ally c ondition . These findings suggest that the effective allies have the potential to increase feelings of inclusion and relief for members of marginalized groups , that these allies are viewed positively from an interpersonal perspective and that they are viewe d as capable of providing help . In addition, these findings su pport the assertion that that ineffective allies exist and are not a neutral force. Indeed, ineffective allies were found to have detrimental effects on the majority of psychological outcomes ex amined. However, there was no effect of ally type on voice. This suggests that regardless of whether an ally is effective or ineffective, marginalized group members feel similar levels of comfort when it comes to giving feedback to said ally. The results of this study also suggest that the perceived effectiveness of all y behavior may depend on the individual differences of the person who is receiving help. Specifically, individuals with higher levels of stigma consciousness and higher levels of identity ce ntrality who were in the effective ally condition reported a higher degree of psychological safety and positive affect and a lower degree of anxiety. Individuals with higher levels of stigma consciousness in the ineffective ally condition were lower on man y of the psychological outcomes whereas identity centrality was not related to lower psychological outcomes in the ineffective ally condition. Individuals with higher levels of perceived discrimination reported a 80 higher degree of anxiety and a lower degree of positive affect compared to individuals with a lower level of perceived discrimination in the effective ally condition and the inverse was found for individuals with higher perceived discrimination in the ineffective ally condition . These findings bols ter the idea that marginalized group members are not a monolith and that when one engages in allyship, the wants and needs of the specific individual they are attempting to help must be considered . 81 STUDY 2 METHOD Study Purpose The purpose of the second study was to create items reflecting typical ally behaviors using themes uncovered during the first study and those uncovered in the literature review from Table 2. These items were then analyzed using an exploratory factor analys is and a reliability analysis to determine the factor structure and to select the best items for the measure. Item Generation Items were generated based on the critical incidents provided by participants in the first study. I created 1 to 2 items from e ach of the 21 7 6 participants. In total, during the initial item generation phase, 422 potential items were generated. From there, I used Atlas.Ti to code the items and sort through them for content redundancy. Specifically, I used the open coding function to manually create codes for the entire set of items. From this process, 119 initial codes emerged. Codes with more than one item categorized within them were retained resulting in 84 codes remaining during the intermediate phase of coding (see Table 1 5 fo r top ten codes) . To ensure that the most frequently reported behaviors were represented and considered, I elected to keep one to three items from each of these codes to consider for analyses. 6 Note that while only the first completion was counted for analysis in the first study, those who completed the survey multiple times were included when creating additional items. Therefore, there is a discrepancy between Ns here. 82 Table 1 5 : ABQ Top 10 Most Frequently Occurring Codes from Study 1 Data Code Number of Occurrences Example Item Ally Intervention Negative Outcome 25 My ally's actions have had a negative effect on me. Spoke up Against Biased Comments/Slurs 21 My ally spoke up when they saw that others were making me uncomforta ble. Offensive Defense 21 My ally said something offensive in an attempt to defend me from bul lying Stereotyping Person in Need 17 My ally used positive stereotypes of my minority group to convince o thers to leave me alone. Aggressive Intervention 13 My al ly engaged in a yelling match with the person who used biased comments against me. Advocating for Fair Treatment 12 My ally made sure I was treated fairly when others tried to discrimin ate against me. Unwanted Help 11 My ally spoke up for me when I did not feel I was being discriminated against. Ignorant Allyship 9 My ally thinks they understand issues that affect my identity group but they do not. Taught majority members about minorities 9 My ally educated others about issues that affect my mino rity group. Voice Amplification 9 My ally amplified my voice as I spoke up against injustice I was facing. From there, I drew one to three items from each of the coding categories with more items being drawn from codes containing more items. Doing this reduced the number of items from 422 to 185 items that reflected the most frequently occurring ally behaviors (helpful and not 83 helpful). Eighty - eight items reflected ineffective behaviors, eighty - seven items reflected effective behaviors and ten items reflected perceptions of the outcomes of ally behavio r . I decided to separate out the outcome perceptions because the outcome may be beyond the control 10 members of the Diversity Lab at Michigan State University who assessed the construct validity and overall quality of these items as s ubject matter experts (SMEs). SMEs were asked to rate the degree to at A G reat E xtent) scale. For the outcome items, SMEs were asked to rate t he degree to which items of this type represented a well phrased or comprehendible item using a (Not at A G reat E xtent) scale. I calculated the grand mean of the scores within each type of item and across all items. The grand mean acros s all items was 3.60. The grand mean of the ineffective behavioral items was 3.48. The grand mean across effective behavioral items was 3.72. The grand mean across outcome items was 3.65. Items with mean scores lower than the grand mean within each item ty pe were removed from consideration. Two items involving yelling at the person engaging in harassment and physically fighting the person who was engaging in harassment were retained because of their overwhelming prevalence in the critical incidents despite all items of this type having lower - than - grand - mean mean scores. Two items involving deescalating the situation and helping minority members navigate predominately majority - centric environments were retained because of their prevalence in the critical inci dents despite items of this type having lower - than - grand mean scores. This resulted in the retention of 53 ineffective behavioral items, 51 effective behavioral items, and four outcome items for a total of 108 items retained. 84 edited those items that they deemed needed improvements or edits. The edited versions of the final items were then presented to participants in the second study ( see Appendix L ) . Participants For the second study, participants were recruited from Amazon Mechanical Turk (Mturk). All participants were Americans who were racial and ethnic minorities or Americans who were sexual or gender minorities (e.g., LGBTQ+). Pur posive sampling was used to be sure that participants from these two specific marginalized groups were sampled. All participants were at least 18 years of age, were American citizens who had been living in the U.S. for the past five years and were fluent i n English. Participants also reported working more than 30 hours per week and that they had experienced allyship behaviors from a member of a member of their respective majority group in the past 12 months in order to qualify. To be certain that participan ts were not bots, a CAPTCHA was also implemented and a qualitative question was asked. Participants who met all of these criteria were selected to participate in this study and received $1.50 for their participation in this study. The 2 25 p articipants in t his study were all members of marginalized racial, ethnic, sexual and gender identity groups . With consideration to race and ethnicity, 3 6.7 % of participants indicated that they were White - American, 2 4 . 2 % indicated that they were Black - American, 1 1 . 9 % indi cated that they were Asian - American, 1 3.8 % indicated that they were Hispanic - or Latino - American 5.7 % indicated that they belong to two or more racial/ethnic identity groups and 4.3 % indicated that they were American - Indian. With consideration to sexual or ientation, 4 5.5 % of participants indicated that they were heterosexual, 18.5 % identified as homosexual, 2 6.1 % identified as bisexual, 5 .7 % identified as pansexual and 1.9 % identified as asexual and 85 1.9% of participants indicated that their sexuality was n ot listed and specified it via a fill in the blank option . With consideration to gender identity, 38.4 % of participants identified as male, 4 9.3 % of participants identified as female, 5.7 % of participants identified as non - binary and 5.7 % of participants i dentified as transgender . On average, participants were 3 1 years old ( SD = 7.92 ) , made 3 0,000 - 3 9,000 dollars annually (min = > $10,000 max = $150,000 <) and a little over half 52 . 8 %). Measures Open - Ended Ally Experience Question. For quality control purposes, all participants were asked to detail a previous experience with allyship. For this study , their experiences could be positive or negative but participants were not purposefully grouped based on this. The quality of written responses was used as a secondary mechanism to ensure that (1) participants were not bots , (2) participants were actually minority group members , and (3) participants understood the question being asked. Allyship Behaviors Questionnaire . All participants were asked to respond to the 10 8 items created for the present study. Perceived Ally Characteristics Scale (PACS). To ass ess the degree to which the allies described in the critical incident have the qualities outlined by Brown and Ostrove (2013), participants were asked to report on the degree to which their ally is affirming and engaged in informed action (see Appendix B). While item content in the appendix was written for participants who are racial/ethnic minorities, I created and provided an altered form of the survey for sexual minorities. This measure was included for validity - related purposes. The affirmation 86 sub - scal e exhibited sufficient reliability ( = . 86) while the informed action subscale exhibit ed moderate reliability ( = . 68 ). Demographic Questionnaire. Lastly, a demographic questionnaire was administered to all participants. These questions included asking gender, and sexual orientation (see Appendix I ). Procedure Participants completed a series of prescreening questions in order to determine whether they qualify for the study. Qualifying participants were for warded on to the informed consent (see Appendix J ). If they chose to participate, participants were then asked to think of a previous time they experienced allyship over the past 12 months and then asked to detail it qualitatively. Participants then respon ded to the new allyship behavior questions and the PACS based on that experience. Lastly, participants were asked to complete demographic questions before receiving their payment code and being debriefed (see Appendix K ). 87 STUDY 2 RESULTS Data Cleaning In order to ensure the quality of the data, some participants were excluded from the analyses. Participants who failed the pre - screening questions, provided nonsensical qualitative responses or otherwise did not qualify were not allowed to partic ipate in the study. These filters and the qualification check at the end of the survey which was used to verify that a participant was indeed a member of the marginalized group resulted in 409 participation attempts being screened out from the racial and e thnic minority sample and 443 participation attempts being screened out from the sexual and gender minority sample before data cleaning. Once these participants were properly excluded, additional participants were screened out for other reasons. This inclu ded participants who completed very little of the survey and those who did not answer two or more of the preliminary ally behavioral questions (14 participants in total). Following these exclusions, a final sample of 225 participants remained (115 racial a nd ethnic minority and 105 sexual and gender minority). Qualitative Data Analyses Participants were not asked to specifically provide an ineffective or effective allyship experience for this study and were allowed to instead provide an ally experience of their choosing . However, they were asked to describe whether their relationship with the ally participants mentioned that their relationship with their ally i mproved following the incident they mentioned (65.4%) while some said that their relationship with their ally stayed the same 88 (32.2%) and few said that the relationship worsened (1.4%). This suggests that the majority of participants in this sample may hav e been describing a positive allyship experience. Quantitative Data Analyses Following data cleaning, I aimed to reduce the number of items to achieve a measure with a practical scale length. I entered all of the 10 8 ally behavioral items into an initial exploratory factor analysis with principle axis factoring and a varimax rotation. In order to determine the number of factors to retain, I relied upon information from the scree plot and the size of the eigenvalues. Upon investigating the scree plot for th is analysis, it appeared that the scree plot suggested that there were three factors. According to the scree plot, the second eigenvalue is roughly four times the size of the third eigenvalue, thus suggesting that these items exhibit a strong two factor st ructure (Ahn & Horenstein, 2013). However, because the scree plot suggested that there were at least three factors, I decided to run a series of exploratory factor analyses (one factor three factors) to determine the structure that would yield the most c onceptual meaning. The results of the one factor EFA did not yield a high degree of conceptual meaning. By this, I mean, items that were designed to represent effective ally behavior and items designed to represent ineffective ally behavior did not all l oad on a single factor in a meaningful way. In addition, the size of the first factor eigenvalue was only twice the size of the second eigenvalue. Because of this, I elected to consider the two - factor structure. The results of the two factor EFA did yield interpretable and parsimonious results. Specifically, all items that were designed to be ineffective ally behaviors emerged on the first factor and all items designed to be effective ally behaviors emerged on the second factor. Only 89 eight of the 103 items exhibited high cross - loadings with the other factor and only four items did not load strongly on their respected factor 7 . While the results of the three factor EFA yielded slightly more variance explained, the factor structure was less parsimonious than the two - factor structure. The first factor to emerge was the ineffective ally behaviors. The second factor to emerge was a subset of the effective ally behaviors with a tendency to include behaviors that involved taking action during instances of overt di scrimination (e.g., speaking up, defending others). The final factor to emerge was the subset of effective ally behaviors with a tendency towards advocacy, emotional support, educating others. While all but four items loaded strongly on their respective fa ctor (.40 or above) many of the items in the latter two factors exhibited large cross loadings on another factor (.30 or above). Most commonly, it was the factors that contained the effective ally behavioral items that were found to have the most cross loa dings with each other. The high cross loadings between the two factors containing effective ally behavioral items and the fact that the eigenvalue of the third factor is roughly one fourth the size of the second factor may suggest that there is evidence fo r the two - factor structure. Because of this, and because the two - factor structure is more parsimonious and theoretically driven, I elected to pursue the two - factor structure rather than pursue other potential structures. Item Selection for the Final Scal e Once I decided to pursue a bidimensional measure, I needed to decide which items to include for the final scale. To further pair down the items, I decided to conduct a series of exploratory factor analyses with principle axis factoring and a varimax rotati on (see Furr & 7 Tw o of the items that did not strongly load on their respective factor also exhibited high cross loadings with the alternate factor. Thus, a total of ten items were considered for factorial removal at this step. 90 Bacharach, 2008 for discussion of approach; see Post & Walma van der Molen, 2019 and et al., 2019 for recent scale development examples employing this approach) . During the first one, I removed items that were deemed problemat ic due to high cross loadings or low loadings on their respective factors. During the second EFA, I elected to remove items that were redundant with other content. Specifically, if there were two or more items that were similar in terms of content, I would remove the item(s) that had the lower factor loading. This resulted in only 50 items remaining retained. During the third EFA, I retained items from the ineffective behavior factor that had factor loadings above .70 and items from the effective behavior f actor that had factor loadings above .60. This resulted in only 34 items remain ing retained. During the fourth EFA, I retained only the top twelve loading items on each of the respective factors that were also the most conceptually distinct from one anothe r. This resulted in the final 2 4 item scale (see Table 1 6 ) in which all items loaded at least .60 on their respective factor and the internal consistency of the effective behavior items was = .90 and the internal consistency of the ineffective behavior i tems was = .95 . Table 1 6 : Factor Structure of Final Selected Items Item Number Item Content Factor 1 Factor 2 1 ABQ Item 70 Shared personal information about me that negatively affected me. .871 - .093 2 ABQ Item 62 Gave me advice I found to be offensive. .856 - .124 3 ABQ Item 49 Refused to listen to my concerns. .811 - .086 4 ABQ Item 91 Asked me inappropriate questions about my identity. .799 - .139 5 ABQ Item 44 Did not put much effort into protecting me. .792 - .054 6 ABQ Item 9 Made offensive comments about my group. .786 - .011 7 ABQ Item 4 Told me I don't act like a normal member of my group. .784 - .025 8 ABQ Item 65 Made my circumstances worse by drawing attention to my minority status. .782 - .206 91 Table 16 : 9 ABQ Item 82 Restricted me from being able to fully express myself. .770 - .181 10 ABQ Item 90 Did not address the offensive behavior of others. .761 - .084 11 ABQ Item 87 Made excuses for the biased comments of others. .709 .029 12 ABQ Item 84 Was patronizing towards me. .708 - .116 13 ABQ Item 99 Made sure that my needs were considered. - .100 .778 14 ABQ Item 74 Made me feel heard. - .140 .701 15 ABQ Item 96 Helped communicate my thoughts effectively to those with privilege. .065 .689 16 ABQ Item 32 Spoke up in a way that provided for me to advocate for myself. - .022 .676 17 ABQ Item 18 Helped others understand more about my identity group. .052 .675 18 ABQ Item 15 Insisted that others treat me like a human being. - .169 .657 19 ABQ Item 60 Defended me against biased statements in an educated way. - .058 .656 20 ABQ Item 75 Made me feel accepted when other people were purposefully excluding me. - .155 .645 21 ABQ Item 16 Advocated for me when I was treated worse than others. - .240 .634 22 ABQ Item 20 Spoke up for me when I was being ignored. .077 .628 23 ABQ Item 56 Protected me from verbal harassment. - .211 .606 24 ABQ Item 93 Rallied support for me when I was being discriminated against. - .100 .602 I conducted a confirmatory factor analysis (CFA) on the same data to confirm the factor structure of the ABQ items using the lavaan package in R. Effective items were loaded on one factor and ineffective items were loaded on the other factor. The results o f the CFA indicated that the two factor structure of ABQ items exhibited good fit (CLI = .93, TLI =.92, RMSEA =. 07, SRMR = .07). Conversely, a follow up, CFA revealed that when all items were loaded onto one factor, the ABQ items exhibited poor fit to the one factor structure (CLI = .64, TLI =.60, RMSEA =. 15, SRMR = .20). These results suggest strong support for a two - factor structure in which effective and ineffective allyship behaviors remain separate. I also conducted additional analyses to determine how the PACS measure related to scores on the newly developed ABQ items. Correlational results revealed that the items selected for the final version of the ABQ correlated highly with the PACS subscale. Both PACS subscales 92 were positively related to the effec tive ABQ score and both were negatively related to the ineffective ABQ score. Although these measures correlated in the expected directions, the magnitude of the correlations suggest that they are also conceptually distinct. This study does not directly te st the predictive power of the ABQ over the PACS, however, this could be an endeavor for future research. To view all bivariate correlations, see Table 1 7 . Table 1 7 : Study 2 Correlations between PACS and ABQ Scores ABQ Effective ABQ Ineffective PACS Informed Action PACS Affirmation ABQ Effective - ABQ Ineffective - .21** - PACS Informed Action .56** - .16* - PACS Affirmation .58** - .43** .48** - I also conducted independent samples t - tests to determine whether there were differences in ABQ scores between the two demographically defined samples in this study (e.g., ethnic and racial minorities and sexual/gender minorities). As depicted in Table 1 8 , there were no significant differences between the samples for the ABQ subscale scores and no significant differences between the samples for the PACS subscale scores. 93 Table 1 8 : T - tests to Examine Mean Differences Between Sample Types Racial and Ethnic Minority Sample Sexual and Gender Minority Sample t F sig Dependent Variable Mean Standard Deviation Mean Standard Deviation ABQ Effective 3.78 . 81 3. 85 .9 1 - .59 .78 .38 ABQ Ineffective 1.74 . 99 1.61 .95 1.03 1.53 .22 PACS Informed Action 3.47 . 74 3. 53 . 80 - .54 .74 .39 PACS Affirmation 4.17 . 79 4.27 . 91 - .93 1.59 .21 When examining the skew of each of the measures in the present study, the ABQ effective items revealed a negative skew (skew = - 1.04) while the ABQ ineffective items revealed a highly positively skew (skew = 1.39). This suggests that overall individuals in this 94 STUDY 2 DISCUSSION The exploratory factor structure of the ABQ revealed a two - factor structure in which effective behaviors fell onto one factor while ineffective behaviors fell onto the other factor. Confirmatory factor analyses confirmed the two - factor structure. Both subscales exhibited sufficient reliability as well. These findings as well as the low correlation between ineffective ABQ and effective ABQ scores suggest that effective and ineffective behaviors do not exist on the same continuum, rather, they are separate constructs. The fact that they are separate constructs suggests that individuals can engage in both effective and ineffective ally behaviors and potentially that they can engage in bot h types of behavior during a single helping attempt . The ABQ effective items correlated positively and moderately with the PACS subscales while the ABQ ineffective items correlated negatively with both PACS subscales and the relationship was stronger for t he affirmation PACS subscale. These findings suggest preliminary evidence of convergent validity for the ABQ subscales. Lastly, t here were nonsignificant differences between demographic sample types and scores on the measures used in this study . This sugge sts that there may be a universal component to perceptions of effective and ineffective ally behaviors. 95 STUDY 3 METHOD Study Purpose The purpose of the third and final study was to determine the motivations and individual differences related to eff ective and ineffective ally behavior. To do this, I provide d participants with scenarios involving workplace discrimination and bias and ask ed them to report what they would do in response. T wo skilled , undergraduate coders then code d these behaviors as ef fective and ineffective and interrater agreement was assessed. I then us ed motivations and individual differences to predict the degree to which majority group members engaged in effective ally behavior . I also test ed to see whether motivations predict beh aviors over and above other individual differences as an exploratory analysis . Because there is very little research on the individual differences related to allyship behavior, I elect ed to include many individual difference measures to build a nomological network of related constructs. Thus, this study test ed Hypotheses 1a, 1b and 2. Participants The 3 00 p articipants in this study were all members of racial, ethnic, sexual and gender ma jority groups (e.g., White, heterosexual and cisgendered) . Thus, 100 % of participants indicated that they were White - American and 100% participants indicated that they were heterosexual . With consideration to gender identity, 4 6.3 % of participants identifi ed as male, 53.7 % of participants identified as female . On average, participants were 3 8 years old ( SD = 11 .6 9 ) , made 50,000 - 59,000 dollars annually (min = > $10,000 max = $150,000 <) and most onal attainment ( 55 . 9 %). 96 All participants in this study were workers from Amazon Mechanical Turk (Mturk). Responses were collected from 300 participants from members of racial, ethnic and sexual majority groups (e.g., white , heterosexual cisgendered parti cipants). All participants were over the age of 18 , indicated that they were American citizens and indicated that they were fluent in English. The y also report ed that they work ed more than 30 hours per week to qualify. Participants receive d $2 .00 for their participation in this study . Measures Behavioral Measure of Allyship . Participants were provided with four scenarios based on the critical incident experiences provided in Study 1. Participants w ere asked to report what they would do if they were a bystander in the situation and to write out any specific actions they would take. The scenarios w ere designed to allow multiple behaviors to emerge. Free - response actions were coded/categorized into two groups: effective behaviors and ineffective behaviors . Afterward, I provide d some effective and ineffective behavioral options for participants to choose from and ask ed them to select the ones that they would engage in. These options w ere comprised of relevant items from the Allyship Behaviors Questionnaire. The sum of the free response effective ally behaviors across all scenarios served as an indicator of effective ally behavior. The sum of the free response ineffective ally behaviors across all scenarios served as an indicator of ineffective ally behavior . Th ese w ere the primary measure s of behavior for this study. Similar indices of effective behaviors and ineffective behaviors were calculated for the ally behavior selection list option (e.g., self - reported ally behaviors) . This was the ancillary measure o f behavior for this study (See Table 1 9 for all scenarios and behavioral response options/codes) . 97 Table 1 9 : Study 3 Scenarios and Response Options Scenario Response Options You are a manager at a fast food restaurant. One of your employees is being yelled at by a customer because your store has run out of chicken tenders. The customer is shouting angrily and using racial slurs against your employee. Ineffective Behaviors - Not listen to employee's concerns - Mention that employee is not typical member their racial group - Not address the offensive behavior of the customer - Mention alternative reasons for the behavior of the customer - Do nothing Effective Behaviors - Make sure needs of employee are considered - Defend employee against discrimination in educated way - Insist that the employee is treated as a human being - Protect employee from verbal harassment - Do nothing in moment but make sure employee is okay You are sitting in a break room at work when you overhear some coworkers making homophobic comments in the presence of John, another coworker, who is gay. While the other coworkers at the table are not directly addressing John, he is present while they discuss how much they disagree with the gay lifestyle and oppose g ay marriage. Ineffective Behaviors - Mention that John is gay and that he is sitting with them - Ask John to share details about his life so that his coworkers could be educated - Not address the offensive behavior of the coworkers - Mention alternative reasons fo r the behavior of the coworkers - Do nothing Effective Behaviors - Make sure John's needs considered - Defend John against biased comments in educated way - Speak up for John in a way that provides an opportunity for him to advocate for himself - Protect John from verbal harassment - Do nothing in moment but report the comment to HR - Do nothing in the moment but make sure John is okay afterward - Do nothing in the moment but privately confront coworkers later on 98 Table 19 : Your friend at work has been passed over for a promotion yet again although he is the most qualified one for the position. As you think back on it, you remember that everyone who was promoted over the past few years has been Non - Hispanic White while your friend is Hispanic. Ineffectiv e Behaviors - Not listen to friend's concerns - Draw attention to the fact the friend is Hispanic and has been passed over for a promotion - Not address the fact that the friend was passed over - Mention alternative reasons for the bosses' hiring choices (or sugge sting that the friend is deficient in some way) - Do nothing Effective Behaviors - Speak up for friend in a way that allows him to advocate for himself - Rally support for friend from employees/managers - Advocate for friend to those who make hiring decisions or HR - Make sure friend feels heard You are sitting in a design meeting for a new commercial featuring a gay couple. After reviewing the commercial, your teammate, Sharon, questions the interactions between the couple. Specifically, she questioned several aspects of the couple's relationship in an offensive and ignorant manner. Charles, who is gay, wrote the dialogue for the couple. He insists that what he wrote is an accurate depiction of a gay couple. Sharon rebuts with continued questi ons that grow more offensive over time. Ineffective Behaviors - - Mention that Charles is not a normal gay person - Say nothing about Sharon's questions - Mention alternative reasons for Sharon's behaviors - Do nothing Effective Behaviors - Make Charles feel heard - Defend Charles against Sharon's biased comments in an educated way - Speak up with Charles so that Sharon does not ignore his comments - Advocate for Charles' work and qualifications - Do nothing in the moment but make sure Charles is okay after the meeting - Do nothing in the moment but report Sharon to HR - Do nothing in the moment but confront Sharon about comments privately 99 Allyship Behaviors Questionnaire Majority Group Perceptions . All participants w ere asked questions about the 24 items in the Allyship Behaviors Questionnaire created previously. The questions examine d majority group perceptions of typical ally behaviors and determine d how frequently participants engage d in and would be willing to engage in these behaviors. l questions w ere responded to using a N ot at A V ery M uch or V ery F requently) scale. This measure was originally created for marginalized group members to report on ally behavior. These measures demonstrated sufficient reliability ( helpful = .9 2, frequency 0, willingness 3 ; see Appendix M ) . Social Dominance Orientation (SDO). To assess the degree to which individuals prefer hierarchy within society and power over low - status groups, the Social Dominance Orientation (SDO) was imp lemented (Pratto et al., 1994; see Appendix N ay in their demonstrate d 5 ). Motivation to Appear Non - Prejudiced. not appearing prejudiced the ten - item, Motivation to Respond without Prejudice Scale (Plant & Devine, 1998) was implemented (see Appendix O ) 8 . Participants respond ed 8 100 (Strongly Disagree) to ere designed to assess internal of my personal values, I believe that using stereotypes about minority groups is were designed to assess external motivation to respond without prejudice. concerned ere averaged within each subscale to create an index of internal and external motivations to appear non - prejudic ed respectfully. Higher scores are indicative of a greater level of internal and external motivation to appear non - prejudice d respectively. In previous research (Plant & Devine, 1998) , t his measure was subjected to an exploratory factor analysis and a conf irmatory factor analysis and there was support for the model fit of the two - factor structure (GFI = .96 and AGFI = .93). It also demonstrated evidence of convergent, discriminant validity and predictive validity in previous research . This measure exhibited 79 92 ) in the present study . Prosocial Motivations. To assess the degree to which participants engage in helping behavior for prosocial reasons, I adapted four items measuring p rosocial motivation from Grant (2008) prosocial motivation subscale . These items w ere edited to represent helping more broadly rather that focus on helping in the work context (see Appendix P ). Participants w ere asked to help they agree with four statemen D isagree S A gree S trongly). Example items 101 ere averaged together to create a total score. Higher sco res indicate d a greater degree of prosocial motivation. In previous research (Grant, 2008) , t hese items were subjected to exploratory and confirmatory factor analyses as part of a larger measure and there was support for model fit (NFI =.94, CFI = .96, SRM R = .04. This measure demonstrate d 1 ) in the present study . Selfishness. To assess the degree to which participants exhibit different forms of selfishness, the Selfishness Questionnaire (Raine & Uh, 2018) w as administered ( see Appendix Q ). While the paper by Raine and Uh (2018) contained a version of this measure in which there were only three response options (Disagree, Neither Agree nor Disagree, and Agree), I elected to expand the response options to the five standard res assessed egocentric items assessed up assessed pathological selfishness. Example items were averaged together wi thin each subscale to create an index of egocentric, adaptive and pathological selfishness. Items w ere also averaged together to create an overall index of selfishness. Higher scores indicate a greater degree of selfishness. This measure was subjected to an exploratory factor analysis and a confirmatory factor analysis and there was support for the model fit of the three - factor structure (CFI =.93, RMSEA = .07). It also demonstrated evidence of convergent, discriminant validity 102 and criterion validity. This measure exhibited sufficient reliability within each subscale (egocentric 5 81 , patholo 6 ) . Because of the high correlations between subscales in the present study, these items were all averaged together to form a single index of selfishness. HEXACO. T wo theoretically relevant subscales from the HEXACO (Ashton & Lee, 2009) were implemented to determine the degree to which participants exhibit two personality facets (see Appendix R ). These facets were openness to new experiences and honesty - humility. Participants w ere asked to respond to 2 Disagree) scale. I elected to use the long form items over the shorter form by de Vries (2013) because the short form had low internal consistency. These subscales demonstrated sufficient 8 3 , honesty - Political Orientation Assessment . This measure is designed to determine the political orientation that each participant best aligns with (Carney, Jost, Gosling, & Potter, 2008). Using a E xtremely L E xtremely C onservative) scale, participants responded to three items (s ee Appendix S following scale of liberalism - were averaged to creat conservative political orientation. This measure was found to demonstrate sufficient internal 92 ). Empathy . To assess the degree to which partic ipants engage in feeling empathic concern, the empathic concern subscale of the interpersonal reactivity index (IRI; Davis, 1980) was employed (see Appendix T ). Participants were (Does N ot D escribe M e W ell) t M e V ery W ell) scale. An example item 103 include s to these items were averaged together to compute a total score. Higher scores on this measure are indicat ive of higher levels of empathy. This measure was found to demonstrate sufficient internal consistency ( 85 ). Perspective Taking . To assess the degree to which participants take the perspective of others, the perspective taking subscale of the interp ersonal reactivity index (IRI; Davis, 1980) was employed (see Appendix T ). Participants were asked to respond to the seven items using a N ot Describe M e W M e V ery W ell) scale. An example item Responses to these items were averaged together to compute a total score. Higher scores on this measure are indicative of higher levels of perspective taking. This measure was found to demonstrate sufficient internal consistency ( 82 ). Moral Identity. The Self Importance of Moral Identity Scale (SIMI) was designed to assess the degree to which being a moral person is important to o sense of self (Aquino & Reed, 2000). Respondents were asked to consider a set of moral characteristics (e.g., caring, compassionate, fair, friendly, generous, helpful, hardworking, honest, and kind) and how a person with these characteristics would th ink, feel and act (see Appendix U - scale, participants were asked to select their good to be a per son who has these characteristics and I often buy products that communicate F ive items w ere averaged together to represent the meaning inter nal feelings regarding moral identity). The last five items w ere averaged to represent the degree to which 104 meaning how one presents their moral identity to other people). This measure wa s found to demonstrate sufficient reliability (internalization 89 73 ) in the present study . Moral Relativism. The Moral Relativism Scale (MRS) was designed to assess individual differences in the endorsement of moral relativism as a moral viewpoint (Collier - Spruel et al., 2019). By this, I mean, this scale assess es the degree to which a person believes that if an individual or culture disagreed with another individual or culture in terms of their moral rules , both parties could be right according to their own respective framework s (see Appendix V ). (Strongly Disagree) (Strongly Agree) , par ticipants were asked the Each person is the final authority on whether his or her actions really are morally correct. Two different cultures could have dissimilar moral rules and bot the responses to the items into a single score. Higher scores on the MRS indicate a higher degree of endorsement for moral relativism as a meta - ethical thinking pattern. In previous research (Collier - Sp ruel et al., 2019), This measure was subjected to an exploratory factor analysis and a confirmatory factor analysis and there was support for the model fit of the bifactor model (CFI = .97, TLI = .96, RMSEA = .069, SRMR = .043). It also demonstrated eviden ce of convergent and discriminant validity in previous research . This measure was found to demonstrate sufficient 5 ) in the present study . Willingness to Learn About Marginalized Groups . To assess the degree to which participants are willing to learn about the experiences of marginalized group members, five items were created (see Appendix W rate d the frequency with which they seek out information about the experiences of marginalized 105 group members Scores for this measure were calculated by averaging the responses to these items and computing a single score. Higher scores indicate a higher propensity to seek out information about marginalized groups. These items were created by me and have not been tested psyc hometrically. This measure was found to demonstrate 9 ). Close Relationships with Marginalized Groups. To assess the degree to which participants feel close to individuals from marginalized groups, two questions from Yip, S eaton were were asked to indicate the proport ion of the races of their close friends similar questions were asked friend groups and two questions were also asked regarding the proportion of gender - identity . Responses to each of these items indicate the self - reported proportion of friendships the participant has with minorities (racial, ethnic , s exual and gender ) and majority group members (White, heterosexual) respectively. Demographic Questionnaire. Lastly, a demographic questionnaire was administered to all participants. These questions included ity, gender, and sexual orientation (see Appendix I ). 106 Procedure Participants w ere asked to complete a series of prescreening questions in order to determine whether they qualify for the study. These questions included asking about their ethnicity and race, sexual orientation and gender identity to determine whether they were indeed majority group members. Qualifying participants w ere then forwarded on to the informed consent (see Appendix J ). A fter indicating that they wanted to participate , participants w ere shown four scenarios depicting workplace depicting another person from a marginalized background experiencing discriminatory or biased actions. Participants were asked to write about how th ey would re act to this situation . After typing their responses, participants w ere then asked to select all of the behaviors that they would engage from a pre - determined list comprised of items from the Allyship Behavior Questionnaire. They w ere also asked questions about their willingness to engage in each of the allyship behaviors and asked to complete several other individual difference measures. Lastly, participants w ere asked to complete demographic questions before receiving their payment code before b eing debriefed (see Appendix K ). 107 STUDY 3 RESULTS Data Cleaning In order to ensure the quality of the data, some participants were excluded from the analyses. A total of 709 attempts to enter and complete the study were made. Participants who failed the pre - screening questions, provided nonsensical qualitative response s or otherwise did not qualify were not allowed to participate in the study (385 participation attempts excluded) . These filters and the qualification check at the end of the survey which was used to verify that a participant was indeed a member of the req uested majority group s (White, heterosexual, cisgender) resulted in 10 participation attempts being screened out . Once these participants were properly excluded, additional participants were screened out for other reasons. This included participants who completed very little of the survey and those who did not answer two or more of the preliminary ally behavioral que stions (14 participants in total). Following these exclusions, a final sample of 300 participants remained . Descriptive Informatio n For a table containing all descriptive statistics for the third study please see Table 20 . For a table containing bivariate correlations between all measures in the present study, please see Table 2 1 . When examining the ways in which majority group members viewed the behaviors of the ABQ items, I found that on average participants reported that effective behaviors were helpful to marginalized group member s ( M = 4.1, SD = .67), that they frequently engage in these behaviors ( M = 3.9, SD = .62), and that they would be willing to engage in these behaviors on behalf of marginalized group members ( M = 4.1, SD = .71). 108 Table 20 : Des criptive Statistics for All Measures in Study 3 Measure Mean Standard Deviation Minimum Maximum Reliability Prosocial Motivation 4.20 .84 1.00 5.00 . 91 External Motivation to Appear Non - Prejudiced 4.47 2.37 1.00 9.00 . 92 Internal Motivation to Appear Non - Prejudiced 7.15 1.64 1.00 9.00 .7 9 Social Dominance Orientation 2.50 1.35 1.00 7.00 . 95 Egocentric Selfishness 2.39 .84 1.00 5.00 . 85 Pathological Selfishness 2.29 .84 1.00 5.00 . 86 Adaptive Selfishness 2.79 .85 1.00 5.00 . 81 Total Selfishness 2.48 .77 1.00 5.00 . 93 Openness 3.60 .77 1.10 5.00 .8 3 Honesty Humility 3.41 .74 1.40 5.00 .79 Political Orientation 3.83 1.81 1.00 7.00 .92 Empathic Concern 3.78 .77 1.00 5.00 .85 Perspective Taking 3.80 .78 1.43 5.00 . 82 Moral Relativism 2.94 .76 1.00 5.00 . 85 Internalization 5.95 1.05 2.60 7.00 . 89 Symbolization 3.92 1.53 1.00 7.00 . 73 109 Table 20. Willingness to Learn About Marginalized Groups 3.45 .98 1.00 5.00 .8 9 Perceptions of Behavior: Helpful 4.10 .67 1.42 5.00 . 92 Perceptions of Behavior: Frequency 3.91 .62 1.42 5.00 .8 8 Perceptions of Behavior: Willing to Act 4.10 .71 1.25 5.00 .9 3 110 Table 2 1 : Bivariate Correlations Between all Measures in Study 3 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 1. Prosocial Motivation - 2. External Motivation to Appear Non - Prejudiced .04 - 3. Internal Motivation to Appear Non - Prejudiced .57** - .12* - 4, Social Dominance Orientation - .47** .20** - .64** - 5. Egocentric Selfishness - .39** .33** - .44** .49** - 6. Pathological Selfishness - .29** .40** - .41** .53** .78** - 7. Adaptive Selfishness - .31** .29** - .26** .36** .73** .75** - 8. Total Selfishness - .36** .38** - .41** .51** .92** .93** .90** - 9. Openness .28** - .21** .27** - .32** - .35** - .29** - .19** - .31** - 10. Honesty Humility .17** - .42** .33** - .29** - .55** - .62** - .55** - .63** .18** - 11. Political Orientation - .21** .19** - .33** .50** .26** .29** .18** .27** - .36** - .21** - 12. Empathic Concern .55** - .23** .54** - .60** - .63** - .59** - .51** - .63** .38** .49** - .31** - 13. Perspective Taking .45** - .18** .44** - .48** - .44** - .46** - .35** - .46** .38** .38** - .17** .71** - 14. Moral Relativism .01 .16** - .06 - .09 .14** .15* .12* .15* .06 - .18** - .15** - .08 .03 - 15. Internalization .42** - .12* .52** - .45** - .41** - .39** - .21** - .37** .25** .27** - .11* .53** .48** - .11 - 16. Symbolization .30** .33** .04 .08 - .05 .08 - .12* - .03 .05 - .13* .15* .05 .08 .10 .01 - 17. Willingness to Learn About Marginalized Groups .50** .03 .40** - .41** - .29** - .21** - .24** - .27** .47** .16** - .37** .43** .35** .16** .19** .30** - 18. Perceptions of Behavior: Helpful .48** - .17** .69** - .68** - .46** - .49** - .26** - .45** .33** .30** - .33** .58** .50** - .03 .64** - .10 .35** - 19. Perceptions of Behavior: Frequency .55** - .15** .70** - .67** - .49** - .51** - .33** - .49** .35** .32** - .35** .62** .57** - .02 .61** .03 .46** .85** - 20. Perceptions of Behavior: Willing to Act .53** - .18** .71** - .70** - .50** - .50** - .29** - .48** .37** .33** - .37** .62** .55** - .02 .62** - .04 .43** .91** .89** - 111 Qualitative Coding Information Two trained, undergraduate coders examined participant responses to each of the four scenarios presented in this study. For each scenario, a set of relevant response options were selected from the Ally Behavior Questionnaire to serve as both response optio ns for participants and codes for the coders (see Table 1 9 for scenario and response options). Coders were asked to sum up the effective behaviors and ineffective behaviors separately to serve as indexes of participant behaviors. For the first round of cod ing, coders were asked to code freely using the categories described in Table 1 9 . T hen, the princip al investigator reviewed the coding work and provided general feedback to both coders regarding more stringent interpretations of the codes. Specifically, I asked coders to try to fit free - response codes into existing codes and clarified the meaning of codes so that both coders would be better able to report the presence of behaviors in participant responses. In order to examine agreement between the coders o n the count data , intra - class - correlations (ICC) were implemented 9 following the second round of coding . Specifically, I used two - way random effects model s with absolute agreement as the index. This version of the ICC was selected because the two raters we re a sample of potential raters rather than the entirety of a population of raters and because both raters rated the entire set of participant responses (see Hallgren, 2012; Koo & Li, 2016; Landers, 2015 for discussion of this analytic technique). There ap pears to be debate regarding interpretation of ICCs and the thresholds for poor, fair, good and cited work (3,292 citations), ICCs above .70 are deemed to indicat 9 categorical. This analytic technique was initially considered but ultimately not selected because we were seeking agreement on con tinuous ratio data in addition to the other considerations discussed in the body of the paper. 112 raters. Thus, this is the interpretation I elected to use for the present study. The results of these models suggest that there was moderate to decent agreement between raters overall across scenarios (see Table 2 2 ) . Table 2 2 : Summary of Int ra class Correlation Results f or Scenario Behavioral Coding Scenario ICC 95% Confidence Interval F p 1 Effective .73 .69 .42 - .79 3.67 <.01 Ineffective .81 .81 .76 - .85 5.33 <.01 2 Effective .72 .72 .65 - .78 3.53 <.01 Ineffective .82 .81 .76 - .86 5.59 <.01 3 Effective .70 .64 .39 - .77 3.31 <.01 Ineffective .53 .43 .09 - .62 2.14 <.01 4 Effective .70 .63 .36 - .76 3.29 <.01 Ineffective .74 .73 .66 - .79 3.86 <.01 Note : ICC refers to average measure ICC After computing the general levels of agreement, the princi pal investigator examine d participant responses to each scenario of available codes and comments left by each coder regarding potential additional positive or negative behaviors to resolve remaining discrepancies between coders. Examining each of these elements revealed that discrepancies between coder sums in scenario 1 were due to one rater systematically adding one additional behavior to sum positive majority codes . Discrepanc ies between coder sums in scenario 2 were primarily due to using the write - in option to describe behavior that was redundant with existing behavioral codes. Lastly, discrepancies between coders in scenario 4 were primarily due to one rater systematically using one of the positive behavioral codes incorrectly, inflating the sum by 113 one. Bec ause of low agreement on scenario 3 after two rounds of coding and because compared to the other three scenarios it is the only one during which the participant needed to advocate to higher levels of the organization rather than interpersonally , I decided to remove it from the criterion for study 3. Instead, count values from scenarios 1, 2 and 4 were summed together as an index of ally behavior for the following analyses. On average, participants engaged in 1.10 1.40 effective behaviors and .30 - .45 ine ffective behaviors suggesting that effective behaviors were more likely to be reported. Quantitative Data Analyses To determine the individual differences related to effective and ineffective ally behavior, a series of multiple regression analyses and hier archal regression analyses were conducted. This series of analyses tested Hypotheses 1a, 1b, and 2. For the first set of analyses, I tested Hypothesis 1a ( Individuals who are higher in other oriented - motivations will be more likely to engage in effective a lly behavior than those lower in other - oriented motivations ) and hypothesis 1b ( Individuals who are higher in self - oriented motivations will be more likely to engage in ineffective allyship behavior than those lower in these motivations ) using a series of multiple regression analyses. Four models were run in which prosocial motivations, external motivation to respond without prejudice and internal motivation to respond without prejudice were predictors and the behaviors (effective coded behaviors, ineffecti ve coded behaviors, effective self - report behaviors, ineffective self - report behaviors) were entered into each model as the dependent variable. For results of all four models, see Table 2 3 . 114 Table 2 3 : MR of Motivations Predicting Allyship Behaviors DV R 2 F Predictor SE p Effective Coded Behaviors .18** 21.71 Prosocial Motivation - .01 .13 .85 External Motivation to Respond Without Prejudice - .06 .04 .26 Internal Motivation to Respond Without Prejudice .42 ** .07 <.01 Ineffective Coded Behaviors .18** 21.19 Prosocial Motivation - .11 .12 .08 External Motivation to Respond Without Prejudice .01 .03 .91 Internal Motivation to Respond Without Prejudice - .34 ** .06 <.01 Effective Self - Report Behaviors .2 2 ** 2 8 . 49 Prosocial Motivation .21 ** .34 <.01 External Motivation to Respond Without Prejudice - .0 8 .10 .1 3 Internal Motivation to Respond Without Prejudice .3 1 ** .18 <.01 Ineffective Self - Report Behaviors . 11 ** 11 . 70 Prosocial Motivation - . 03 .1 0 . 63 External Motivation to Respond Without Prejudice . 09 .03 . 12 Internal Motivation to Respond Without Prejudice - . 29 ** .05 <.01 115 Results indicate that internal motivation to respond without prejudice was the strongest predictor of effective and ineffective ally behaviors. This construct positively predicted both coded and self - reported effective allyship behaviors. This construct negatively predicted both coded and self - reported ineffective allyship behaviors. In addition, effective self - reported allyship behavior was also positively predicted by prosocial motivation . However, I note that this predictive relationship may have occurred due to social desirability when participants were reporting self - reported behaviors or due to common method bias r elated to self - report. I note this particularly because this finding did not emerge for the coded effective behaviors. These findings suggest that allies who are motivated to respond without prejudice for internal reasons are more likely to engage in effec tive ally behaviors and less likely to engage in ineffective behaviors. For the second set of analyses, I tested Hypothesis 2 ( Effective allies will be (a) more empathetic, (b) be more likely to engage in perspective taking, (c) be more liberal, (d) have lower social dominance orientation (e) exhibit a lower degree of selfishness (f) exhibit a lower degree of moral relativism ( g ) have a greater degree of internalized moral identity ( h ) have a lower degree of symbolized moral identity ( i ) be more willi ng to learn about marginalized groups ( j ) have more close relationships with people in marginalized groups ( k ) will have a higher degree of honesty/humility and openness to new experiences compared to ineffective allies ) using a series of multiple regressi on analyses. Four models were run in which empathy, perspective taking, political orientation, moral relativism, moral identity, selfishness, social dominance orientation, honesty - humility, openness to new experience, willingness to learn about marginalize d groups and close relationships with marginalized groups were predictors and the behaviors (effective coded behaviors, ineffective coded behaviors, effective self - report 116 behaviors, ineffective self - report behaviors) were entered into each model as the dep endent variable. For results of all four models, see Table 2 4 - 2 7 . Table 2 4 : MR of Individual Differences Predicting Effective Coded Behaviors DV R 2 F Predictor SE P Effective Coded Behaviors .26** 7.08 Social Dominance Orientation - .18 ** .09 .02 Selfishness .08 .17 .31 Openness to New Experience .08 .14 .20 Honesty - Humility .12 .16 .09 Political Orientation - .14 ** .06 .04 Empathy .02 .20 .85 Perspective Taking .02 .17 .79 Moral Relativism .04 .12 .44 Internalization .16 ** .11 .02 Symbolization - .05 .06 .38 Willingness to Learn About Marginalized Groups .07 .12 .35 Racial Diversity of Friends - .05 .11 .39 Sexual Diversity of Friends .05 .16 .49 Gender - Identity Diversity of Friends - .10 .14 .16 117 Results from the multiple regression predicting effective cod ed behavior indicate that lower social dominance orientation and political orientation (e.g., conservatism) negatively predict these behaviors . Self - importance of moral identity for internalized reasons positively predicted effective coded behavior . Th ese findings suggest that , allies who are lower on social dominance orientation, less politically conservative and had higher levels of internalized self - importance of moral identities are more likely to engage in effective behaviors. Table 2 5 : MR of Individual Differences Predicting Ineffective Coded Behaviors DV R 2 F Predictor SE P Ineffective Coded Behaviors .23** 6.00 Social Dominance Orientation .28 ** .09 .00 Selfishness - .02 .16 .82 Openness to New Experience .01 .13 .82 Honesty - Humility - .04 .15 .61 Political Orientation .04 .06 .50 Empathy - .15 .19 .12 Perspective Taking .15 .15 .06 Moral Relativism - .07 .11 .21 Internalization - .08 .10 .27 Symbolization - .01 .06 .92 Willingness to Learn About Marginalized Groups - .13 .11 .08 118 Table 25. Racial Diversity of Friends - .04 .10 .52 Sexual Diversity of Friends - .11 .14 .14 Gender - Identity Diversity of Friends .10 .13 .17 Results from the multiple regression predicting ineffective coded behavior indicate that higher social dominance orientation positively predicted ineffective coded behavior. This finding suggests that, allies who are higher on social dominance orientation are more likely to engage in ineffective behaviors. Table 2 6 : MR of Individual Differences Predicting Self - Report Effective Behaviors DV R 2 F Predictor SE P Effective Self - Report Behaviors .34** 10. 52 Social Dominance Orientation - . 21 ** .25 < .01 Selfishness .0 9 .45 . 25 Openness to New Experience .06 .35 .3 0 Honesty - Humility .03 .41 .6 5 Political Orientation - .11 .16 .0 9 Empathy .00 .52 .99 Perspective Taking .01 .43 .85 119 Table 26. Moral Relativism - .02 .31 .66 Internalization .22 ** .28 < .0 1 Symbolization - .0 6 .16 . 30 Willingness to Learn About Marginalized Groups .2 0 ** .31 < .0 1 Racial Diversity of Friends .1 3* .29 .0 2 Sexual Diversity of Friends .1 1 .4 0 . 11 Gender - Identity Diversity of Friends - .1 7 * .36 .0 2 Results from the multiple regression predicting self - reported effective behavior indicate that higher levels of social dominance orientation and lower levels gender identity diversity in - reported effective ally behavior while internalized self - importance of moral identity and willingness to learn about marginalized groups positively predicted self - reported effective ally behavior. This finding suggests that, allies who are lower on social dominance orientation, have lower gender identity diversity in their friend groups, have a higher degree of internalized self - importance of moral identity and are willing to learn about marginalized groups are more likely to engage in ineffective behaviors. 120 Table 2 7 : MR of Individual Differences Predicting Self - Report Ineffective Behaviors DV R 2 F Predictor SE P Ineffective Self - Report Behaviors .14** 3.23 Social Dominance Orientation .1 8* .08 .0 3 Selfishness .0 6 .14 . 49 Openness to New Experience - .04 .11 .5 2 Honesty - Humility - . 08 .13 . 31 Political Orientation .1 0 .05 . 15 Empathy - . 10 .16 .3 4 Perspective Taking .1 4 .1 4 .1 0 Moral Relativism - .0 4 .10 . 50 Internalization - .0 1 .09 . 85 Symbolization .0 2 .05 . 7 8 Willingness to Learn About Marginalized Groups .0 3 .10 . 74 Racial Diversity of Friends - . 10 .09 . 12 Sexual Diversity of Friends .0 0 .13 . 9 8 Gender - Identity Diversity of Friends .1 5 .11 .0 6 Results from the multiple regression predicting self - reported ineffective behavior indicate that social dominance orientation positively predicted self - reported ineffective ally behavior. 121 This finding suggests that, allies who are higher on social dominanc e orientation are more likely to engage in ineffective behaviors. For the final set of analyses, I tested my exploratory research question (Do individual differen ces predict allyship behavior over and above ally motivations?) using a series of hierarchal r egression analyses. Four models were run in which prosocial motivations, external motivation to respond without prejudice and internal motivation to respond without prejudice were entered into the model in step one as predictors and empathy, perspective ta king, political orientation, moral relativism, moral identity, selfishness, social dominance orientation, honesty - humility, openness to new experience, willingness to learn about marginalized groups and close relationships with marginalized groups were ent ered into the model in step two as predictors. The behaviors (effective coded behaviors, ineffective coded behaviors, effective self - report behaviors, ineffective self - report behaviors) were entered into each hierarchal regression model as the dependent va riable. For results of all four models, see Tables 2 8 - 3 1 . 122 Table 2 8 : Hierarchal Regression Predicting Effective Coded Behaviors Step 1 Motivation Step 2 Add Individual Differences Prosocial Motivation - .01 (.13) - .07 (.15) External Motivation to Respond Without Prejudice - .06 (.04) .06 (.04) Internal Motivation to Respond Without Prejudice .42 ** (.07) .22 ** (.08) Social Dominance Orientation - .10 (.10) Selfishness .05 (.18) Openness to New Experience .10 (.14) Honesty - Humility .10 (.16) Political Orientation - .13 ** (.06) Empathy .02 (.20) Perspective Taking .03 (.16) Moral Relativism .06 (.12) Internalization .12 (.11) Symbolization - .07 (.07) Willingness to Learn About Marginalized Groups .04 (.12) Racial Diversity of Friends - .06 (.11) Sexual Diversity of Friends .02 (.16) Gender - Identity Diversity of Friends - .07 (.14) R 2 . 18** . 24 * * p of R 2 Change < . 01** <.01** Note: * p < 0.05 ; ** p < 0.0 1 . Values in parentheses are the standard errors of the regression estimates. The results of the hierarchal regression predicting effective coded allyship behaviors revealed that when considering motivations and individual differences together, internal motivation to respond without prej udice emerged as a positive predictor and political orientation (conservatism) emerged as a negative predictor of these ally behaviors. This suggests that allies 123 who engage in effective behavior tend to have higher levels of self - importance of moral identi ty and lower levels of conservatism. Table 2 9 : Hierarchal Regression Predicting Ineffective Coded Behaviors Step 1 Motivation Step 2 Add Individual Differences Prosocial Motivation - .11 (.12) - .03 (.13) External Motivation to Respond Without Prejudice .01 (.03) - .03 (.04) Internal Motivation to Respond Without Prejudice - .35 ** (.06) - .19 ** (.07) Social Dominance Orientation .19 ** (.10) Selfishness .00 (.16) Openness to New Experience .00 (.13) Honesty - Humility - .02 (.15) Political Orientation .04 (.06) Empathy - .13 (.19) Perspective Taking .15 (.15) Moral Relativism - .09 (.11) Internalization - .03 (.10) Symbolization .02 (.06) Willingness to Learn About Marginalized Groups - .08 (.11) Racial Diversity of Friends - .03 (.10) Sexual Diversity of Friends - .09 (.14) Gender - Identity Diversity of Friends .08 (.13) R 2 .18** .25** p of R 2 Change <.01** .03** Note: * p < 0.05 ; ** p < 0.0 1 . Values in parentheses are the standard errors of the regression estimates. The results of the hierarchal regression predicting ineffective coded allyship behaviors revealed that when considering motivations and individual differences together, internal motivation to respond without prejudice emerged as a neg ative predictor and social dominance orientation emerged as a positive predictor of these ally behaviors. This suggests that allies who 124 engage in in effective behavior tend to have lower levels of self - importance of moral identity and higher levels of socia l dominance orientation . Table 30 : Hierarchal Regression Predicting Self - Reported Effective Behaviors Step 1 Motivation Step 2 Add Individual Differences Prosocial Motivation .2 1 ** (.34) .1 3 (.38) External Motivation to Respond Without Prejudice - .0 7 (.10) .0 0 (.11) Internal Motivation to Respond Without Prejudice .3 1 ** (.18) .06 (.20) Social Dominance Orientation - .1 5 (.27) Selfishness .0 9 (.46) Openness to New Experience .0 7 (.36) Honesty - Humility .03 (.42) Political Orientation - .11 (.16) Empathy - .0 4 (.53) Perspective Taking .01 (.4 3 ) Moral Relativism - .01 (.32) Internalization .1 8 ** (.29) Symbolization - . 09 (.17) Willingness to Learn About Marginalized Groups .1 6 * (.32) Racial Diversity of Friends .1 3* (. 29 ) Sexual Diversity of Friends .1 0 (.41) Gender - Identity Diversity of Friends - .1 6 * (.36) R 2 .2 3 ** .35** p of R 2 Change <.01** <.01** Note: * p < 0.05 ; ** p < 0.0 1 . Values in parentheses are the standard errors of the regression estimates. The results of the hierarchal regression predicting effective self - reported allyship behaviors revealed that when considering motivations and individual differences together, internalized self - importance of moral identity, willing ness to learn about marginalized groups ositive predictors and gender - identity 125 emerged as a negative predictor of these ally behaviors. This suggests that allies who engage in effective behavior tend to have higher levels of self - importance of moral identity a re more willing to learn about marginalized groups, tend to have a more racially diverse friend group and tend to have a less gender identity diverse f riend group . Table 3 1 : Hierarchal Regression Predicting Self - Reported Ineffective Behaviors Step 1 Motivation Step 2 Add Individual Differences Prosocial Motivation - .03 (.1 0 ) - .0 5 (.12) External Motivation to Respond Without Prejudice . 08 (.03) - .0 2 (.04) Internal Motivation to Respond Without Prejudice - . 29 ** (.05) - . 1 8 * (.0 6 ) Social Dominance Orientation . 09 (.0 8 ) Selfishness .0 8 (.1 4 ) Openness to New Experience - .0 6 (.11) Honesty - Humility - . 0 7 (.1 3 ) Political Orientation .1 0 (.05) Empathy - . 07 (.1 7 ) Perspective Taking .1 4 (.1 4 ) Moral Relativism - .0 6 (.10) Internalization .0 4 (.09) Symbolization .0 5 (.0 5 ) Willingness to Learn About Marginalized Groups . 0 7 (.10) Racial Diversity of Friends - . 10 (.09) Sexual Diversity of Friends .0 2 (.13) Gender - Identity Diversity of Friends . 1 2 (.11) R 2 . 11 ** .1 7 ** p of R 2 Change <.01** . 11 Note: * p < 0.05 ; ** p < 0.0 1 . Values in parentheses are the standard errors of the regression estimates. 126 The results of the hierarchal regression predicting self - reported ineffective allyship behaviors revealed that when considering motivations and individual differences together, internal motivation to respond without prejudice emerged as a negative predictor of these ally behaviors. This suggests that allies who engage in ineffective behavior tend to have lower levels of internal motivation to respond without prejudice. Lastly, another exploratory analysis was conducted to determine the relative importance of each of the predictors used in this study in relation to each other. I used the method described in Tonidandel and LeBreton (2015) to conduct a relative weights analysis. Table 32 provides a summary of the relative raw and rescaled weights for each predictor by criterion. Through examination of the weights, it appears that internal motivation to respond without prejudice and social dominance orientation emerge as two of the most important predictors of both coded and self - reported of effective and ineffective allyship behaviors. Table 32 : Relative Weights Analyses of Predictors in Study 3 Effective Coded Behavior Ineffective Coded Behavior Self - Reported Effective Behavior Self - Reported Ineffective Coded Behavior Raw RS Raw RS Raw RS Raw RS Prosocial Motivation 0.01 2.84 0.02 8.10 0.03 9.99 0.01 5.61 External Motivation to Respond Without Prejudice 0.00 0.54 0.00 0.43 0.00 0.62 0.00 1.33 Internal Motivation to Respond Without Prejudice 0.05 19.46 0.05 20.82 0.04 11.65 0.03 20.02 127 Table 32. Social Dominance Orientation 0.04 12.86 0.05 20.51 0.04 12.44 0.02 13.62 Selfishness 0.01 2.70 0.01 3.90 0.01 2.03 0.02 9.88 Openness to New Experience 0.02 7.06 0.01 2.08 0.02 6.12 0.01 3.01 Honesty - Humility 0.01 5.18 0.01 2.05 0.00 1.21 0.01 6.19 Political Orientation 0.04 13.36 0.02 8.63 0.03 8.91 0.02 12.68 Empathy 0.02 5.72 0.02 8.94 0.02 5.14 0.01 6.46 Perspective Taking 0.01 4.81 0.01 2.40 0.01 4.31 0.00 2.38 Internalization 0.03 10.86 0.01 5.13 0.05 13.90 0.01 3.00 Symbolization 0.01 2.54 0.00 0.56 0.01 1.70 0.00 2.53 Willingness to Learn About Marginalized Groups 0.01 4.53 0.02 8.63 0.04 11.42 0.00 1.78 Moral Relativism 0.00 1.38 0.01 3.10 0.00 0.27 0.00 0.86 Racial Diversity of Friends 0.01 1.79 0.00 1.08 0.02 4.49 0.00 2.05 Sexual Diversity of Friends 0.00 0.65 0.01 2.23 0.01 2.01 0.00 1.45 Gender - Identity Diversity of Friends 0.01 3.71 0.00 1.40 0.01 3.79 0.01 7.15 128 STUDY 3 DISCUSSION The results of the third study partially support Hypotheses 1a and 1b. Internal motivation to respond without prejudice emerged as a significant motivational predictor ally behavior. Specifically, it positively predicted effective ally behaviors and negatively predicted ineffective ally behavior. These effects were found for self - reported and coded behaviors. This suggests that individuals who are motivated to behave in a non - biased manner for internal reasons may engage in a greater degree of effective allyship behaviors and may thus be better allies. While the majority of in dividual differences were non - significant in the present study, social dominance orientation emerged as a significant predictor of both effective and ineffective allyship behaviors. Specifically, social dominance orientation positively predicted engaging i n ineffective ally behaviors and negatively predicted engaging in effective ally behaviors. These effects were found for self - reported and coded behaviors. This suggests that individual who prefer hierarchy and have no qualms with the domination of lower s tatus groups tend to be less likely to engage in effective allyship behaviors. Internalized self - importance of moral identity also emerged as a significant predictor of self - reported and coded effective allyship behaviors. This suggests that individuals wh o view being moral as important to their identity tend to engage in a greater degree of effective allyship behaviors . Taken together, the results of this study suggest that majority group members who are motivated by internal factors and equality - related b elief systems tend to be more likely to engage in effective ally behaviors and thus be better allies. 129 GENERAL DISCUSSION This series of studies examined the effects of ineffective allyship behaviors on members of marginalized groups and examined the mo tivations and individual differences that are related to effective and ineffective allyship behaviors. Study 1 examined the effects of effective and and voi Ineffective allies were found to be inversely related to positive psychological outcomes for marginalized group members and they were also found to be evaluated less posit ively than effective allies. Using the critical incidents from Study 1, items representing frequently occurring ally behaviors were examined in Study 2 . E xploratory factor analyses revealed that ally behaviors do not exist on a continuum but are factoriall y distinct. Study 3 examined an array of predictors that were theoretically thought to be related to effective and ineffective allyship behaviors. Although few predictors were significant when all other variables were accounted for in the model, social dom inance orientation and internal motivation to appear non - prejudiced emerged as robust predictors of allyship behavior. Through the set of studies conducted, definitions of effective and ineffective allyship emerged empirically. Effective allyship behavior consists of culturally competent actions that directly address the biased behaviors of others in a manner that considers the needs and desires of the person one attempt s to help. Conversely ineffective allyship consists of actions that are not culturally competent, do not address the biased behaviors of others or fail to consider the needs or desires of the person they are trying to help . When effective and ineffective behaviors are conceptualized at the individual level it is difficult if not impossible t o extract behavior from the outcome as is possible when looking at conceptualizations of job performance compared to 130 effectiveness ( Campbell, McCloy, Oppler, & Sager, 1993 ). This is particularly true when asking the beneficiary of the ally behavior to rate their ally. Leadership effectiveness ratings from direct reports, particularly those ratings for actions considered relations - oriented behaviors (e.g., supporting, developing, recognizing, empowering), may be similarly difficult to disentangle from outcom es at an interpersonal level (Yukl, 2012). However, it is possible that when allies and leaders strive for outcomes beyond the interpersonal level, the distinction between behaviors and outcomes becomes distinguishable similar to conceptualizations of job performance (Campbell et al., 1993). Future research is needed to help disentangle ally behavior from outcomes at various levels and to examine whether this feat is possible at the individual level. The present set of studies focuses on ally at the interpersonal, individual level. In the next section, I will discuss the se findings in depth. Effects Allyship Behavior on Marginalized Group Members The resu lts of Study 1 support hypotheses 3 a and 3 b . Allyship behaviors were significantly related to the majority of psychological outcomes (e.g., psychological safety, anxiety, positive affect and negative affect). This suggests that the type of behaviors an all y engages in has the potential to impact marginalized group members. Ineffective ally behaviors are not neutral and can result in negative ramifications for members of marginalized groups Although no known research has considered the effects of ineffective allyship behaviors, these findings align with the assertions made in the popular press (Ariel, 2017; Charles, 2016; Muniz, 2016; Zivad; 2017 M orrison , 201 3 ; Patton , 201 7 ; Threads of Solidarity, 201 7) . These findings suggest that merely being an ally is not enough for marginalized group members to feel included. When individuals take on the ally role, it is important that they engage in effective behaviors and refrain from engaging in ineffective behaviors. In addition , the results of Study 1 also suggest 131 th at marginalized group members evaluate allies who engage in ineffective behaviors more negatively than those who engage in effective ally behaviors. Although this finding may feel intuitive, it has the potential to have long lasting effects. When marginali zed group members experience ineffective ally behaviors it results in negative psychological states. Because of this, when people experience these ineffective behaviors, they may be motivated to keep themselves from being affected by these behaviors in the future which may lead to overall distrust in majority group members who claim to be allies. Future research would do well to examine the longitudinal effects of experiences with allies on the views of future allyship behavior for members of marginalized g roups. Of all relationship categories, allies in this sample were mostly reported as being close friends (26.6 %) or strangers (21.6%) . Post - hoc evaluation of count data revealed that close friends and strangers were both found to be effective and ineffect ive allies. This suggests that there is a range of relationships between allies and those they ally with. Future research could further consider the dyadic relationships between allies and marginalized group members to determine whether this influences the degree to which allies are effective. Characteristics Related to Effective and Ineffective Ally Behaviors The results of study 2 and 3 partially support H ypotheses 1a, 1b and 2 . Although there were there were many characteristics and motivations exami ned that were theoretically related, few emerged as significant predictors. Some of this may have occurred due to some of the predictors capturing overlapping construct space. For example, empathic concern and perspective taking ( r = .71) may have cancelled each other out. However, there was also a great deal of distinction between constructs with constructs like openness to new experience ( = .19 - .35) , moral relativism ( = .01 - .18) and symbolized self - importance of moral id entity ( = .01 - .33) exhibiting relatively low correlations with the other predictors . Because highly correlated 132 variables and variables with lower correlations with other predictors yielded similarly non - significant results, it may suggest that there a re other motivations and individual differences that are more predictive of ally behavior. It can be helpful to consider the overall predictive power of the total models as well . The R 2 values of models containing all predictors and predicting coded behavi ors ranged from .24 to .25 suggesting that these models predicted a sizable degree of variance within the realm of social science research (Cohen, 1988). R 2 values of models containing all predictors and predicting self - reported behaviors ranged from .17 t o .35 suggesting that these models predicted a moderate to sizable degree of variance (Cohen, 1988). Thus, although f ew of the predictors of ally behaviors were significant the total models yielded rather sizable predictive power. In addition, the predicto rs that were significant remained so across all analyses suggesting that they are robust predictors of ally behavior . However, because 100% of the variance was not explained, it is clear that there are additional predictors of allyship behavior that were n ot considered for the present study. Future research could consider additional predictors of allyship behavior outside of the individual difference domain. Potential directions include considering p revious participation in organizations that support the ma rginalized group members , participation in structured dialogues on privilege and demographic factors and participation in other programs that facilitate ally development. It would also be interesting to determine whether volitional and involuntary particip ation in these programs result in greater degrees of effective and ineffective ally behaviors. When majority group members had a higher degree of motivation to appear non - prejudiced for internal reasons and were lower in social dominance orientation, they were more likely to engage in effective allyship behaviors and were less likely to engage in ineffective allyship behaviors. Both of these characteristics have been historically linked to empirical 133 research in the realm of prejudice and i ngroup - outgroup relations (Butz & Plant, 2009; Ho et al., 2012) . In addition, internalized self - importance of moral identity was also found to be a significant, positive predictor of engaging in effective ally behaviors. Taken together, these findings sugg est that majority group members who do not support or endorse hierarchies and dominating lower - status groups and are intrinsically motivated to treat marginalized group members well tend to be better allies. Unfortunately, not all majority group members ha ve these characteristics. Because of this, future research on allyship training may help discover ways in which interested majority group members can learn to become better allies. Research in the area of diversity training in organizations exhibits an oft en contradictory set of results ( see Bezrukova, Jehn & Spell, 2012 for review). Some of this may be explained by the fact that employees do not come to the training as tabula e rasa e , but they come in with their own characteristics, values and beliefs. As d iscovered in this study, individuals with certain characteristics are more likely to engage in effective allyship behaviors but not all majority group members share these characteristics. A recent meta - analysis ( Kalinoski et al., 2013) revealed that effect s of diversity training were stronger for skill based and cognitive outcomes as opposed to attitudinal outcomes . In the context of allyship training, it could be helpful to develop training programs based on the stronger findings of diversity training more broadly. Specifically, focusing on training workers to engage in skills and to encourage them to practice these skills in a non - evaluative setting during more than one timepoint ( Bezrukova, Spell, Perry, & Jehn, 2016) may result in stronger cognitive and skill - based outcomes. Although there is a possibility that change can occur through attempting to increase levels of relevant traits or altering the motivations of individuals in the majority group , previous research on diversity training suggests that the se effects are small and do not tend to be robust across studies ( Kalinoski et al., 2013) . 134 Findings from Preliminary Scale Development In addition to the other empirical findings, the scale development process also provided insights into the nature of ally behaviors. First, at the outset of this investigation it was unknown whether allyship behaviors existed on a linear spectrum from effe ctive to ineffective or if these behaviors existed as separate constructs. The present study provided strong evidence that effective and ineffective allyship behaviors are separate constructs. Indeed, t he two - factor structure of the ABQ and the low, negati ve correlation between factors suggests this. As a result of examining the critical incidents in Study 1, I also decided to separate out interactions between peopl e, characteristics within people and interactions between people and the environment. Even when an ally engages in effective behaviors it is possible that they do so in an environment that is not receptive. This can include standing up for marginalized group members when other individuals are not willing to change their views or behaviors. Thus, although allies may engage in effective and ineffective behaviors, the degree to which these behaviors result in positive change in the larger envir onment is not yet known. Future research could help examine the person - environment interactions that result in the greatest degree of positive change for members of marginalized groups beyond the psychological ramifications considered in the present study. The ABQ may also be different from current measures of allyship behaviors. Although the ABQ and the PACS measures were correlated with each other in the predicted directions, there was not complete overlap between the measures. This suggests that the AB Q may be capable of explaining additional variance over and above PACS. Future research is needed to 135 determine whether the ABQ predicts outcomes over and above the PACS and the magnitude of these predictions. Although the exploratory factor structure has been evaluated and a confirmatory factor analysis has been evaluated on the same data , it is highly advised that additional psychometric and validation wor k be c onsidered before wide - ranging use of the ABQ . Specifically, future work would do well to confi rm the factor structure of the ABQ using confirmatory factor analyses and examining the fit indices in a unique sample . In addition, determining whether there are structural differences in the measure depending on the sample type (e.g., marginalized group members vs . majority group members) may determine whether it is better suited for use as a self - report for allies or as an other - report measure for marginalized group members to complete. Limitations While the present work has revealed more about the perc eptions of ally behaviors and the individual differences that may be related to allies who engage in effective behaviors versus allies who engage in ineffective behaviors, this study is not without limitations. Specifically, the present study uses cross - se ctional data, combined individuals from various marginalized groups despite the fact that individuals from these various groups go through different experiences , one scenario from the criterion in the third study had to be dropped for conceptual and statis tical reasons and participants were asked to report on an event up to a year after it occurred . The use of cross - sectional data limits my ability to infer causal connections from any of the results of the present stud y. Although the results of present stu dy are an informative foray into empirical research on allyship , I cannot be certain that predictors and relationships are causal . In addition, it is arguable that self - reported data may have certain biases and this study 136 relied almost exclusively on self - report. For the third, study, majority group members were asked to self - report on their behaviors and characteristics. The fact that a greater number of predictors were related to self - reported effective behavior signals that there may be social desirabili ty bias influencing some relationship s . Similarly, because self - reported ineffective behaviors were less frequently reported, there was less variance in the criterion space to be predicted. This again speaks to a potential social desirability bias. In a s imilar vein, discussion of how to classify marginalize d group members reports of allyship behavior is warranted. S ome may argue that reports of ineffective and effective ally behaviors from marginalized group members should be classified as perceptions of ally behavior rather than indices of the behavior in and of themselves . I argue against this assertion. Individuals in marginalized groups are more likely to have experienced biased treatment or discrimination. Because of this, they may be more cognizant o f which behaviors would be helpful in such situations and which ones would make the situation worse. In addition, drawing from research on personality more broadly, other reports are thought to exhibit unique predictive power above self - ratings and tend to be more reliable (Balsis, Cooper, & Oltmanns, 2015; Luan et al., 2019). Thus, for the present study, marginalized group members served as subject matter experts on ally behaviors and I maintain that their experiences should be considered as such. Future research should continue to include the perspective of marginalized group members as their unique perspective is under - considered in this nascent research area. Important nuances may have been missed by aggregating across multiple marginalized gro up. Although racial, ethnic , sexual and gender non - conforming marginalized group members experience stigma, discrimination and biased treatmen t, the stereotypes and treatment of each group are different. For example, in the racial/ethnic minority sample in Study 1, Asian 137 participants would mention being othered by ineffective allies in the form of culturally insensitive gifts and racial appropriation (e.g., a majority group member claiming to be part - Asian). Black participants mentioned being followed aroun d shopping centers and being accused of theft or having to wait inordinate amounts time to be served in establishments. Hispanic and Latino participants mentioned stereotypic forms of cultural appropriation and being called by the names of celebrities from their ethnic group in an unwanted manner. However, individuals from all groups mentioned being called slurs or experiencing mistreatment and harassment more broadly. Similarly, individuals in the LGBTQ + sample experienced different treatment. Lesbian indi viduals mentioned being sexualized by men and being harassed by them. Gay individuals mentioned being chastised for their clothing and being humiliated in professional setting s . Individuals in each of these identity groups mentioned being threatened with slurs, outed in an unwanted manner, and in many cases, experienced physical attacks. However, during the first two studies, differences between the samples were neglig ible in terms of most outcomes with the exception of negative affect in Study 1. This suggests that although individuals from marginalized groups experience different forms of discrimination and biased treatment, it results in negative psychological impact s regardless of the form. Although it would be insightful to delve into the nuances of each marginalized group using data from the present study, the sample sizes are not large enough to examine the data by individual identity group. Thus, t he present stud y examined universal questions including whether or not ineffective ally behaviors exist and what the effects of these behaviors were on marginalized group members broadly . Now that it is know n that these behaviors d o in fact exist and that ineffective beh aviors result in negative 138 psychological outcomes for marginalized group members , future research can work to determine the specific behaviors that are ineffective to specific marginalized groups. Third, the scenario in the third study needed to be dropped due to low interrater agreement and questionable validity. The ICC for the third scenario was quite low after two rounds of coding and feedback. Although my decision to drop the scenario was primarily due to the low agreement, this scenario was also the o nly one that potentially required participants to advocate for someone to higher levels of the organization being that the target was passed over for a promotion. In addition, helping the target navigate a biased promotion system inherently had more steps in the process and would take more time to complete. Conversely all other scenarios occurred at the interpersonal level during a single event in time. Although the present study yielded informative results using the available scenarios, it is arguable that the scenario would have increased the strength of this measure. In addition, the scenarios that were included in the criterion represented a limited context. Future research could consider broadening the context of the scenarios to include a variety of wo rkplace occurrences or manipulate the - worker, client) to determine in the workplace . Lastly, pa rticipants were asked to recount discriminatory or biased events up to 12 months after they occurred. Time between the event and reporting psychological reactions to the event may have resulted in participants need ing to gues s timate their responses to thei r reported events. However, it is also arguable that having to recount such an event re - triggered the emotions and feelings from experiencing the event. Indeed, having people recount events or write about emotionally valanced topics is a technique that is often used as a manipulation in 139 psychological research ( Strack, Schwarz & Gschneidinger , 1985 ; Lerner & Keltner, 2001 ). If allyship closer to when it occurred and to determine the frequency of ally intervention when discrimination is experienced by marginalized group members. F uture research could potentially examine the occurrence of ally intervention studies or experience sampling method ( ESM ) to determine the base - rate of these events for individuals from marginalized groups . This would allow for a n increased understanding of the psychological effects of discrimination over time and reveal whether allies serve as a buffer in these circumstances . Previous research using daily diary studies linked discrimination to increases in negative psychological outcomes directly after experiencing a discriminatory event ( Broudy et al., 2007; Torres & Ong, 2010) . Just as effective ally behavior was linked to increased psychological outcomes in the present study, future research could determ ine whether effective ally behavior buffers against the negative impacts of discrimination and bias or perhaps even results in positive psychological outcomes in the moment. In addition, the present study revealed that ineffective ally behavior was linked to decreased psychological outcomes and future research in this area would help determine whether experiencing ineffective ally behavior during an event of discrimination results makes matters worse in the moment. Future Research There are many opportuniti es for future research on the topic of allyship behaviors more broadly. First, while there is a growing body of work involving allyship in organizational psychology , the topic area is nascent and there is a great deal left unknown. During the most recent c onvention of the Society of Industrial and Organizational Psychology, only two session s featured work on allies ( Demsky, Rineer, Crain & Ellis , 2019 ; Liu, Dray & Sabat , 2019) 140 Although research within the subfield is off to a strong start, there are numerous additional questions that may allow us to further understand allyship and help majority group members to become better allies to members of marginalized groups. Although man y suggestions for future research have been discussed in previous sections of the discussion, I will briefly outline a few additional avenues including : ally feedback , considering looking at allyship at multiple levels of an organization , larger environmen tal factors that may influence allyship in organizations , the generality of the items in the current ABQ measure , examining status differentials between marginalized group members and allies and the use of socially desirable responding measures when engagi ng in research on this topic First, it is not entirely clear how well allies are able to calibrate their behaviors when provided with feedback from others. It would be particularly interesting to learn more about whether the person providing the feedback ( e.g., similar other, marginalized group member) or the type of feedback (e.g., constructive criticism, positive comments or neutral comments) influences the degree to which majority group members feel willing to engage in future allyship behaviors. It is a mindset (fixed and growth; Dweck, 1986) may play a role in terms of their willingness to accept and act on feedback. Allies who are more growth mindset oriented may be more willing to accept feedback and change their behaviors when they are deemed to be ineffective. Allies who are more fixed mindset oriented may feel discouraged when given feedback that is less than positive and may shy away from attempting allyship behaviors in the future. Second, it may be good to examine the complexities of allyship spanning multiple levels of the organization. The present investigation was limited in its scope in that I only considered allyship at the individual, interpersonal level. Allies and advocates can work to change larger 141 systems and structures within the organization so that marginalized group members not only feel included but are also treated equitably. This may include persistently supporting fair pay for marginalized group members, supporting the promotion of marginalized group m embers and advocating for the investment of organizational inclusion efforts like employee resource groups (Welbourne, Rolf, & Schlachter, 2017). Because it is likely that individuals in decision making positions within organizations will have similar demo graphic characteristics to majority group members, allies could be more persuasive due to their similarity to the decision maker and perceived lack of self - interest (Gardner, 2018). Because it is not entirely clear that the findings from this study would t ransfer beyond levels, future work could examine effective and ineffective ally behaviors and characteristics that are relevant to increasing inclusion and fair treatment beyond the individual level to the organizational level . Third , future research may wish to explore the environmental factors such as organizational approach to diversity and climate for diversity when considering allyship behavior in the workplace. Colorblind and multicultural approaches to diversity have been long considered in organiz ational psychology with the former being related to negative outcomes in racism and discrimination (Offermann et al., 2014; Plaut, Thomas, Hurd, & Romano, 2018). Or ganizations that have a colorblind approach to diversity may thus have lower awareness of the biases faced by employees from marginalized groups. According to the CPR model ( Ashburn - Nardo, Morris & Goodwin , 2008) for allies to act on an event it must be de emed discriminatory and deemed to be an emergency. Because of this, it is possible that organizations with a colorblind approach to diversity will have fewer majority group members serving as allies. 142 s approach to diversity helps or hinders majority group members who wish to serve as allies on behalf of marginalized group members. Diversity climate may also impact the degree to which individuals can a c t as allies in the organization. Climate has been theorized to impact individual level factors for members of marginalized groups that then effect their career outcomes which then, in turn ultimately impact organizational performance (Cox, 1993; Cox & Beele, 1997). In climates such as these, majority grou p members may be best suited to advocating for organizational changes on behalf of marginalized groups. However, they may feel penalized and potentially even be ostracized themselves by the organization for behaving in a way that is consistent with the cli mate. Future research would do well to examine the allyship in workplaces with a poor diversity climate to gain a better understanding of allies who are willing to help while putting themselves at increased risk or cost compared to those in healthy diversi ty climates. Another future direction may entail considering whether the current items in the ABQ reflect allyship unique to marginalized group s or whether these items could apply to individuals from all groups. Although several of the ABQ items contain content that is relevant to identity, Told me I don't act like a normal member of my group, Made my circumstanc es worse by drawing attention to my minority status, Defended me against biased statements in an content is more broad and may be applicable to any identity group . Si milarly, the PACS affirmation subscale items may also be more broad and thus applicable to anyone. It is therefore possible that item content may reflect effective and ineffective behaviors that a person may engage in on beha lf of anyone experiencing workp lace incivility regardless of marginalized 143 status. Because of societal norms, more subtle forms of discrimination and biases are more common and are harder to spot . Thus, the more subtle forms may occur through selective - incivility (Cortina, 2008; Kabat - Fa rr & Cortina, 2012) . Although it is highly likely that the majority of people would benefit from effective allyship when experiencing incivility in the workplace, marginalized group members may experience incivility at a greater frequency. The more general items from the ABQ may be useful for examining the degree to which both majority group members and marginalized group members benefit from another person intervening during occurrences of workplace incivility. Future research comparing the effects of ally ship behavior as measured by the ABQ for majority and marginalized groups will help determine whether the ABQ is a measure that examines behaviors that are uniquely helpful to individuals from certain groups or whether it is a measure of behaviors that hel pful to all people experiencing incivility more broadly. In addition, considering the implementation of social desirability measures when conducting research on allyship may be valuable. The degree to which a person is able to accurately self - report on th eir behaviors may be affected by social desirability bias (Paulhus, 1991). When asking people to report on behaviors that are considered to be particularly good or bad, it is possible that they may alter their responses in order to appear a certain way to others states or behaviors (e.g., self - deception). In the present study, I attempted to circumvent socially desirable responding of allyship behaviors by aski ng participants to qualitatively report their behaviors in the scenario and then have trained, independent coders rate their behaviors. However, there are also a variety of methods researchers can use to evaluate the degree to which participants are engagi ng in socially desirable responding without having to engage in taxing 144 qualitative coding ( see Fleming, 2012 and Nederhof, 1985 for review ) . For example, the Balanced Inventory of Desirable Responding (BIDR) is a highly used measure of social desirability (Li & Bagger, 2007; Paulhus, 1988) and assesses both impression management and self - deception. Because of the nature of allyship behaviors, future research on this topic may benefit from including measures of social desirability. Lastly, considering the ro le of status differentials in allyship could be an interesting future direction. In the first study, there were only three participants who were not relatively equal in status with their ally (e.g., the participant had a lower status) so I was unable to ex amine this in the present study. However, it is highly possible that the degree to which one feels comfortable providing their ally with feedback (e.g., voice) may be dependent on the degree to which one is of a similar status to their ally. In the first s tudy, the majority of allies described in both the ineffective and effective ally conditions were close friends and strangers. The lack of status differential may have factored in to the non - significance of voice outcome in the first study. Future research could examine the role of status differentials between marginalized group members and allies in terms of voice and other outcomes including persuasion of similar others and advocating for change at higher levels of the organization. 145 CONCLUSION The present investigation examined the effects of effective and ineffective allyship behaviors. Through this work, the negative psychological ramifications of ineffective ally behaviors were elucidated. In addition, a pattern of individual differences and motivations related to effective and ineffective allyship behaviors emerged following a broad examination of theoretically related characteristics. The study of allyship is nascent and although the present series of studies are i nformative, they barely scratch the surface. Future empirical research on the development of allies, examining additional characteristics of allies and examining effective ways to provid e allies with feedback is warranted. 146 APPENDICES 147 APPENDIX A : Ally Manipulations and Questions for Study 1 Ineffective Ally: We are interested in learning about people who consider themselves to be allies from the perspective of people from marginalized groups. Please use the space be low to write about a time during which a member of majority group advocated on your behalf when you were experiencing bigotry, unfairness or inequality as a result of being a member of a marginalized group (e.g., racial/ethnic minority, LGBTQ+). Please wri te about a time when the ally attempted to advocate or intervene but did so ineffectively. That is, please describe a time when someone considered themselves an ally but actually was not very helpful, or was even harmful from your perspective. What are so me of the specific things your ally did that you found to be not helpful or even hurtful? That is, what did he/she/they do or say? Exploratory Question: Why do you think your ally intervened on your behalf? That is, what do you think motivated them to help ? Effective Ally: We are interested in learning about people who consider themselves to be allies from the perspective of people from marginalized groups. Please use the space below to write about a time during which a member of majority group advocated o n your behalf when you were experiencing discrimination, bigotry, unfairness or inequality as a result of being a member of a marginalized group (e.g., racial/ethnic minority, LGBTQ+). Please write about a time when the ally attempted to advocate or interv ene and did so effectively. That is, please describe a time when someone considered themselves as an ally and their actions were helpful to you, either in changing the outcome of a situation or in making you feel supported. What are some of the specific t hings your ally did that you found helpful? That is, what did he/she/they do or say? Exploratory Question: Why do you think your ally intervened on your behalf? That is, what do you think motivated them to help? 148 APPENDIX B : Perception of Ally Characteristics Scale (PACS) Using a 1 (not at all characteristic) to 5 (very characteristic) scale, please indicate the degree to which the statements below describe your ally during the situation you previously described. 1. My ally proposes possible actio ns to address potentially racist {homophobic} situations affecting me. 2. My ally acknowledges differences between us. 3. My ally understands his or her own racial/ethnic {sexual and gender} identity. 4. My ally is knowledgeable about racial/ethnic communities {sex ual orientations and gender identities} other than his or her own. 5. My ally is active in racial/ethnic communities {sexual and gender communities (e.g., LGBTQ+ communities)} other than his or her own. 6. My ally takes action to address bias among his or her ow n racial/ethnic group {heterosexual or cisgendered people}. 7. My ally creates a feeling of connection with me. 8. My ally is respectful towards me 9. My ally is interested in what happens to me. 10. My ally is nonjudgmental toward me. Note: words in brackets refer to the substitutio ns used for the sexual minority version of the scale 149 APPENDIX C : Trust Scale Using the scale , please indicate how much you agree or disagree with the following statements: 1. Most people are basically honest. 2. Most people are trustworthy. 3. Most people are basically good and kind. 4. Most people are trustful of others. 5. I am trustful. 6. Most people will respond in kind when they are trusted by others. 150 APPENDIX D : Positive and Negative Affect Schedule (PANAS) Instructions: Using the answer choices provided, please indicate the extent to which you felt each emotion as a result of your ally's behaviors. 1 = Very Slightly or Not at All 2 = A Little 3 = Mod erately 4 = Quite a Bit 5 = Extremely 1. Interested 2. Distressed 3. Excited 4. Upset 5. Strong 6. Guilty 7. Scared 8. Hostile 9. Enthusiastic 10. Proud 11. Irritable 12. Alert 13. Ashamed 14. Inspired 15. Nervous 16. Determined 17. Attentive 18. Jittery 19. Active 20. Afraid 151 APPENDIX E : Voice Questions Instructions: The following questions are about how comfortable you would feel voicing your concerns to the ally you described in the previous section. Please respond to each statement honestly using the answer choices below (1 = Strongly Disagree, 5 = Str ongly Agree). I would feel comfortable... 1. 2. Proactively voicing out constructive suggestions that help my ally improve their behaviors. 3. behaviors. 4. Advising my ally against undesirable behaviors that would hamper their helpfulness. 5. Speaking up honestly when my ally engages in problematic behavior, even when/though my ally may disagree with me. 6. aviors, even if that would hamper the relationship with that ally. 152 APPENDIX F : Stigma Consciousness Instructions: Use the numbers below to indicate how much you agree or disagree with each statement ( 1 = Strongly Disagree, 5 = Strongly Agree ). 1. Stereotypes about my racial group {sexual orientation} [gender identity] have not affected me personally. 2. I never worry that my behaviors will be viewed as stereotypical of my racial minority {sexual minority group} [gender identity group ] . 3. When interacting with White {heterosexual} [cisgendered] people, I feel like they interpret all my behaviors in terms of the fact that I am a racial minority {sexual minority group member} [gender minority group member] . 4. Most White {heterosexual} [cisge ndered ] people do not judge minorities on the basis of their race {sexual orientation} [gender identity] . 5. My being a racial minority {sexual minority group member } [gender minority group member] does not influence how White {heterosexual} [cisgendered] peop le act with me. 6. I almost never think about the fact that I am a racial minority {sexual minority group member} [gender minority group member] when I interact with White {heterosexual} [cisgendered] people. 7. My being a racial minority {sexual minority group member} [gender minority group member] does not influence how people act with me. 8. Most White {heterosexual} [cisgendered] people have a lot more racist {homophobic} [transphobic] thoughts than they actually express. 9. I often think that White {heterosexual} [cisgendered] people are unfairly accused of being racists {homophobic} [transphobic] . 10. Most White {heterosexual} [cisgendered] people have a problem viewing people from minority groups {sexual minority groups} [gende r minority group members] as equals. Note: words in brac e s refer to the substitutions used for the sexual minority version of the scale and words in brackets refer to the substitutions used for the gender minority version of the scale 153 APPENDIX G : Id entity Centrality Instructions: Use the numbers below to indicate how much you agree or disagree with each statement ( 1 = Strongly Disagree, 5 = Strongly Agree ). 1. I have spent time trying to find out more about my ethnic group {sexual orientation } [gender identity group] , such as its history, traditions, and customs. 2. I am active in organizations or social groups that include mostly members of my own ethnic group {sexual orientation} [gender identity group] . 3. I have a clear sense of my ethnic background {sex uality} [gender identity] and what it means for me. 4. I think a lot about how my life will be affected by my ethnic group membership {sexual orientation} [gender identity group membership] . 5. I am happy that I am a member of the group I belong to. 6. I have a str ong sense of belonging to my own ethnic group {sexual orientation} [gender identity group] . 7. I understand pretty well what my ethnic group membership {sexual orientation} [gender identity group membership] means to me. 8. In order to learn more about my ethnic background {sexuality} [gender identity] , I have often talked to other people about my ethnic group {sexual orientation} [gender identity] . 9. I have a lot of pride in my ethnic group {sexual orientation} . 10. I participate in cultural practices of my own group, such as special food, music, or customs. 11. I feel a strong attachment towards my own ethnic group {sexual orientation} [gender identity group] . 12. I feel good about my cultural or ethnic background {sexuality} [gender identity] . Note: words in braces refer to the substitutions used for the sexual minority version of the scale and words in brackets refer to the substitutions used for the gender minority version of the scale 154 APPENDIX H : Perceived Discrimination Instructions: In your day - to - day l ife, how often do any of the following things happen to you (1=Never, 2 = Less than Once a Year, 3 = A Few Times a Year, 4 = A Few Times a Month, 5 = A Few Times a Week, 6 = Almost Every day)? 1. You are treated with less courtesy than other people are. 2. You are treated with less respect than other people are. 3. You receive poorer service than other people at restaurants or stores. 4. People act as if they think you are not smart. 5. People act as if they are afraid of you. 6. People act as if they think you are dishones t. 7. 8. You are called names or insulted. 9. You are threatened or harassed. 155 APPENDIX I : Demographic Questions 1. How old are you? 2. How would you describe your gender identity? a. Male b. Female c. Non - Binary d. Transgender e. Agender f. Prefer not to answer 3. How would you describe your race? Select ALL that apply. a. American Indian b. Asian c. Black or African American d. Native Hawaiian or Pacific Islander e. White (including all European heritage) f. Other 4. How would you describe your ethnicity? a. Hispanic b. Latino c. Neither Hispanic not Latino 5. How would you describe your sexual orientation? a. Heterosexual b. Homosexual c. Bisexual d. Pansexual e. Asexual f. Other 6. How long have you worked at your current job (in years)? 7. How long have you worked in your current industry (in years)? 156 APPENDIX J : Consent Forms for All Studies Research Participant Information and Consent Form Study 1 You are being asked to participate in a research study. Researchers are required to provide a consent form to inform you about the research study, to convey that participation is voluntary, to explain risks and benefits of participation, and to empower you to make an informed decision. You should feel free to ask the researchers any questions you m ay have. Study Title: Allyship Effectiveness Study 1 1. PURPOSE OF RESEARCH The purpose of this research study is to develop a measure that will reveal how individuals think about allyship more broadly. 2. WHAT YOU WILL DO Consenting participants will be asked to report their previous experience with allies. Participants will be asked to describe the experience quali tatively and report their perceptions of the ally. Questions about these events and several personality - related questions will also be asked. 3. POTENTIAL BENEFITS You may not benefit personally from being in this study. However, we hope tha t this research may eventually benefit others hoping to learn more about how allyship is conceptualized. 4. POTENTIAL RISKS The researchers do not expect that there are any potential risks to completing this study. 5. PRIVACY AND CONFIDENTIALITY All data will be stored on the hard drive of a secure computer and will only be accessed by trained experimenters. Data will be stored for five years after the publication of resear ch stemming from this project --- as specified by the American Psychological Association. 6. YOUR RIGHTS TO PARTICIPATE, SAY NO, OR WITHDRAW You have the right to say no to participate in the research. You can stop at any time after it has already sta rted. There will be no consequences if you stop and you will not be criticized. You will not lose any benefits that you normally receive. However, participants will only be paid in proportion to the percentage of the survey they complete (see below). It i s important to provide quality answers to the open ended/fill in the blank questions. Failure to provide quality answers to these questions will result in your disqualification from the study. Participants who complete the study but do not meet the qualifi cations specified in the hit will not be compensated. 157 7. COSTS AND COMPENSATION FOR BEING IN THE STUDY Qualifying participants will receive $1.00 for participating in this study. The study is projected to take no more than 30 minutes. Please be sure t o check the original hit page to make sure you qualify. Participants who complete the study but do not meet the qualifications specified in the hit will not be compensated . Providing poor quality responses to the open ended/fill in the blank questions will disqualify you from participation and you will not be compensated. 8. CONTACT INFORMATION If you have concerns or questions about this study, such as scientific issues, how to do any part of it, or to report an injury, please contact Lauren Collier at colli719@msu.edu OR Ann Marie Ryan Ph. D., Department of Psychology, Michigan State Universi ty, East Lansing, MI 48824, phone: 517 - 355 - 0203, e - mail: ryanan@msu.edu. If you have questions or concerns about your role and rights as a research participant, would like to obtain information or offer input, or would like to register a complaint about th is study, you Protection Program at 517 - 355 - 2180, Fax 517 - 432 - 4503, or e - mail irb@msu.edu or regular mail at 4000 Collins Rd, Suite 136, Lansing, MI 48910. 9. DOCUMENTA TION OF INFORMED CONSENT If you would like a copy of the consent form, please email Lauren Collier (colli719@msu.edu). 158 Research Participant Information and Consent Form Study 2 You are being asked to participate in a research study. Researchers are required to provide a consent form to inform you about the research study, to convey that participation is voluntary, to explain risks and benefits of participation, and to empower you to make an informed decision. You should feel free to ask the researchers any questions you may have. Study Title: Allyship Effectiveness Study 2 1. PURPOSE OF RESEARCH The purpose of this research study is to develop a measure that will reveal how individuals think about allyship more broadly. 2. WHAT YOU WILL DO Cons enting participants will be asked to detail a previous and recent experience with allyship and respond to several items regarding their beliefs and views about allyship. 3. POTENTIAL BENEFITS You may not benefit personally from being in this study. However, we hope that this research may eventually benefit others hoping to learn more about how allyship is conceptualized. 4. POTENTIAL RISKS The researchers do not expect that there are any potential risks to completi ng this study. 5. PRIVACY AND CONFIDENTIALITY This study is confidential. Your answers will only be associated with an anonymous ID. To help us protect your confidentiality, please do not write or give your name or any other identifying information during the study. All data will be stored on the hard drive of a secure computer and will only be accessed by trained experimenters. Data will be stored for five years after the publication of research stemming fro m this project --- as specified by the American Psychological Association. 6. YOUR RIGHTS TO PARTICIPATE, SAY NO, OR WITHDRAW You have the right to say no to participate in the research. You can stop at any time after it has already started. There will be no consequences if you stop and you will not be criticized. You will not lose any benefits that you normally receive. 7. COSTS AND COMPENSATION FOR BEING IN THE STUDY Participants will receive $1.50 for participating in this brief study. The study i s projected to take no more than 20 minutes. Compensation Rules (PLEASE READ) The following are reasons why we would not be able to compensate you for your participation . By following these compensation rules, we hope to be as fair as possible to 159 surve y respondents who meet the study criteria, who access the survey only once, and who provide quality data for our study. Please note: If you do not include your MTurk ID in the online survey we cannot identify you and so you will not be compensated if you fail to correctly enter your Mturk ID in the online survey. If we have no record of your Mturk ID in our data, we cannot compensate you. If you are not eligible to take this research surv ey based on the prescreening questions , we cannot compensate you for your participation. The quality of our scientific study depends on participants meeting these criteria. If we find that you have re - entered the survey multiple times after initially faili ng the prescreening questions, we also cannot compensate you. If your survey responses include poor qualitative (written) responses , we cannot compensate you for your participation. Poor quality qualitative responses include, but are not limited to, nons ensical text or lines copied and pasted from other internet sources. The rigor of our scientific study depends on high quality data. If your survey responses include poor quantitative (bubbles) responses , we cannot compensate you for your participation. P oor quality quantitative responses include, but are not limited to, selecting the same numerical answer choice over and over again (e.g., 3,3,3,3,3,3,3,3,3,3,3,3,3) and answering in a way that does not make psychological sense given the questions asked. Th e rigor of our scientific study depends on high quality data. If you type the wrong survey code into the Mturk survey code box , we cannot compensate you for your participation as we cannot ensure you are a human participant who is eligible for this research survey. If you fail the CAPTCHA check , we cannot compensate you for your participation as we cannot ensure you are a human participant who is eligible for this research survey. If you do not correctly answer attention check items , we cannot co mpensate you for your participation as we cannot be sure you have provided quality data. 8. CONTACT INFORMATION If you have concerns or questions about this study, such as scientific issues, how to do any part of it, or to report an injury, please contact Lauren Collier at colli719@msu.edu or Ann Marie Ryan Ph. D., Department of Psychology, Mic higan State University, East Lansing, MI 48824, phone: 517 - 353 - 8855, e - mail: ryanan@msu.edu. If you have questions or concerns about your role and rights as a research participant, would like to obtain information or offer input, or would like to registe r a complaint about this study, you Protection Program at 517 - 355 - 2180, Fax 517 - 432 - 4503, or e - mail irb@msu.edu or regular mail at 4000 Collins Rd, Suite 136, Lansing, MI 48910. 9. DOCUMENTATION OF INFORMED CONSENT. 160 If you would like a copy of the consent form, please email Lauren Collier (colli719@msu.edu). 161 Research Participant Information and Consent Form Study 3 You are being asked to participate in a research study. Researchers are required to provide a consent form to inform you about the research study, to convey that participation is voluntary, to explain risks and benefits of participation, and to empower you to make an informed decision. You should feel free to ask the researchers any questions you may have. Study Title: AE Study 3 1. PURPOSE OF RESEARCH The purpose of this research study is to reveal how people respond to a variety of scenarios that may occur in everyday life and learn more about what people think of certain behaviors. 2. WHAT YOU WILL DO Consenting participants will be asked to read four scenarios and type their responses to each one. Participants will also be asked to respond to several behavioral and perso nality questions regarding certain actions. 3. POTENTIAL BENEFITS You may not benefit personally from being in this study. However, we hope that this research may eventually benefit others hoping to learn more about behaviors people are more or less likely to engage in. 4. POTENTIAL RISKS The researchers do not expect that there are any potential risks to completing this study. 5. PRIVACY AND CONFIDENTIALITY This study is c onfidential. Your answers will only be associated with an anonymous ID. To help us protect your confidentiality, please do not write or give your name or any other identifying information during the study. All data will be stored on the hard drive of a sec ure computer and will only be accessed by trained experimenters. Data will be stored for five years after the publication of research stemming from this project --- as specified by the American Psychological Association. 6. YOUR RIGHTS TO PARTICIPATE, SAY NO, OR WITHDRAW You have the right to say no to participate in the research. You can stop at any time after it has already started. There will be no consequences if you stop and you will not be criticized. You will not lose any benefits that you norma lly receive. 7. COSTS AND COMPENSATION FOR BEING IN THE STUDY Participants will receive $2.00 for participating in this brief study. The study is projected to take no more than 40 minutes. Compensation Rules (PLEASE READ) 162 The following are reasons w hy we would not be able to compensate you for your participation . By following these compensation rules, we hope to be as fair as possible to survey respondents who meet the study criteria, who access the survey only once, and who provide quality data for our study. Please note: If you do not include your MTurk ID in the online survey we cannot identify you and so you will not be compensated if you fail to correctly enter your Mturk ID in the online survey. If we have no record of your Mturk ID in our da ta, we cannot compensate you. If you are not eligible to take this research survey based on the prescreening questions , we cannot compensate you for your participation. The quality of our scientific study depends on participants meeting these criteria. If we find that you have re - entered the survey multiple times after initially failing the prescreening questions, we also cannot compensate you. If your survey responses include poor qualitative (written) responses , we cannot compensate you for your part icipation. Poor quality qualitative responses include, but are not limited to, nonsensical text or lines copied and pasted from other internet sources. The rigor of our scientific study depends on high quality data. If your survey responses include poor quantitative (bubbles) responses , we cannot compensate you for your participation. Poor quality quantitative responses include, but are not limited to, selecting the same numerical answer choice over and over again (e.g., 3,3,3,3,3,3,3,3,3,3,3,3,3) an d answering in a way that does not make psychological sense given the questions asked. The rigor of our scientific study depends on high quality data. If you type the wrong survey code into the Mturk survey code box , we cannot compensate you for your par ticipation as we cannot ensure you are a human participant who is eligible for this research survey. If you fail the CAPTCHA check , we cannot compensate you for your participation as we cannot ensure you are a human participant who is eligible for this r esearch survey. If you do not correctly answer attention check items , we cannot compensate you for your participation as we cannot be sure you have provided quality data. 8. CONTACT INFORMATION If you have concerns or questions about this study, such as scientific issues, how to do any part of it, or to report an injury, please contact Lauren Collier at colli719@msu.edu or Ann Marie Ryan Ph. D., Department of Psychology, Michigan State University, East Lansing, MI 48824, phone: 517 - 353 - 8855, e - mail: ryanan@msu.edu. If you have questions or concerns about your role and rights as a research participant, would like to obtain information or offer input, or would like to register a complaint about this study, you Protection Program at 517 - 355 - 2180, Fax 517 - 432 - 4503, or e - mail irb@msu.edu or regular mail at 4000 Collins Rd, Suite 136, Lansing, MI 48910. 163 9. DOCUMENTATION OF INFORMED CONSENT. If you w ould like a copy of the consent form, please email Lauren Collier (colli719@msu.edu). 164 APPENDIX K : Debriefing Form s for All Studies Debriefing Form Studies 1 and 2 Thank you for participating in our study. This form is designed to provide you with information about the purpose and importance of this study. The purpose of this study was to learn more about allyship. Specifically, we hoped to determine how people conceptualize allyship and to learn whether people engage in effective allyship behaviors. The experimental design was relatively straightforward and is of the type often encountered in psychological research. Given the mild nature of the experimental design, we anticipate that there are and will be no risks involved for any of our participants. However, if you did recall an event that negatively impacted you, please contact the appropriate number below: National suicide hotline (phone: 1 - 800 - 273 - 8255) Emergency number (phone: 911) Additionally, if you have questions or concerns regarding this study, please do not hesitate to contact the investigators. Additionally, if you would like more information about the study or have further questions about it, please feel free to contact: L auren Collier, Department of Psychology, Michigan State University, East Lansing, MI 48824, e - mail: colli719@msu.edu. OR Ann Marie Ryan, Ph.D., Department of Psychology, Michigan State University, East Lansing, MI 48824, phone: 517 - 355 - 0203, e - mail: ryan an@msu.edu 165 Debriefing Form Study 3 Thank you for participating in our study. This form is designed to provide you with information about the purpose and importance of this study. The purpose of this study was to learn more about allyship. Specifically, we hoped to determine how people conceptualize allyship and to learn whether people engage in effective allyship behaviors. To learn more about allyship and other research conducted on this topic, please feel free to visit the fo llowing information sources: Ostrove, J. M., & Brown, K. T. (2018). Are allies who we think they are?: A comparative analysis. Journal of Applied Social Psychology, 48(4), 195 204. https://doi.org/10.1111/jasp.12502 Brown, K. T., & Ostrove, J. M. (2013 ). What does it mean to be an ally?: The perception of allies from the perspective of people of color. Journal of Applied Social Psychology, 43(11), 2211 2222. http://dx.doi.org.proxy2.cl.msu.edu/10.1111/jasp.12172 Collins, J. C., & Chlup, D. T. (2014). Criticality in Practice: The Cyclical Development Process of Social Justice Allies at Work. Advances in Developing Human Resources, 16(4), 481 498. https://doi.org/10.1177/1523422314544295 The experimental design was relatively straightforward and is of the type often encountered in psychological research. Given the mild nature of the experimental design, we anticipate that there are and will be no risks involved for any of our participants. However, if you did recall an event that negatively impacted yo u, please contact the appropriate number below: National suicide hotline (phone: 1 - 800 - 273 - 8255) Emergency number (phone: 911) Additionally, if you have questions or concerns regarding this study, please do not hesitate to contact the investigators. Add itionally, if you would like more information about the study or have further questions about it, please feel free to contact: Lauren Collier, Department of Psychology, Michigan State University, East Lansing, MI 48824, e - mail: colli719@msu.edu. OR A nn Marie Ryan, Ph.D., Department of Psychology, Michigan State University, East Lansing, MI 48824, phone: 517 - 355 - 0203, e - mail: ryanan@msu.edu 166 APPENDIX L : Items Presented to Participants in Study 2 Item Number Item Content Ineffective 1 Used slurs. Ineffective 2 Emphasized positive stereotypes of my group. Ineffective 3 Reinforced stereotypes about my group. Ineffective 4 Told me I don't act like a normal member of my group. Ineffective 5 Exhibited bias. Ineffective 6 Spoke up for me in a way that was ignorant. Ineffective 7 Spoke in a manner that is stereotypical to my group to try to make me feel accepted. Ineffective 8 Treated all members of my group as though we are interchangeable. Ineffective 9 Made offensive comments about my group. Ineffective 10 Made uneducated comments about me based on stereotypes of my group. Ineffective 11 Made assumptions about me based on my group. Ineffective 12 Physically fought someone who mistreated me. Ineffective 13 Used derogatory language ag ainst others in the situation. Ineffective 14 Stepped in to defuse the situation when I did not want them to. Ineffective 15 Defended me when I did not want them to. Ineffective 16 Spoke up for me when I was not offended. Ineffective 17 Tried to defend me without a full understanding of the situation. Ineffective 18 Thought they know what was best for me, even if it went against my wishes. Ineffective 19 Acted like they knew everything about my identity group, despite being very uninformed. Ine ffective 20 Thought they understood issues that affect my identity group but they do not. Ineffective 21 Attempted to defend me without paying attention to the context Ineffective 22 Made light of a discriminatory event that hurt me. Ineffective 23 Dismissed concerns I had about the treatment of my group. Ineffective 24 Did not put much effort into protecting me. Ineffective 25 Agreed with the stereotypes others were using against me. Ineffective 26 Joined forces with those who were discriminating against me. Ineffective 27 Cared more about sharing their opinion than listening to me. Ineffective 28 Focused more on their own thoughts rather than my experiences Ineffective 29 Refused to listen to my concerns. Ineffective 30 Did not address the discrimination I experienced in front of them Ineffective 31 Stayed silent while I was experiencing discrimination. Ineffective 32 Gave me advice I found to be offensive. Ineffective 33 Gave me advice from a place of p rivilege. Ineffective 34 Gave me advice that I feel I cannot use as a minority. Ineffective 35 Made my circumstances worse by drawing attention to my minority status. Ineffective 36 Brought unwanted attention to the situation. 167 Ineffective 37 Used positive stereotypes of my group to convince others to leave me alone. Ineffective 38 Provided unnecessary information to others, making the situation worse. Ineffective 39 Revealed personal information about me without my permission. Ineffective 40 Sha red personal information about me that negatively affected me. Ineffective 41 Tried to control the way I expressed myself. Ineffective 42 Restricted me from being able to fully express myself. Ineffective 43 Treated me in a condescending manner. Ineffective 44 Was patronizing towards me. Ineffective 45 Ineffective 46 Intervened for me in a way that made me look weak. Ineffective 47 Made excuses for the biased comments of others. Ineffective 48 Spok e for me without understanding my point of view. Ineffective 49 Tried to speak for me in a way that was inaccurate. Ineffective 50 Did not address the offensive behavior of others. Ineffective 51 Asked me inappropriate questions about my identity. Ineffective 52 Enabled others when they made biased comments towards me. Effective 1 Spoke up for me when I was being harassed because of my group. Effective 2 Intervened when others were making disparaging comments about my group. Effective 3 Spoke up for me when others used slurs against me. Effective 4 Spoke up when others were treating me unfairly due to being a member of my group. Effective 5 Insisted that others treat me like a human being. Effective 6 Advocated for me when I was treated worse than others. Effective 7 Educated others about issues that affect my minority group. Effective 8 Helped others understand more about my identity group. Effective 9 Made me feel understood by others in a privileged group. Effective 10 Spoke up for me when I was being ignored. Effective 11 Spoke up with me against injustices faced by my group. Effective 12 Spoke up in a way that provided for me to advocate for myself. Effective 13 Allowed me to speak up for myself. Effec tive 14 Amplified my voice as I spoke up against injustice. Effective 15 Helped me get opportunities that I was denied because of my identity group. Effective 16 Advocated for me to receive opportunities. Effective 17 Deescalated the situation. Effective 18 Helped me fit in socially with those in their identity group. Effective 19 Vouched for my abilities in front of others. Effective 20 Defended my identity group when others made disparaging comments. Effective 21 Called out those who spoke ill of my identity group. Effective 22 Defended me by asking my attacker to leave the area. Effective 23 Continued to speak up for me, even in the face of criticism. Effective 24 Advocated for me persistently. Effective 25 Did not let the issue go when others discriminated against me. Effective 26 Protected me from verbal harassment. 168 Effective 27 Spoke up for me when other people harassed me. Effective 28 Made it known that intolerance was unacceptable. Effective 29 Made it clea r that they would not abide with any mistreatment of those in my identity group. Effective 30 Defended me against biased statements in an educated way. Effective 31 Challenged the thinking of people who made biased comments against me. Effective 32 Competently explained why my harasser's comments were not acceptable. Effective 33 Listened to my feelings about experiencing negative treatment as a result of my identity group. Effective 34 Made me feel heard. Effective 35 Made me feel ac cepted when other people were purposefully excluding me. Effective 36 Made it clear to others that I was welcome. Effective 37 Affirmed my identity. Effective 38 Validated my feelings about my mistreatment. Effective 39 Affirmed that the biased treatment I experienced was not acceptable. Effective 40 Persuaded others to support the cause of people from my identity group. Effective 41 Rallied support for me when I was being discriminated against. Effective 42 Used their position of privilege to defend me against biased comments. Effective 43 Used their position to defend me against discriminatory treatment. Effective 44 Helped communicate my thoughts effectively to those with privilege. Effective 45 Defended me when others threatened my sense of safety. Effective 46 Made me feel like I wasn't alone. Effective 47 Made sure that my needs were considered. Effective 48 Stopped others from asking invasive questions of me. Effective 49 Protected me from physical attacks. Effective 50 Corrected others who made biased comments against me. Effective 51 Reported those who discriminated against me to the proper authorities. Outcome 1 My ally speaking up for me made things worse for me. Outcome 2 My ally's attempt to help me made the situation worse. Outcome 3 My ally's advocacy changed how others treated me for the better. Outcome 4 My ally's actions resulted in positive change for me. 169 APPENDIX M : Majority Group Perceptions of Ally Behaviors Instructions: Using the ow helpful do you believe each of the following behaviors are to members of marginalized groups (e.g., racial, ethnic, sexual and gender minorities)? Instructions: Using the response option ow frequently do you complete each of the following behaviors when engaging with members of marginalized groups (e.g., racial, ethnic, sexual and gender minorities)? Instructions: Using the response op ow willing are you to complete each of the following behaviors when engaging with members of marginalized groups (e.g., racial, ethnic, sexual and gender minorities)? 1. Sharing personal information a bout the person without asking. 2. Giving advice from a place of privilege. 3. Not listening to the person's concerns. 4. Asking the person inappropriate questions about their identity. 5. Putting little effort into helping the person when they are in need. 6. Making offensive comments about the person's identity group. 7. Telling the person that they do not act like a normal member of their identity group. 8. Making a person's circumstances worse by drawing attention to their minority status. 9. Restricting the pe rson from being able to fully express themselves. 10. Not addressing the offensive behavior of others. 11. Making excuses for the biased comments of others. 12. Being patronizing. 13. Making sure that the person's needs are considered. 14. Making the person feel hea rd. 15. Helping communicate the person's thoughts effectively to those with privilege. 16. Speaking up in a way that provided the person an opportunity to advocate for themselves. 17. Helping others understand more about the person's identity group. 18. Insisting that others treat the person like a human being. 19. Defending the person against biased statements in an educated way. 20. Making the person feel accepted when other people are purposefully excluding them. 21. Advocating for the person when they are treated wor se than others. 22. Speaking up for the person when they are being ignored. 23. Protecting the person from verbal harassment. 24. Rallying support for the person when they are being discriminated against. 170 APPENDIX N : Social Dominance Orientation Instructions : feeling towards? Beside each object or st atement, place a number from '1' to '7' which represents the degree of your positive or negative feeling." The scale was labeled very positive (7), positive (6), slightly positive (5), neither positive nor negative (4), slightly negative (3), negative (2), and very negative (1). 1. Some groups of people are simply inferior to other groups. 2. In getting what you want, it is sometimes necessary to use force against other groups. 3. It's OK if some groups have more of a chance in life than othe rs. 4. To get ahead in life, it is sometimes necessary to step on other groups. 5. If certain groups stayed in their place, we would have fewer problems. 6. It's probably a good thing that certain groups are at the top and other groups are at the bottom. 7 . Inferior groups should stay in their place. 8. Sometimes other groups must be kept in their place. 9. It would be good if groups could be equal. 10. Group equality should be our ideal. 11. All groups should be given an equal chance in life. 12. We should do what we can to equalize conditions for different groups. 13. Increased social equality i s beneficial to society . 14. We would have fewer problems if we treated people more equally. 15. We should strive to make incomes as equal as possible. 16. No one g roup should dominate in society. 171 APPENDIX O : Motivation to Respond Without Prejudice Instructions: The following questions concern various reasons or motivations people might have for trying to respond in nonprejudiced ways towards minority groups. Minority groups include racial and ethnic minorities as well as members of sexual minority groups (LGBTQ+). Please (Strongly Agree) scale to indicate yo ur responses. 1. - prejudice towards minority groups. 2. I try to hide any negative thoughts about minority groups in order to avoid negative reactions from others. 3. If I acted prejudiced t owards minority groups, I would be concerned that others would be angry with me. 4. I attempt to appear nonprejudiced toward minority groups in order to avoid disapproval from others. 5. I try to act nonprejudiced towards minority groups because of pressure from others. 6. I attempt to act in nonprejudiced ways towards minority groups because it is important personally to me. 7. According to my personal values, using stereotypes about minority groups is OK. 8. I am personally motivated by my beliefs to be nonprejudiced to wards minority groups. 9. Because of my personal values, I believe that using stereotypes about minority groups is wrong. 10. Being nonprejudiced towards minority groups is important to my self - concept. 172 APPENDIX P : Prosocial Motivations Scale Instructions: Think back over the situation you described. Please respond using the scale statement applies to the following question: Why did you decide to help? 1. Because I care about benefiting others. 2. Because I want to help others. 3. Because I want to have positive impact on others. 4. Because it is important to me to do good for others. 173 APPENDIX Q : The Selfishness Questionnaire after your own self - interests. Answer the following questions as honestly as you can by indicating whether you: Strongly Disagree (1) Disagree (2), Neither Agr ee nor Disagree (3), Agree (4) or Strongly Agree (5) with each statement. 1. 2. 3. ends to gain an advantage. 4. At the end of the day I care mostly for myself, my family, and friends who can help me. 5. 6. 7. Even if it meant giving my kids an unfair advantage ove 8. Sometimes you need to take advantage of other people before they take advantage of you. 9. 10. When it comes to helping myself or helping others, I tend t o help myself. 11. 12. take it for myself and my family. 13. Quite often in life, it is more impor tant to receive than to give. 14. I know I love rewards in life, even if there is a cost to others. 15. misspent. 16. 17. 18. I care for myself much more than I care for others. 19. 20. I sometimes lie to others for my own good, and theirs too. 21. 22. I go out of the way to exploit situations for my own advantage. 23. h person. 174 APPENDIX R : Short Form HEXACO (Honesty - Humility and Openness) Instructions: On the following pages you will find a series of statements about you. Please read each statement and decide how much you agree or disagree with that statement. Then w rite your response in the space next to the statement using the following scale: Please answer every statement, even if you are not completely sure of your response . 1. I would be quite bored by a visit to an art gallery. 2. I wouldn't use flattery to get a raise or promotion at work, even if I thought it would succeed. 3. I'm interested in learning about the history and politics of other countries. 4. If I knew that I could never get caught, I would be willing to steal a million dollars. 5. I would enjoy creating a work of art, such as a novel, a song, or a painting. 6. Having a lot of money is not especially important to me. 7. I think that paying attention to radical ideas is a waste of time. 8. I think that I am entitled to more respect than the average person is. 9. If I had the opportunity, I would like to attend a classical music concert. 10. If I want something from someone, I will laugh at that person's worst jokes. 11. . 12. I would never accept a bribe, even if it were very large. 13. People have often told me that I have a good imagination. 14. I would get a lot of pleasure from owning expensive luxury goods. 15. I like people who have unconventional views. 16. I want people to know that I am an important person of high status. 17. 18. 19. I find it boring to discuss philosophy. 20. tempted to use counterfeit money, if I were sure I could get away with it. 175 APPENDIX S : Political Orientation Assessment Instructions: Please answer the following questions openly and honestly ervative) scale. 1. Overall, where would you place yourself, on the following scale of liberalism - conservatism? 2. In terms of social and cultural issues (e.g. abortion, separation of Church and State, affirmative action) where would you place yourself on the following scale? 3. In terms of economic issues (e.g. taxation, welfare, privatization of social security) where would you place yourself on the following scale? 176 APPENDIX T : Interpersonal Reactivity Index (IRI) Instructions: The following statements inquire about your thoughts and feelings in a variety of situations. For each item, indicate how well it describes you. Answer as honestly as you can. Please indicate how well ea ch item describes you using a (Describes me very well) scale. Empathic Concern I often have tender, concerned feelings for people less fortunate than me. Sometimes I don't feel very sorry for other people when they are having problems. When I see someone being taken advantage of, I feel kind of protective towards them. Other peopl e's misfortunes do not usually disturb me a great deal. When I see someone being treated unfairly, I sometimes don't feel very much pity for them. When I see someone being treated unfairly, I sometimes don't feel very much pity for them. I am often quit e touched by things that I see happen. I would describe myself as a pretty soft - hearted person. Perspective Taking 1. I sometimes find it difficult to see things from the "other guy's" point of view. 2. I try to look at everybody's side of a disagreement before I make a decision. 3. I sometimes try to understand my friends better by imagining how things look from their perspe ctive. 4. If I'm sure I'm right about something, I don't waste much time listening to other people's arguments. 5. I believe that there are two sides to every question and try to look at them both. 6. When I'm upset at someone, I usually try to "put myself in his shoes" for a while. 7. Before criticizing somebody, I try to imagine how I would feel if I were in their place. 177 APPENDIX U : Self - Importance of Moral Identity Scale (SIMI) Instructions: LISTED BELOW ARE SOME CHARACTERISTICS THAT MAY DESCRIBE A PERSON: C aring Compassionate Fair Friendly Generous Helpful Hardworking Honest Kind For a moment, visualize in your mind the kind of person who has these characteristics. Imagine how that person would think, feel, and act. When you have a clear image of what this person would be like, answer the following questions by selecting the approp riate answer choice using the scale provided ( = Strongly Disagree , = Strongly Agree) . 1. It would make me feel good to be a person who has these characteristics. 2. Being someone who has these characteristics is an important part of who I am. 3. I would be ashamed to be a person who has these characteristics. 4. Having these characteristics is not really important to me. 5. I strongly desire to have these characteristics. 6. I often buy products that communicate the fact that I have these characte ristics. 7. I often wear clothes that identify me as having these characteristics. 8. The kinds of books and magazines that I read identify me as having these characteristics. 9. The fact that I have these characteristics is communicated to others by my membership in certain organizations. 10. I am actively involved in activities that communicate to others that I have these characteristics. 178 APPENDIX V : Moral Relativism Scale (MRS) Instructions: You are about to read ten statements regarding the way people think about morality. Please rate the degree to which you personally agree (or disagree) with each statement using the following scale: 1 = Strongly Disagree 2 = Somewhat Disagree 3 = Neither Agree nor Disagree 4 = Somewhat Agree 5 = Strongly Agree Remember that there are no right or wrong answers. We would only like to know your honest response to each statement. 1. Different people can have opposing views on what is moral and immoral without anyone being wrong 2. People can disagree on what is morally right without anyone being wrong. 3. Two different cultures could have 4. 5. 6. There is a moral standard that all actions sho uld be held to, even if cultures disagree. 7. Each person is the final authority on whether his or her actions really are morally correct 8. An action is only morally wrong if a person believes it is morally wrong. 9. There are moral rules that apply to everyone r egardless of personal beliefs. 10. The same moral standards should be followed by people from all cultures. 179 APPENDIX W : Willingness to Learn About Marginalized Groups please indicate the frequency with which you engage in the behaviors outlined in the statements on this minorities (e.g. LGBTQ+). 1. I seek out information about indivi duals from marginalized groups. 2. I enjoy learning about people from different minority backgrounds. 3. I ask people from marginalized groups about their life experiences. 4. I read books, blogs and online articles by authors from marginalized groups. 5. I am always open to learning more about the experiences of people from marginalized groups. 180 REFERENCES 181 REFERENCES Ariel, C. (2017, August 16). For our white friends desiring to be allies. Retrieved October 11, 2018, from https://sojo.net/articles/our - white - friends - desiring - be - allies Ashton , M. C., & Lee, K. (2009). The HEXACO - 60: a short measure of the major dimensions of personality. Journal of Personality Assessment , 91(4), 340 345. https://doi.org/10.1080/00223890902935878 Ashbur n - Nardo, L., Morris, K. A., & Goodwin, S. A. (2008). The confronting prejudiced responses (CPR) model: Applying CPR in Organizations. Academy of Management Learning & Education , 7 (3), 332 342. Aquino, K., Freeman, D., Reed, A. I. I., Felps, W., & Lim, V. K. G (2009). Testing a social cognitive model of moral behavior: The interaction of situational factors and moral identity centrality. Journal of Personality and Social Psychology, 97, 123 141. doi:10.1037/a0015406 Aquino, K., & Reed, A. II. (2002). The self - importance of moral identity. Journal of Personality and Social Psychology, 83 (6), 1423 - 1440. http://dx.doi.org/10.1037/0022 - 3514.83.6.1423 Avery, D. R., McKay, P. F. & Wilson, D. C., Tonidandel S. (2007). Unequal attendance: The relationships between race, organizational diversity cues, and absenteeism. Pers. Psychol . 60(4):875 902. Batson, C. D. (1987). Prosocial motivation: Is it ever truly altruistic? In L. Berkowitz (Ed.), Advances in Experimental Social Psychology (Vol. 20, pp. 65 122). Academic Press. https://doi.org/10.1016/S0065 - 2601(08)60412 - 8 Batson, C. D., Lishner, D. A., & Stocks, E. L. (2015). The empat hy altruism hypothesis. In The Oxford handbook of prosocial behavior (pp. 259 281). New York, NY, US: Oxford University Press. https://doi.org/10.1093/oxfordhb/9780195399813.013.023 Ba tson, D., Lange, P., Ahmad, N., & Lishner, D. (2003). Altruism and helping behavior (pp. 279 295). https://doi.org/10.4135/9781848608221.n11 Blau, P. M. (1964). Justice in Social Exchange. Sociolog ical Inquiry , 34 (2), 193 206. https://doi.org/10.1111/j.1475 - 682X.1964.tb00583.x Bishop, A. (2002). Becoming an ally: Breaking the cycle of oppression in people (2nd ed.). City, Canada: Fe rnwood Publishing. 182 Broido, E. M. (2000). The development of social justice allies during college: A phenomenological investigation. Journal of College Student Development , 41 (1), 3 18. Brooks, A. K., & Edwards, K. (2009). Allies in the workplace: Including LGBT in HRD. Advances in Developing Human Resources; San Francisco , 11 (1), 136. Broudy, R., Brondolo, E., Coakley, V., Brady, N., Cassells, A., Tobin, J. N., & Sweeney, M. (2007). Per ceived ethnic discrimination in relation to daily moods and negative social interactions. Journal of Behavioral Medicine , 30(1), 31 43. https://doi.org/10.1007/s10865 - 006 - 9081 - 4 Brown, K. T., & Ostrove, J. M. (2013). What does it mean to be an ally?: The perception of allies from the perspective of people of color. Journal of Applied Social Psychology; Hoboken , 43 (11), 2211 2222. http://dx.doi.org.proxy2.cl.msu.edu/10.1111/jasp.12172 Butz, D. A., & Plant, E. A. (2009). Prejudice control and interracial relations: The role of motivation to respond without prejudice. Journal of Personality , 77(5), 1311 1342. https://doi.org/10.1111/j.1467 - 6494.2009.00583 .x Campbell, J.P., McCloy, R.A., Oppler, S.H. and Sager, C.E. (1993) A Theory of Performance. In: Schmitt, N. and Borman, W.C., Eds., Personnel Select ion in Organizations, Jossey - Bass, San Francisco, 3570. Carney, D. R., Jost, J. T., Gosling, S. D., & Potter, J. (2008). The secret lives of liberals and conservatives: Personality profiles, interaction styles, and the things they leave behind. Political Psychology , 29 (6), 807 840. https://doi.org/10.1111/j.1467 - 9221.2008.00668.x Charles, C. (2016, October 15). 10 Common things well - intentioned allies do that are actually counterproductive. Retrieved October 5, 2018, from https://medium.com/reclaiming - anger/10 - common - things - well - intentioned - allies - do - that - are - actually - counterproductive - af3dda187822 Chou, S. Y., & Stauffer, J. M. (2016). A theoretical classification of helping behavior and helping motives. Personnel Review; Farnborough , 45 (5), 871 888. Cohen, J. (1988). Statistical Power Analysis for the Behavioral Sciences. Routledge. Cohen, S., & McKay, G. (1984). Social support, stress and the buffering hypothesis: A theoretical analysis. In A. Baum, S. E. Taylor, & J. E. Singer (Eds.), Handbook of psychology and health (pp. 253 - 267). Hillsdale, NJ: Erlbaum. Cohen, T. R., Panter, A. T., Turan, N., Morse, L., & Kim, Y. (2013). Agreement and similarity in self - other perceptions of moral character (SSRN Scholarly Paper No. ID 2428561). Rochester, NY: Social Science Research Network. Retri eved from https://papers.ssrn.com/abstract=2428561 183 Collier, L.A. (2017). Towards the development of a belief in moral relativism scale. Unpublished Collier - Spruel, Hawkins, Jayawickreme, Furr & Fleeson, (2019). R elativism or tolerance? Defining, assessing, connecting, and distinguishing two moral personality features with prominent roles in modern societies, 1 - 19 . Collins, J. C., & Chlup, D. T. (2014). Crit icality in Practice: The cyclical development process of social justice allies at work. Advances in Developing Human Resources , 16 (4), 481 498. https://doi.org/10.1177/1523422314544295 Costello, A. B., & Osborne, J. W. (2005). Best practices in exploratory factor analysis: Four recommendations for getting the most from your analysis. Practical Assessment, Research & Evaluation , 10 , 173 178. Cortina, L. M. (2008). Unseen i njustice: Incivility as modern discrimination in organizations. The Academy of Management Review , 33 (1), 55 75. https://doi.org/10.2307/20159376 Curtin, N., Kende , A., & Kende, J. (2016). Navigating multiple identities: The simultaneous influence of advantaged and disadvantaged identities on politicization and activism. Journal of Social Issues , 72 (2), 264 285. htt ps://doi.org/10.1111/josi.12166 Davidson, M. N. (2012, November 1). How hard should you push diversity? Retrieved October 7, 2018, from https://hbr.org/2012/11/how - hard - should - you - push - diversity Davies P. G., Spencer S. J. & Steele C. M., (2005). Clearing the air: Identity safety moderates the J. Personal. Soc. Psychol . 88:276 87. Davis, M.H. (1980) A multidimensional approach to individual differences in empathy. J SAS Catalog of Selected Documents in Psychology , 10, 85. Davis, M. H., & Maitner, A. T. (2010). Perspective taking and intergroup helping. In S. Stürmer & M. Snyder (Eds.), The psychology of prosocial behav ior: Group processes, intergroup relations, and helping (pp. 175 - 190). : Wiley - Blackwell. Demsky, C. A., Rineer, J. R., Crain, T. L., Ellis., A. M. (2019). Advancing the Women of I - O: Allies in Action. Symposium presented at the annual conference of the S ociety for Industrial and Organizational Psychology, National Harbor, MD. De Vries, R. (2013). The 24 - item Brief HEXACO Inventory (BHI). Journal of Research in Personality , 47 , 871 880. https://doi .org/10.1016/j.jrp.2013.09.003 Dickey, M. (2018). Diverse teams are still *really* good for business, McKinsey says. Retrieved August 15, 2018, from http://social.techcrunch.com/2018/01/18/diverse - teams - are - still - really - good - for - business - mckinsey - says/ 184 Dovidio, J. F., Piliavin, J. A., Schroeder, D. A., & Penner, L. (2006). The social psychology of prosocial behavior. Mahwah, NJ, US: Lawr ence Erlbaum Associates Publishers. Ehrhart, M. G., & Naumann, S. E. (2004). Organizational Citizenship Behavior in Work Groups: A Group Norms Approach. Journal of Applied Psychology , 89 (6), 960 974. http://dx.doi.org.proxy2.cl.msu.edu/10.1037/0021 - 9010.89.6.960 Ferdman, B. (2014). The practice of inclusion in diverse organizations: Toward a systemic and inclusive framework. In Diversity at Work: The Practice of Inclusion (pp. 3 54). https://doi.org/10.1002/9781118764282.ch1 Fisher, R. J., & Katz, J. E. (2000). Social - desirability bias and the validity of self - reported values . Psychology & Marketing , 17(2), 105 120. https://doi.org/10.1002/(SICI)1520 - 6793(200002)17:2<105::AID - MAR3>3.0.CO;2 - 9 Fleming, P. (2012). Social desirability, not what it seems: A r eview of the implications for self - reports. The International Journal of Educational and Psychological Assessment , 11 (1), 3 22. Gardner (2018). The role of race and gender in effectively promoting organizational diversity Goffman, E. (1971). Relations in public: Microstudies of the public order. New York: Basic Books. Goldstein, N. J., Vezich, I. S., & Shapiro, J. R. (2014). Perceived perspective takin g: When others walk in our shoes. Journal of Personality and Social Psychology; Washington , 106 (6), 941. http://dx.doi.org.proxy2.cl.msu.edu/10.1037/a0036395 Grant, A. M. (2008). Does intrinsic motivation fuel the prosocial fire? Motivational synergy in predicting persistence, performance, and productivity. The Journal of Applied Psychology , 93 (1), 48 58. https://doi.org/10.1037/0021 - 9010.93.1.48 Griffin, P. (1997). Introductory module for the single issue course. In M. Adams, L. A. Bell, & P. Griffin (Eds.), Teaching for diversity and social justice: A sourcebook. New York, NY: Routledge Hartley, A. G., Furr, R. M., Helzer, E. G., Jayawickreme, E., Velasquez, K. R., & Fleeson, W. Social Psychological and Personality Scienc e , 7 (7), 648 657. https://doi.org/10.1177/1948550616655359 Hewlett, S. A., Marshall, M., & Sherbin, L. (2013, December 1). How diversity can drive innovation. Retrieved August 15, 2018, from https://hbr.org/2013/12/how - diversity - can - drive - innovation 185 Ho, A. K., Sidanius, J., Pratto, F., Levin, S., Thomsen, L., Kteily, N., & Sheehy - Skeffington, J. (2012). Social dominance orien tation: Revisiting the structure and function of a variable predicting social and political attitudes . Personality and Social Psychology Bulletin . https://doi.org/10.1177/0146167211432765 Hofhuis, J., van der Zee, K., & Otten, S. (2016). Dealing with differences: The impact of perceived diversity outcomes on selection and assessment of minority candidates. The International Journal of Human Resource Management , 27 , 1318 1339. https://doi.org/10.1080/09585192.2015.1072100 Hofhuis, J., Zee, K. I. V. der, & Otten, S. (2014). Comparing antecedents of voluntary job turnover among majority and minority employees. Equality, Diversity an d Inclusion: An International Journal; Birmingham , 33 (8), 735 749. http://dx.doi.org.proxy2.cl.msu.edu/10.1108/EDI - 09 - 2013 - 0071 Ilgen D. R., Youtz M. A., (1986). Factors affectin g the evaluation and development of minorities in organizations. In Research in Personnel and Human Resource Management: A Research Annual, ed. K Rowland, G Ferris, pp. 307 37. Greenwich, CT: JAI Inzlicht M, & Schmader T, (2012). Stereotype Threat: Theory , Process, and Application . New York: Oxford Univ. Press James E. H., (2000). Race - related differences in promotions and support: Underlying effects of human and social capital. Organ. Sci. 11:493 508. Johns J, Inzlicht M, Schmader T. 2008. Stereotype threat and executive resource depletion: Examining the influence of emotion regulation. J. Exp. Psychol. Gen . 137:691 705. Kabat - Farr, D., & Cortina, L. M. (2012). Selective incivility: Gender, race, and the discriminatory workplace. I n New Horizons in Management . Gender and the dysfunctional workplace (pp. 120 134). https://doi.org/10.4337/9780857932600.00014 teinke, J., & Bowling, N. A. (2013). A meta - analytic evaluation of diversity training outcomes. Journal of Organizational Behavior , 34(8), 1076 1104. https://doi.org/10.1002/job.1839 Kyriazos, T. A. (2018). Applied psychometrics: Sample size and sample power considerations in factor analysis (EFA, CFA) and SEM in General. Psychology , 09 , 2207. https://doi.org/10.4236/psych.2018.98126 Landau J. (1995). potential. Journal of Organizational Behavior, 16 , 391 400. Larson, E. (2017). New Research: Diversity + inclusion = better decision making at work. Retrieved August 15, 2018, from 186 https://www.forbes.com/sites/eriklarson/2017/09/21/new - research - diversity - inclusion - better - decision - making - at - work/#6bb38b214cbf Lerner, J. S., & Keltner, D. (2001). Fear, anger, and risk. Journal of Personality and Social Psychology , 81 (1), 146 159. https://doi.org/10.1037/0022 - 3514.81.1.146 Li, A., & Bagger, J. (2007). The balanced inventory of desirable responding (BIDR): A reliability generalization study. Educational and Psychological Measure ment , 67 (3), 525 544. https://doi.org/10.1177/0013164406292087 Liu S. - N., Dray K., and Sabat I. (2019). Building better workplace allies: Where are we lacking and how can we get there? Symposium pre sented at the annual conference of the Society for Industrial and Organizational Psychology , National Harbor, MD. Luan, Z., Poorthuis, A. M. G., Hutteman, R., Denissen, J. J. A., Asendorpf, J. B., & van Aken , M. A. G. (2019). Unique predictive power of other - rated personality: An 18 - year longitudinal study. Journal of Personality , 87 (3), 532 545. https://doi.org/10.1111/jopy.12413 Major, B., Quinton, W. J., & Schmader, T. (2003). Attributions to discrimination and self - esteem: Impact of group identification and situational ambiguity. Journal of Experimental Social Psychology , 39 (3), 220 231. https://doi.org/10.1016/S0022 - 1031(02)00547 - 4 Engaging Cultural Differences: The Multicultural Challenge in Liberal Democracies , eds Shweder R. A., Minow M., editors. (New York, NY: Russell Sage Foundation), 453 472. Marteau T.M. & Bekker H. The development of a six - item short - form of the state scale of the Spielberger State - Trait Anxiety Inventory (STAI). British Journal of Clinical Psychology. 1992; 31:301 - 306 May, D. R., Gilson, R. L., & Harter, L. M. (2004). The psychological conditions of meaningfulness, safety and availability and the engagement of the human spirit at work. Journal of Occupat ional and Organizational Psychology , 77 (1), 11 37. https://doi.org/10.1348/096317904322915892 McKay, P. F., Avery, D. R., & Morris, M. A. (2009). A Tale of Two Climates: Diversity Climate from Sub Performance. Personnel Psychology , 62 (4), 767 791. https://doi.org/10.1111/j.1744 - 6570.2009.01157.x Morrison, M. (2013). Becoming trustworthy white allies | Reflections. Retrieved May 3, 2018, from https://reflections.yale.edu/article/future - race/becoming - t rustworthy - white - allies 187 make things worse. Retrieved October 12, 2018, from https://thebodyisnotanapology.com/magazine/when - helping - turns - into - hovering - 6 - times - being - an - ally - can - make - things - worse/ Murphy, M., & Jones Taylor, V. (2012). The role of situational cues in signaling and maintaining stereotype threat. Stereotype Threat: Theory, Process, and Application . https://doi.org/10.1093/acprof:oso/9780199732449.003.0002 Murphy C. M., & Walton G. M., (2013). F rom prejudiced people to prejudiced places. In Stereotyping and Prejudice, ed. C Stangor, CS Crandall, pp. 181 203. New York: Psychol. Press. Nederhof, A. J. (1985). Methods of coping with social desirability bias: A review. European Journal of Social Psy chology , 15 (3), 263 280. https://doi.org/10.1002/ejsp.2420150303 Nishii, L. H. (2012). The benefits of climate for inclusion for gender - diverse groups. Academy of Management Journal , 56 (6), 1754 1774 . https://doi.org/10.5465/amj.2009.0823 Offermann, L. R., Basford, T. E., Graebner , R., Jaffer, S., De Graaf, S. B., & Kaminsky, S. E. (2014). See no evil: Color blindness and perceptions of subtle racial discrimination in the workplace. Cultural Diversity & Ethnic Minority Psychology , 20(4), 499 507. https://doi.org/10.1037/a0037237 Paulhus, D. L. (1988). Balanced inventory of desirable responding (BIDR). Acceptance and Commitment Therapy. Measures Package , 41 , 79586 7. Paulhus, Delroy L. (1991). Measurement and control of response bias. In Measures of Social Psychological Attitudes, Vol. 1. Measures of personality and social psychological attitudes (pp. 17 59). https:// doi.org/10.1016/B978 - 0 - 12 - 590241 - 0.50006 - X Patton, S. (2017, June 12). Beware of the so - https://www.damemagazine.com/20 17/06/12/beware - so - called - woke - allies/ Patrick, R. B., Bodine, A. J., Gibbs, J. C., & Basinger, K. S. (2018). What accounts for prosocial behavior? Roles of moral identity, moral judgment, and self - efficacy beliefs. The Journal of Genetic Psychology , 1 15 . https://doi.org/10.1080/00221325.2018.1491472 Pavey, L., Greitemeyer, T., & Sparks, P. (2011). Highlighting relatedness promotes prosocial motives and behavior. Personality & Social Psycholog y Bulletin , 37 (7), 905 917. https://doi.org/10.1177/0146167211405994 Pew Research Center (October, 188 Pettigrew, T. F., & Tropp, L. R. (2006). A meta - analytic test of intergroup contact theory. Journal of Personality and Social Psychology , 90 (5), 751 783. https://doi.org/10.1037/0022 - 3514.90.5.751 Phinney, J. S. (1992). The multigroup ethnic identity measure: A new scale for use with diverse groups. Journal of Adolescent Research , 7 (2), 156 176. https://doi.org/10.1177/074355489272003 Pinel, E. C. (1999). Stigma consciousness: the psychological legacy of social stereotypes. Journal of personality and social psychology , 76 (1), 114. Plant, E.A. and Devine, P.G. (1998) Internal and external motivation to respond without prejudice. Journal of Personality and Social Psychology , 75, 811 - 832. http://dx.doi.org/10.1037/0022 - 3514.75.3.811 Plaut, V. C., Thomas, K. M., Hurd, K., & Romano, C. A. (2018). Do color blindness and multiculturalism remedy or foster discrimination and racism ? Current Directions in Psychological Science , 27(3), 200 206. https://doi.org/10.1177/09 63721418766068 Podsakoff, P. M., & MacKenzie , S. B. (1997). Impact of organizational citizenship behavior on organizational performance: A review and suggestion for future research. Human Performance , 10 (2), 133 151. https://doi.org/10.1207/ s15327043hup1002_5 Post, T., & Walma van der Molen, J. H. (2019). Development and validation of a questionnaire questionnaire). Motivation and Emotion , 43 (1), 159 178 . https://doi.org/10.1007/s11031 - 018 - 9728 - 9 Pratto, F., Sidanius, J., Stallworth, L. M., & Malle, B. F. (1994). Social dominance orientation: A personality variable predicting social and political attitudes. Journal of Personality and Social Psychology, 67 (4), 741 - 763. http://dx.doi.org/10.1037/0022 - 3514.67.4.741 Purdie - Vaughns , V., Steele, C. M., Davies, P. G., Ditlmann, R., & Crosby, J. R. (2008). Social identity contingencies: How diversity cues signal threat or safety for African Americans in mainstream institutions. Journal of Personality and Social Psychology , 94 (4), 615 6 30. http://dx.doi.org.proxy1.cl.msu.edu/10.1037/0022 - 3514.94.4.615 Raine, A., & Uh, S. (2018). The selfishness questionnaire: Egocentric, adaptive, and pathological forms of se lfishness. Journal of Personality Assessment , 0 (0), 1 12. https://doi.org/10.1080/00223891.2018.1455692 Reason, R. D., Millar, E. A. R., & Scales, T. C. (2005). Toward a model of racial justice ally development. Journal of College Student Development, 46(5), 530 546. 189 October 11, 2018, from https://www.teenvogue.com/story/how - to - tell - if - youre - being - a - good - ally Shore, L. M., Cleveland, J. N., & Sanchez, D. (2018a). Inclusive workplaces: A review and model. Hu man Resource Management Review , 28 (2), 176 189. http://dx.doi.org.proxy2.cl.msu.edu/10.1016/j.hrmr.2017.07.003 October 5, 2018, from http://feministing.com/2013/10/01/the - case - against - allies/ Spielberger, C. D., Gorsuch, R. L., Lushene, R., Vagg, P. R., & Jacobs, G. A. (1983). Manual for t he State - Trait Anxiety Inventory . Palo Alto, CA: Consulting Psychologists Press. Steele, C. M., & Aronson, J. (1995). Stereotype threat and the intellectual test performance of African Americans. Journal of Personality and Social Psychology , 69 (5), 797 81 1. Steele D. M., & Cohn - Vargas B., (2013). Identity safe classrooms: Places to belong and learn. Thousand Oaks, CA: Corwin Steele C. M. 2010. Whistling Vivaldi: And Other Clues to How Stereotypes Affect Us . New York: Norton. Swann, W. B. (2012). Self - ve rification theory. In Handbook of theories of social psychology, Vol. 2 (pp. 23 42). Thousand Oaks, CA: Sage Publications Ltd. https://doi.org/10.4135/9781446249222.n27 Tonidandel, S., & LeBreton, J. M. (2015). RWA Web: A Free, Comprehensive, Web - Based, and User - Friendly Tool for Relative Weight Analyses. Journal of Business and Psychology , 30 (2), 207 216. https://doi.org/10.1007/s10869 - 014 - 9351 - z Torres, L., & Ong, A. D. (2010). A daily diary investigation of latino ethnic identity, discrimination, and depression. Cultural Diversity and Ethnic Minorit y Psychology , 16 (4), 561 568. https://doi.org/10.1037/a0020652 Threads of Solidarity: WOC AgainstRacism . (2017, July 5). Dear women of color who are told https://www .huffingtonpost.com/entry/dear - women - of - color - who - are - told - to - be - nicer - to - white_us_595d000be4b0f078efd98da3 Vazire, S. (2006). Informant reports: A cheap, fast, and easy method for personality assessment. Journal of Research in Personality , 40 (5), 472 481 . https://doi.org/10.1016/j.jrp.2005.03.003 Walton G. M. & Cohen G. L., (2007). A question of belonging: Race, social fit, and achievement . J. Personal. Soc. Psychol. 92:82 96 . 190 Walton, G. M., Murphy, M. C., & Ryan, A. M. (2015). Stereotype threat in organizations: Implications for equity and performance. Annual Review of Organizational Psychology and Organizational Behavior , 2 (1), 523 550. https://doi.org/10.1146/annurev - orgpsych - 032414 - 111322 Washington, J., & Evans, N. J. (1991). Becoming an ally. In N. J. Evans & V. A. Wall (Eds.), Beyond tolerance: Gays, lesbians, and bisexuals on campus (pp. 195 204). Was hington, DC: American College Personnel Association. 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 (6), 1063 1070. https://doi.org/10.1037/0022 - 3514.54.6.1063 Welbourne, T. M., Rolf, S., & Schlachter, S. (2017). The case for employee resource groups: A review and social identity theory - based research a genda. Personnel Review , 46 (8), 1816 1834. https://doi.org/10.1108/PR - 01 - 2016 - 0004 Weinstein, N., & Ryan, R. M. (2010). When helping helps: Autonomous motivation for prosocial behavior and its influe nce on well - being for the helper and recipient. Journal of Personality and Social Psychology , 98 (2), 222 244. https://doi.org/10.1037/a0016984 Wijeyesinghe , C. L., Griffin, P., & Love, B. (1997). Racism curriculum design. In M. Adams, L. A. Bell, & P. Griffin (Eds.), Teaching diversity and social justice: A sourcebook (pp. 82 107). New York, NY: Routledge. Williams, D. R., Yu, Y., Jackson, J. S., & Anderson , N. B. (2016). Racial differences in physical and mental health: Socio - economic status, stress and discrimination. Journal of Health Psychology . https://doi.org/10.1177/135910539700200305 Winters , M.F. (2014). From diversity to inclusion: An inclusion equation. In Diversity at Work: The Practice of Inclusion (pp. 205 228). Wiley - Blackwell. https://doi.org/10.1002/9781118764282.ch7 Winters, M. - F. (2015, November 5). Retaining top talent Part 3: Mentoring is key to retaining african american employees. Retrieved October 7, 2018, from http://www.theinclusionsolution.me/retaining - top - talent - part - 3 - mentoring - is - key - to - retaining - african - american - employees/ Wolf, E. J., Harrington, K. M., Clark, S. L., & Miller, M. W. (2013). Sample size requirements for structural equation models: An evaluation of power, bias, and solution propriety. Educational and Psychological Measurement , 73 (6), 913 934. https://doi.org/10.1177/0013164413495237 Yamagishi, T., & Yamagishi, M. (1994). Trust and commitment in the united states and japan. Motivation and Emotion , 18 , 129 166. https://doi.org/10.1007/BF02249397 191 Yip, T., Seaton, E. K., & Sellers, R. M. (2010). Interracial and intraracial contact, school - level diversity, and change in racial identity status among African American adolescents. Child Development , 81 (5), 1431 1444. https://doi.org/10.1111/j.1467 - 8624.2010.01483.x Yukl, G. (2012). Effective Leadership Behavior: What we know and what questions need more attention. Academy of Management Perspectives , 26 (4), 66 85. https://doi.org/10.5465/amp.2012.0088 Ziyad, H. (2 in order to show up. Retrieved October 11, 2018, from http://afropunk.com/2017/04/youre - not - really - an - ally - if - you - must - be - spoken - to - in - a - nice - tone - in - order - to - show - up/