Thesis on Organizational Climate

Description
Organizational climate (sometimes known as Corporate Climate) is the process of quantifying the “culture” of an organization, it precedes the notion of organizational culture.

ABSTRACT

Title of thesis:

THE EFFECT OF ORGANIZATIONAL CLIMATE ON THE ATTRIUBTION TO DISCRIMINATION PROCESS Lisa Michelle Leslie, Master of the Arts, 2004

Thesis Directed by:

Dr. Michele Gelfand Dr. Charles Stangor Department of Psychology

Research on the multi-stage attribution to discrimination process (construct accessibility, perceiving, and reporting of discrimination) focuses on individual difference antecedents and tends to examine one stage in each study (e.g. Major et al., 2002; Stangor, Sechrist, & Swim, 1999; Swim & Hyers, 1999). The current study extends research on this process by examining the interactive effect of individual differences and organizational climate on all three stages of the attribution to discrimination process in an organizational simulation study. Findings indicate that Climate for Intolerance for Discrimination interacts with individual based sensitivity to sexism to predict perceptions of discrimination. Furthermore, perceptions of discrimination fully mediate the relationship between the climate by sensitivity interaction and reporting of discrimination to the organization.

THE EFFECT OF ORGANIZATIONAL CLIMATE ON THE ATTRIBUTION TO DISCRIMINATION PROCESS

by Lisa Michelle Leslie

Thesis submitted to the Faculty of the Graduate School of the University of Maryland, College Park in partial fulfillment of the requirements for the degree of Master of the Arts 2004

Advisory Committee: Professor Michele Gelfand, Co-Chair Professor Charles Stangor, Co-Chair Professor Paul Hanges

ii ACKNOWLEDGEMENTS I would like to thank my advisor and co-chair, Dr. Michele Gelfand, for all of her advice on the project. Working with her has been an extremely rewarding experience. I would also like to thank my co-chair, Dr. Charles Stangor, and committee member, Dr. Paul Hanges, for all of their help. Their insightful comments and questions strengthened the quality of the final product.

iii TABLE OF CONTENTS List of Tables………………………………………………………………………... v List of Figures………………………………………………………………………. vi List of Appendices…………………………………………………………………. vii The Attribution to Discrimination Process………………………………………….. 3 Three Stage Model of the Attribution to Discrimination Process…………... 4 Individual Difference Antecedents………………………………………….. 5 Construct Accessibility…………………………………………….. ..5 Perceiving…………………………………………………………….6 Reporting……………………………………………………………. 6 Links between steps in the model…………………………………….7 Context Antecedents………………………………………………………… 8 Construct accessibility……………………………………………..... 8 Perceiving and Reporting…………………………………………… 8 Organizational Context Variables……………………………………………………8 Organizational Climate………………………………………………………….. .. 11 Climate for Diversity………………………………………………………. 11 Climate for Intolerance for Discrimination…………………………………12 Formal Aspects…………………………………………………….. 13 Informal Aspects……………………………………………………14 The Effect of Organizational Variables on Individual Processes………….. 15 Hypotheses………………………………………………………………………….17 Method……………………………………………………………………………... 22 Sample………………………………………………………………………22 Procedure…………………………………………………………………... 22 Pre-test measures…………………………………………………... 22 Cover Story………………………………………………………… 23 Experimental Manipulation………………………………………... 24 Measure of Construct Accessibility………………………………... 26 Filler Application Materials………………………………………...27 Ambiguous Negative Event………………………………………... 27 Measurement of Perceiving………………………………………... 28 Measurements of Reporting………………………………………...28 Manipulation Checks………………………………………………. 29 Pilot Testing………………………………………………………………... 29 Results………………………………………………………………………………31 Suspicion and Manipulation Checks………………………………………..31 Deletion of Participants Due to Suspicion…………………………. 31

iv Manipulation Checks………………………………………………. 32 Descriptives…………………………………………………………………33 Individual Difference Variables…………………………………….33 Climate for Intolerance for Discrimination…………………………36 Construct Accessibility Data………………………………………..36 Perceiving………………………………………………………….. 38 Reporting……………………………………………………………38 Correlations among the Variables…………………………………..39 Analyses at Each of the Three Attribution to Discrimination Stages ……... 39 Stage 1: Construct Accessibility as a Dependent Variable………… 39 Stage 2: Perceiving as a Dependent Variable……………………… 39 Stage 3: Reporting as a Dependent Variable………………………. 41 Mediation…………………………………………………………... 42 Discussion………………………………………………………………………….. 44 Implications for Organizations…………………………………………….. 47 Limitations…………………………………………………………………. 48 Future directions…………………………………………………………… 53 Conclusions…………………………………………………………………54

v LIST OF TABLES

Table 1:

The Dimensions of Climate for Diversity and Climate for Inclusion………………………………………….. 56 Factor Loadings for Sensitivity to Sexism (SS)…………………… 57 Factor Loadings for Individual Sensitivity to Sexism (ISS)………. 57 Factor Loadings for Gender Identity (GI)…………………………. 57 Factor Loadings for Past Experiences with Gender Discrimination………………………………………...57 Correlations among the Variables…………………………………. 58 Hierarchical Regression Results…………………………………… 59 Hierarchical Logistics Regression…………………………………. 60

Table 2a: Table 2b: Table 2c: Table 2d:

Table 3: Table 4: Table 5:

vi LIST OF FIGURES

Figure 1:

The Effect of Climate for Intolerance for Discrimination on the Attribution to Discrimination Process………………………. 61 The Effect of the Individual Sensitivity to Sexism by Climate for Intolerance for Discrimination Interaction on Perceiving and Reporting of Discrimination…………………… 62 Revised Model of the Effect of Individual Sensitivity to Sexism and climate for Intolerance for Discrimination on Perceiving and Reporting of Discrimination…………………… 63 A Comprehensive Model of Organizational Antecedents to the Attribution to Discrimination Process………………………. 64

Figure 2:

Figure 3:

Figure 4:

vii LIST OF APPENDICES

APPENDIX A. Pre-measures………………………………………………............ 65 APPENDIX B. Climate for Intolerance for Discrimination Manipulation and Fillers………………………………………….. 68 APPENDIX C. List of Words for Construct Accessibility Measure………………. 76 APPENDIX D. Application………………………………………………………... 77 APPENDIX E. Pre-feedback Questionnaire………………………………………..82 APPENDIX F. Feedback Form……………………………………………………..83 APPENDIX G. Measurement of Perceiving………………………………………..84 APPENDIX H. Measurement of Reporting………………………………………...85 APPENDIX I. Manipulation/Suspicion Check……………………………………..88

1 The Effect of Organizational Climate on the Attribution to Discrimination Process Forty years after the Civil Rights Act of 1964 deemed it illegal for organizations “to discriminate against any individual because of his race, color, religion, sex, or national origin” (Title VII), the goal of equal opportunity remains unachieved. While the blatant discrimination plaguing the country during the first part of the 20th century no longer persists, research suggests that privately held discriminatory attitudes remain and result in subtle forms of discrimination (e.g. Dovidio & Gaertner, 1998). Thus, in spite of changes in reported attitudes, evidence indicates that subtle discrimination is still a problem. An abundance of research supports the notion that manifestations of discrimination have shifted from overt to covert (e.g. Brief, Dietz, Cohen, Pugh, & Vaslow, 2000; Crosby, Bromley & Saxe, 1980; Dovidio, & Gaertner, 1998; Dovidio, Kawakami & Gaertner, 2002; James, Brief, Dietz, & Cohen, 2001; Swim, Aikin, Hall, & Hunter, 2002). Due to this change, members of minority groups are faced with ambiguity when interpreting negative events. Although an employer cannot claim gender or racial group membership as the reason why an individual is fired from a job or passed over for a promotion, this does not prove that group membership does not influence organizational decisions. Because current norms governing behavior toward protected groups (e.g. racio-ethnic minorities, women, elderly, disabled, etc) discourage blatant discrimination, protected group members face increasing difficulties when judging the motivation for a behavior. The body of research on the antecedents and consequences of making attributions to discrimination and acting based on those attributions addresses this dilemma (e.g. Crocker, Voelkl, Testa, & Major, 1991; Friedman & Davison, 1999; Major

2 et al., 2002; Sechrist, & Stangor, 2001; Stangor, Swim, Sechrist, Van Allen, & Ottenbreit, in press). The attribution to discrimination process literature focuses on perceptions and behavior in ambiguous situations. Specifically, this research examines the antecedents to and consequences of attributing a negative event to discrimination when the negative event could also be attributed to an internal characteristic. Due to the potential for negative events with attribution ambiguity to occur in work contexts (e.g. hiring decisions, promotion decisions, informal networking), organizations present an appropriate milieu for assessing continuing discrimination. The majority of research on the attribution to discrimination process, however, lacks consideration of organizational variables. Therefore, I proposed a model that brings the attribution to discrimination process into the organizational context by predicting the relationship between an organizational level variable and Stangor et al.’s (in press) three-stage attribution to discrimination process. Specifically, I use a laboratory-based organizational simulation to assess the influence of Climate for Intolerance for Discrimination, in addition to individual difference variables, on each of the three stages involved in the attribution to gender discrimination process (see Figure 1). Developing a model of the interactive effect of climate and individual differences on the attribution to discrimination process necessitates knowledge of several areas of psychological research. I begin with a review of the literature on the attribution to discrimination process. Next, I discuss the current status of research on climate as it relates to diversity issues and present the construct of Climate for Intolerance for

3 Discrimination. Finally, I formulate hypotheses based on the integration of these two areas of research. The Attribution to Discrimination Process Research on the attribution to discrimination process focuses on predicting under what circumstances an individual will or will not attribute an ambiguous negative event to discrimination. The common paradigm involves bringing participants, who are members of protected groups, into the laboratory, asking them to complete a task (e.g. write an essay), and having a confederate give them negative feedback on that task (e.g. Crocker, Voelkl, Testa, & Major, 1991; Major et al., 2002; Stangor, Swim, VanAllen & Sechrist, 2002). Participants know that their group membership information is provided to the confederate before the confederate makes judgments about the participants’ work. After receiving negative feedback, participants are asked whether the feedback resulted from an internal source (e.g. a poorly written essay) or discrimination on the part of the confederate. Variations of this paradigm also exist in the literature, such as exposing participants to a negative comment made by a confederate instead of giving participants negative feedback (e.g. Swim & Hyers, 1999). Using the paradigm described above researchers have identified a variety of factors that influence whether an individual will make an attribution to discrimination or an internal attribution, although most variables examined are individual level and nonorganizational. Stangor et al. (in press) propose a three stage model for studying the attribution to discrimination process that provides a framework for organizing the literature. Therefore, I begin the review of research on the attribution to discrimination process with an overview of the three stage model.

4 Three Stage Model of the Attribution to Discrimination Process The Stangor et al. (in press) model subdivides the attribution process into the three stages of asking (construct accessibility), answering (perceiving) and announcing (reporting/confronting). Outcomes during the asking stage are determined by the construct accessibility of discrimination at the time of an ambiguous negative event. The construct accessibility of discrimination will determine whether or not an individual will become suspicious that an ambiguous negative event constitutes discrimination. The second stage of the model, answering, involves labeling the negative event as discriminatory or perceiving discrimination. Stangor et al. propose that this occurs when an individual has the motivation and ability to process an ambiguous negative event. Perceiving discrimination occurs when an individual decides that an ambiguous negative event constitutes discrimination. The third stage, announcing, involves reporting that discrimination has occurred or confronting the perpetrator. Each stage in the model is contingent on the previous stage. An individual cannot perceive discrimination without first being suspicious about an ambiguous negative event. Similarly an individual cannot report discrimination without first perceiving an event as discriminatory. Furthermore, an individual will not necessarily process an ambiguous negative event at each stage of the model. If an individual has initial suspicion about an event but subsequently decides the event is not discriminatory, that individual will not reach the reporting stage of the model. Although no single study examines all three stages, researchers have investigated the antecedents to the attribution to discrimination process at all stages. Stangor et al. (in press) note that both individual difference and contextual variables can influence each

5 stage. In the case of construct accessibility, although discrimination may be chronically accessible to some individuals and not to others, the situation can also influence accessibility. For example, the presence or absence of ingroup versus outgroup members may affect the salience of group membership, which in turn affects construct accessibility (Cohen & Swim, 1995). Stangor et al. argue that individual differences and the context similarly influence the latter two stages of the model. I review the influence of individual difference and contextual antecedents on the attribution to discrimination process in the following section. Research on these antecedent variables assesses members of a variety of protected groups including ethnic minorities, women, and homosexuals. Although the current study investigates the attribution to gender discrimination process, I include research on the attribution to discrimination process based on membership in any protected group. Individual Difference Antecedents Construct accessibility. Research supports the influence of individual difference variables on each of Stangor et al.’s (in press) three stages in the attribution process. Although construct accessibility is the least studied stage, Stangor et al. (1999) isolated an individual level antecedent to construct accessibility. They measured sensitivity to sexism among women and then exposed the women to a series of headlines. One-fourth of the headlines dealt with sexism, and the remaining three-fourths did not. They found that women who scored high on the sensitive to sexism measure overestimated the number of sexist headlines and women who scored low on the sensitive to sexism measure underestimated the number of sexist headlines. If estimates of the number of sexist headlines are considered a measure of construct accessibility, this study suggests

6 that individual differences in sensitivity to sexism do influence construct accessibility of gender discrimination. Perceiving. Similarly, research suggests individual difference variables influence the perceiving stage of the attribution to discrimination process. In a laboratory study Major et al. (2002) studied the influence of group status and ethnic group identity on perceiving discrimination. Groups were dichotomized as either high status (EuropeanAmericans) or low status (African- and Latino-Americans). They measured ethnic identity with four items assessing identification with other ingroup members, the importance of ethnicity to individual identity, thinking of oneself as belonging to a particular ethnicity, and closeness to other ingroup members. They found that Africanand Latino-Americans who identified strongly with their ingroups were more likely to perceive discrimination than those who did not identify strongly with their ingroups. Reporting. Swim and Hyers (1999) found that individual differences influence reporting of discrimination. In a laboratory study they measured women’s experiences with confronting gender discrimination and then exposed the women to sexist comments made by a male confederate. Those participants with a history of experiences with gender discrimination were more likely to confront the confederate than those participants without a history of experiences with gender discrimination. In sum, the influence of sensitivity to discrimination, group status, ingroup identification and past experiences with discrimination on construct accessibility, perceiving, and reporting of discrimination substantiates the importance of individual differences. Several of the individual difference variables found to influence the attribution to discrimination process are associated with group membership. In general,

7 members of protected groups (or low status groups) are more likely than members of non-protected groups to attribute an ambiguous negative event to discrimination and act on the attribution (Major et al., 2002). Evidence that members of protected groups are more likely to make and act on an attribution to discrimination than members of nonprotected groups follows logically from the premise that discrimination occurs to members of protected groups. However, research also indicates that group membership alone does not predict outcomes of the attribution to discrimination process (Major et al., 2002; Stangor et al., 1999; Swim & Hyers, 1999). Individuals within protected groups vary according to likelihood of making an attribution to discrimination and acting on that attribution. Therefore, because sensitivity to discrimination, ingroup identification and past experiences with discrimination explain variation within protected groups, these variables merit inclusion in a model examining the interactive effect of individual differences and climate on the attribution to discrimination process. Links between steps in the model. Research on individual differences and construct accessibility, perceptions, and reporting of discrimination tends to focus on the influence of antecedents to one of the three stages. One study, however, did assess the relationship between perceiving and reporting of discrimination. Swim and Hyers (1999) showed that even when an individual perceived an event as discriminatory, the perceiver did not always act on that perception. They found that three-quarters of women who did not actively confront sexist comments still perceived the comments as discriminatory. Furthermore, a higher percentage of women reported they would confront sexist comments, than the percentage of women who did confront sexist remarks in a laboratory study. Therefore, Swim and Hyers (1999) provided initial support for contrasting

8 outcomes at different steps in the three stage model. This research, however, ignored the influence of organizationally relevant variables on the attribution to discrimination process and did not consider the role of construct accessibility. Context Antecedents Construct Accessibility. In addition to individual level factors, laboratory research suggests that context can influence the attribution to discrimination process. Cohen and Swim (1995) assessed construct accessibility of discrimination, measured via self-report questions about the likelihood that an individual would be the victim of discrimination, among protected group members. They manipulated whether each participant was a token (a single protected group member in the presence of only outgroup members) or a nontoken (a protected group members in presence of other ingroup members). Cohen and Swim found that token participants were significantly more likely to expect to be stereotyped than nontoken participants. Perceiving and Reporting. The presence of outgroup versus ingroup members also influences perceiving discrimination. Stangor et al. (2002) showed that minority group members were more likely to perceive an ambiguous negative event as discriminatory in private or in the company of other minority group members than in the presence majority group members. Additionally, Swim and Hyers (1999) found an effect of context on reporting discrimination. Specifically, women were more likely to confront a man who made a sexist remark when they were the only woman present than when other women were present. Organizational context variables. Although research on contextual antecedents to the attribution to discrimination process focuses on the presence or absence of other

9 ingroup members, organizational contexts may also influence the attribution to discrimination process. Stangor et al. (in press) note “organizational climates in which sexism is a frequent topic of conversation among individuals is likely to produce an overall increase in asking” (p. 10). No study has directly investigated the role of organizational climate in the attribution to discrimination process. Two studies do assess the influence of individual perceptions of organizational context on the attribution process (Ragins & Cornwell, 2001; Goldman, 2001). Ragins and Cornwell measured the impact of perceptions of organizational characteristics on perceiving discrimination among homosexuals and bisexuals by administering surveys to members of several gay rights groups. They found perceptions of organizational policies and practices influenced perceptions of discrimination such that individuals who perceived supportive policies and practices in their organizations also perceived less discrimination than individuals who did not. Similarly, Goldman (2001) assessed factors influencing individuals’ decisions to file a discrimination claim with the Equal Employment Opportunity Commission (EEOC) after termination from their jobs. Goldman found individual perceptions of organizational justice to be negatively correlated with individual reporting of discrimination to the EEOC. The organizational literature on the attribution to discrimination process is limited in several ways. Although organizational justice and organizational policies and practices can be conceptualized as organizational variables, Ragins and Cornwell (2001) and Goldman (2001) measured individual perceptions of these variables, making the constructs individual level variables instead of organization level variables. For the

10 constructs to be considered organizational level variables, perceptions must be shared among employees. These studies used employees from many organizations and therefore could not assess whether or not the constructs were shared within a particular organization. Furthermore, the findings in both studies are limited by single source bias. The researchers conducting these studies collected data on perceptions of organizational variables and perceptions/reporting of discrimination at the same time and from the same participants. The collection of data on two variables at the same time and from the same participant creates a confound. It is unclear whether the relationships result from a true association or from another factor influencing the participants at the time when they completed the survey. Although it is possible to avoid single source bias in field research (e.g. measure the independent and dependent variables at different times), such measures were not used in these two studies. Finally, as with research on individual difference variables, research on contextual variables fails to examine the influence of a variable on all phases of the three-stage model within a single study and fails to explore the interaction between organizational context and individual difference variables. In conclusion, a review of the literature on the attribution to discrimination process reveals several gaps, especially when applying this research to organizations. In the current study, I begin to fill these gaps by investigating the influence of climate, in addition to individual difference variables, on the attribution to discrimination process. This study’s contribution is two-fold. First, because the literature currently lacks a test of the complete model, I examine the role of antecedents on all three stages of the attribution to discrimination process within a single study. Second, instead of focusing solely on individual difference antecedents, I explore the interaction between individual

11 difference and contextual antecedents. An explanation of the organizational variable of interest, Climate for Intolerance for Discrimination, and its appropriateness to the present study, follows. Organizational Climate Before defining Climate for Intolerance for Discrimination a review of organizational climate is necessary. Schneider and Reichers (1983) define organizational climate as shared employee perceptions of organizational policies, practices, and procedures. Climate is shared among employees and therefore defined by how employees collectively experience the organization. Due to the broad nature of organizational climate, a literature has developed on specific climates within organizations (Schneider, 1975). Organizations have targeted climates such as a climate for safety, sexual harassment, well-being, service, and diversity (Barak, Cherin & Berkman, 2000; Burke, Borucki, & Hurley, 1992; Fitzgerald, Drasgow, Hulin, Gelfand, & Magley, 1997; Schneider, Bowen, Ehrhart, & Holcombe, 2000; Zohar, 1980). The goal of the present study is to determine how organizational climate influences the attribution to discrimination process among members of protected groups. In searching for an aspect of organizational climate relevant to the attribution to discrimination process, climate for diversity is the obvious place to begin because climate for diversity deals with the treatment of members of protected groups within an organization. Climate for Diversity Over the last 10 years, a literature on climate for diversity and other related constructs has developed (e.g. Barak et al., 1998; Cox, 1993; Hicks-Clark & Iles, 2000;

12 Kossek & Zonia, 1993; Nishii & Raver, 2001; Pelled, Ledford, & Mohrman, 1999). However, no clear definition or well-accepted measure of climate for diversity exists in the literature. While research on climate for diversity generally defines the construct in terms of the shared experiences of protected groups within an organization, conceptualizations vary in terms of the content and labels of the factors that constitute climate for diversity. Moreover, related climates such as climate for inclusion exist in the literature. Climate for inclusion differs from climate for diversity by focusing on the informal shared experiences of protected groups in an organization. To provide some clarity to the literature, I grouped the dimensions of measures of climate for diversity and inclusion into three categories including formal aspects, informal aspects, and value placed on diversity (see Table 1). In my conceptualization, formal organizational aspects represent the extent to which employees’ shared perceptions of formal, or regulated, practices reflect fair or unfair treatment of protected groups in an organization. Informal organizational aspects represent the extent to which employees’ shared perceptions of informal, or unregulated, practices reflect fair or unfair treatment of protected groups in an organization. Finally, value placed on diversity represents the extent to which an organization appreciates, values, and active seeks to employ a diverse workforce. Climate for Intolerance for Discrimination The climate most likely to influence the attribution to discrimination process is a climate concerning organizational policies, practices, and procedures that address discrimination. While such a climate includes the shared experiences of protected groups in an organization and is therefore part of climate for diversity, climate for diversity is

13 broader than a climate for organizational experiences with discrimination. Because discrimination can occur formally (e.g. hiring decisions) and informally (e.g. inclusion in social networks) I conclude that the dimensions of formal and informal organizational aspects constitute a climate for experiences with discrimination in an organization. Furthermore, I labeled this new construct the Climate for Intolerance for Discrimination (CID). Value placed on diversity is not included in CID because an organization can be free of discrimination while not actively seeking to recruit diverse employees and valuing a diverse workforce. The definition of CID as formal and informal organizational aspects falls within Schneider and Reichers’ (1983) definition of climate as shared perceptions of organizational policies, practices and procedures. For example, a company’s protocol for hiring, which is a formal aspect of organizations, exemplifies an organizational procedure. Excluding certain groups from social interactions within a company, which is an informal aspect of organizations, exemplifies an organizational practice. A more detailed description of the two factors included in organizational CID follows. Formal aspects. The formal aspects dimension of CID is conceptualized as the shared perceptions of the extent to which regulated organizational policies, practices, and procedures result in equitable versus inequitable outcomes for protected group members. Formal organizational aspects of CID include hiring, promotion, allocation of resources, and determination of salary because these outcomes are at least partially monitored by organizations. Kossek and Zonia (1993), Hicks-Clarke and Iles (2000), Barak et al. (1998), and Nishii and Raver (2001) include formal aspects in their conceptualizations of climate for diversity. For example, Nishii and Raver provide an example of the formal

14 aspects dimension of Climate for Intolerance to Discrimination with the following item: “In this organization, promotions are influenced by people’s group identities.” Informal Aspects. Informal aspects of CID are conceptualized as shared perceptions of the extent to which unregulated organizational policies, practices, and procedures result in equitable versus inequitable outcomes for protected group members. More generally, the informal aspects of CID are defined as the extent to which employees perceive an organization as including or excluding members of protected groups in unregulated organizational interactions. Examples of informal aspects include access to the advice of coworkers and integration into social networks. Like formal aspects, informal aspects of CID have negative work outcomes in the form of fewer unregulated resources. Pelled et al., (1999), Barak et al. (1998), and Nishii and Raver (2001) include informal aspects in their conceptualizations of climate for inclusion and diversity. For example, Nishii and Raver provide an example of the informal aspects dimension of Climate for Intolerance to Discrimination with the following item: “In general, individuals in positions of power only help others whom they perceive to be similar to them.” The two dimensions of formal and informal organizational aspects constitute Climate for Intolerance for Discrimination. Within a particular organization the valence of the formal and informal dimensions can be in the same or opposite directions. For example, organizations can potentially have formal policies and practices that treat minorities positively (e.g. fair hiring and promotion processes) and at the same time have informal policies and practices that treat minorities negatively (e.g. informally exclude minorities from informal social networks). Therefore, the separation of Climate for

15 Intolerance to Discrimination into two dimensions is necessary. In the current study, however, I only examine organizations with both positive and negative formal and informal organizational aspects. The Effect of Organizational Variables on Individual Processes The present study assesses the influence of an organizational variable, CID, on an individual outcome, the attribution to discrimination process. Although few examples exist, levels research suggests that higher level variables can influence individual level outcomes. For example, Schneider, Smith and Sipe (2000) note the importance of examining cross-level direct effects relationships to determine which organizational level variables influence individual level performance. Perry, Davis-Blake, and Kulik (1994) also provide an example of how organizational level variables may influence individual processes. They explore the role of both organizational context and cognitive factors in gender segregation, the differential distribution of men and women across jobs. In addition to contextual and cognitive factors separately affecting gender segregation, Perry et al. propose that contextual factors influence the cognitive processes used in making hiring decisions. The main cognitive variable discussed is the content and activation of job schemata, prototypes for the person appropriate for a particular job, held by employees making hiring decisions. Perry et al. propose that context variables including organizational demography, organizational structure, and the power of human resources within an organization may affect job schemata. For example, they propose that controlling for organization size the job schemata of decision makers in organizations with many job titles are more likely to contain a gender component than the job schemata of decision makers in organizations

16 with few job titles. The more job titles a company has, fewer individuals that occupy each job. When a small number of individuals hold a particular job, this increases the specificity of the job schema an individual will form for that particular job and then use in making hiring decisions. For example, if only one person occupies a particular job, the job schema may include the gender of the individual in the job. If many individuals of both genders occupy a particular job, the schema for that job is less likely to include gender. In this way, organizational structure influences an individual level cognitive process. Several aspects of Perry et al.’s (1994) theory parallel the current study. First, the notion that high level organizational variables such as organizational demography, organizational structure and the power of human resources affect individual cognition, is consistent with the hypothesis that CID will influence the attribution process. Like Perry et al.’s context variables, CID is an organization level variable capable of being measured in terms of the shared perceptions of employees. Similarly, like the discussion of schema content and activation, the attribution to discrimination process occurs individually and has cognitive components. Although the contexts differ, gender segregation versus discrimination, this research bolsters the argument that organization level variables will affect individual level processes. In sum, the current study has several implications for the literature on organizational climate. First, I introduce the new construct of Climate for Intolerance for Discrimination, a narrow and well-defined subdivision of climate for diversity. Second, investigating the influence of climate on the attribution to discrimination process contributes to levels research by showing the ability of an organizational level variable to

17 influence individual level processes. Finally, this study advances the literature by studying the influence of climate at the organizational level, instead of measuring individual perceptions of climate. Hypotheses Although the primary goal of this study is to assess the influence of an organizational context variable, CID, on the attribution to discrimination process the role of individual difference variables remains relevant. In the absence of contextual antecedents, individual difference variables will influence the attribution to discrimination process (Major et al., 2002; Stangor et al., 1999; Swim & Hyers, 1991). The meaningful individual difference variables isolated in the literature include group status, sensitivity to discrimination, ingroup identity, and past experiences with discrimination (Major et al., 2002; Stangor et al., 1999; Swim & Hyers, 1999). The present study assesses the attribution to discrimination process among protected group members, specifically women. Therefore all participants included in the study belong to the same low status groups, eliminating variance due to group status. The remaining three variables are all relevant. Due to the use of women as subjects, the current study will specifically assess sensitivity to sexism, gender identity, and past experiences with gender discrimination. Sensitivity to sexism is defined as the extent to which women are consciously aware of and concerned with potentially being the victim of discrimination (Stangor et al., 1999). Gender identity is defined as the extent to which being female is important to the identity of women (Luhtanen & Crocker, 1992). Finally, past experiences with gender discrimination is defined as the extent to which women have been the target of discrimination in the past (Schmitt, Branscombe, Kobrynowicz, &

18 Owen, 2002). I hypothesize that these individual difference variables will directly influence construct accessibility, perceiving and reporting of discrimination. Hypothesis 1a-c (H1a-c): (a) Sensitivity to sexism, (b) gender identity, and (c) past experiences with discrimination will influence construct accessibility such that the gender discrimination construct will be more accessible to women with high scores on these variables than it will be to women with low scores on these variables. Hypothesis 2a-c (H2a-c): (a) Sensitivity to sexism, (b) gender identity, and (c) past experiences with discrimination will influence perceptions of discrimination such that women with high scores on these variables will be more likely to perceive gender discrimination, than women with low scores on these variables. Hypothesis 3a-c (H3a-c): (a) Sensitivity to gender discrimination, (b) gender identity, and (c) past experiences with gender discrimination will influence reporting of gender discrimination such that women with high scores on these variables will be more likely to report discrimination than to women with low scores on these variables. Stangor et al. (in press) indicate that context and individual differences influence construct accessibility of discrimination, perceptions of discrimination, and reporting of discrimination. Therefore, I predict the contextual organizational variable of CID will influence the attribution to discrimination process. I predict people who work in an environment with a positive CID will have less suspicion of discrimination than people working in an environment with a negative CID. In organizations that promote fair treatment of people regardless of gender or ethnic background, employees will have less

19 concern that they may become a target of discrimination than employees who work in an organization that does not promote just treatment regardless of background. Therefore, I hypothesize that a positive CID will decrease suspicion of discrimination. Hypothesis 4 (H4): Controlling for individual differences, the construct of gender discrimination will be more accessible for women in a negative CID than for women in a positive CID. Similarly, individuals who work in a positive CID are less likely to perceive discrimination as the cause of negative outcomes than individuals who work in a negative CID. If an individual experiences a negative event in an organization with policies, practices, and procedures that do not tolerate discrimination that individual will be less likely to interpret the negative event in a way that contradicts the aim of the policies, practices, and procedures. Hypothesis 5 (H5): Controlling for individual differences, women will be more likely to perceive an ambiguous negative event as discriminatory in a negative CID than in a positive CID. As noted, each of the steps in the three stage model is dependent on the previous stage. Therefore, in addition to having a direct influence on perceiving discrimination, CID will also affect perceiving discrimination through the relationship between CID and construct accessibility of discrimination. Hypothesis 6 (H6): Construct accessibility will mediate the relationship between CID and perceiving discrimination I also hypothesize that CID will influence reporting of discrimination. I predict that the influence of CID on reporting discrimination to the organization in which it

20 occurred will be in the opposite of the influence of CID on construct accessibility and perceiving. In organizations that promote fair treatment individuals are more likely to feel that, if reported, their concerns will be taken seriously by the organization. In organizations where unjust treatment is the norm, however, employees will see reporting instances of discrimination as futile. Hypothesis 7 (H7): Controlling for individual differences, women will be more likely to report gender discrimination in a positive CID than in a negative CID. Similar to Hypothesis 6, construct accessibility and perceiving will also influence the relationship between climate and reporting. Hypothesis 8 (H8): Construct accessibility and perceiving will mediate the relationship between CID and reporting of discrimination. In addition to the direct influence of individual difference variables and CID on the three stages involved in the attribution to discrimination process, I predict individual differences will interact with CID. For example, I expect CID to have more of an effect on construct accessibility and perceiving of discrimination for individuals with low gender identification as compared to individuals with high gender identification. Individuals with high gender identification will be likely to have high construct accessibility and perceive an ambiguous negative event as discriminatory regardless of the CID. Although individuals with low gender identification may still have high construct accessibility and perceive negative events as discriminatory in a negative CID, low gender identification individuals will be less likely to have high construct accessibility and perceive discrimination in a positive CID than individuals with high gender identification will be. Stated another way, I predict individuals high and low on

21 gender identity to have high construct accessibility and perceive discrimination in a negative CID. In a positive CID, however, only those individuals with high gender identity will have high construct accessibility and perceive discrimination. I expect sensitivity to sexism and past experiences with gender discrimination to interact with CID in the same way. Hypothesis 9a-9c (H9a-H9c): CID will interact with (a) sensitivity to discrimination, (b) gender identity, and (c) past experiences with discrimination to predict construct accessibility. Hypothesis 10a-10c (H10a-H10c): CID will interact with (a) sensitivity to discrimination, (b) gender identity, and (c) past experiences with discrimination to predict perceiving. Similarly, in the presence of a negative CID, I expect no difference in reporting of discrimination between individuals who are high versus low on the individual difference variables. For example in a negative CID, regardless of gender identity, individuals will not feel comfortable reporting a discriminatory event. In a positive CID, however, I predict that individuals with high gender identification will report discrimination but individuals with low gender identity will not. Furthermore, a similar relationship will hold for sensitivity to sexism and past experiences with gender discrimination. Hypothesis 11a-11c (H11a-H11c): CID will interact with (a) sensitivity to discrimination, (b) gender identity, and (c) past experiences with discrimination to predict reporting.

22 Method Sample I recruited a total of 141 female undergraduate psychology students from a large Mid-Atlantic university to participate in a laboratory study during the fall 2003 semester. The majority of the students (N = 115) were enrolled in an undergraduate introductory psychology course, and the remaining students (N = 26) were enrolled in an upper level undergraduate psychology course. All participants received course extra credit for participation in the study. Of the 141 participants, 122 (86.5%) were Caucasian, 7 (5.0%) were Biracial, 6 (4.3%) were Hispanic, 1 (.7%) was Native American, 0 were African American or Asian, and 5 (3.5%) listed their race as “other.” Procedure Pre-test measures. I measured the individual difference variables of interest during mass testing at the beginning of the fall of 2003 semester (see Appendix A). The participants enrolled in the introductory psychology course completed the pre-measures as part of a packet of questionnaires containing measures from many researchers within the psychology department. The participants enrolled in the upper level psychology course completed only the pre-measures used in this study. I collected the pre-measures in the first two weeks of the semester, before anyone participated in the study. Furthermore, participants were given no indication that the pre-measures were connected with the laboratory study. The pre-measures included demographic information in addition to the three individual difference variables of interest. I measured sensitivity to gender discrimination with the sensitivity to sexism measure from Stangor et al. (1999); history with

23 discrimination with Schmitt, et al.’s (2002) past experiences with gender discrimination scale; and gender identity with the importance to identity subscale of Luhtanen and Crocker’s (1992) collective self-esteem scale. Each of the scales has been used previously in research on the attribution to discrimination process and demonstrated internal reliability of = .77, = .82, and = .79, respectively (Major et al., 2002;

Schmitt et al., 2002; Stangor et al., 1999) Cover Story. I based the methodology for this study on the procedure used in Study 2 of Major et al. (2002). The procedure was altered and expanded to include examination of organizationally relevant variables. Participants signed up to participate in an organizational simulation study focusing on how decisions are made in organizations. The experimenter brought each participant to one of three rooms with a computer station. Some of the participants (62%) were guided through the study by a female undergraduate experimenter, and the remaining participants (38%) were guided through the study by a male undergraduate experimenter. In each session of the study, the experimenter told each participant to assume the role of an associate at RLK Consulting and conveyed that although the name had been changed, all information received about RLK Consulting was based on an actual company. The experimenter told each participant that she would be participating in a group of three participants. Furthermore, one of the other participants had been randomly assigned the role of manager. The manager had the task of deciding which of the remaining two participants would be given a promotion to the role of co-manger. The participant not receiving the promotion would remain in the role of associate. The manager would determine who would be given the promotion to the role of co-manager

24 based on performance on an application. In reality, no participant was assigned the role of manager and instead a male confederate played the role of manager. Furthermore, each participant competed with a fictitious student for the promotion to co-manager. While providing an oral description of the study, the experimenter portrayed the role of co-manager as more favorable than the role of associate. The experimenter told each participant that the manager and co-manager would work together on solving a series of organizational dilemmas. The associate would work alone and make a series of decisions relating to mundane topics such as choosing among a number of different of benefits packages. The experimenter described the co-manager’s task as exciting, and the associate’s task as dull. Furthermore, the experimenter told each participant that the comanager would learn managerial skills that may be useful after graduation, while the associate would not learn such skills. To increase psychological fidelity, the participants were told that the co-manager would be entered in a lottery with the chance of winning $100, but the associate would not be. In reality, all participants were entered in the lottery. Experimental manipulation. The experimental manipulation was embedded in the first task and indicated that RLK Consulting had a positive or negative CID. I used previous research that has successfully manipulated organizational climate to design the climate manipulation (Chatman, Polzer, Barsade, & Neale, 1998; Mannix, Neale, & Northcraft, 1995). I operationalized positive CID as an organization in which shared perceptions of both formal and informal policies, practices, and procedures promote fair treatment of protected group members. I operationalized negative CID as an organization in which shared perceptions of both formal and informal polices, practices and

25 procedures promote unfair treatment of protected group members. The manipulation was administered between subjects. The experimenter instructed participants that the first task was to become familiar with the company. Each participant received a packet of information about RLK Consulting. The experimenter instructed each participant to read the packet of information carefully to gain an accurate perception of the company. The information about the company in the packet came from The Insider Scoop, an (actually fictitious) website that provides information about various companies to individuals looking for jobs. Therefore, the information ostensibly came from a source motivated to portray RLK Consulting objectively. The packets included seven items, two of which manipulated CID. One item manipulated the formal aspects of CID through a list of top employees at RLK Consulting. In the positive condition women held half of the top positions in the company. In the negative condition men held all of the top positions in the company. A second item manipulated the informal aspects of CID. A list of quotes about the company made by employees was provided. In the positive condition one quote indicated that women feel they have equal power to men in the company. In the negative condition one quote from a female employee referred to the organization as an “old boys' network.” The critical quote in each condition was embedded in a longer list of quotes not related to CID. The remaining items in the packet that did not manipulate CID included the average salary earned in each job within the company, a statement about the company, a summary of the profits earned by the company over the last 10 years, and a list of the size of each department within the company (see Appendix B).

26 Measure of Construct Accessibility. After reading the information packet, the participants received an application packet. The experimenter explained that the application packet contained three parts. The first part of the application was created using the program Inquisit. The task, administered on a computer, consisted of a lexical decision task requiring participants to decide whether or not a string of letters formed a word (Neeley, 1977). The experimenter explained that that after completing the task the computer would calculate a score reflecting the speed and accuracy of each participant’s responses. The experimenter instructed participants to record their score on the task on the application packet. The critical trials on this task consisted of words related to discrimination (sexism, bias, discrimination, prejudice, unjust, inequality , and partial). The task also contained a total of 42 filler words. I matched the critical trials to the filler words in terms of length. Also, I scrambled the letters of each of the critical and filler words to form 49 nonwords. The critical words, filler words, and nonwords resulted in a total of 98 trials (See Appendix C). The computer program randomly interspersed critical trials throughout the task and randomized the order of presentation of trials for each subject. I used the response latencies for the critical trials as a measure of construct accessibility. I based this measure on work by other researchers who have used a word recognition task as a measure of construct accessibility (e.g. Dovidio, Evans, & Tyler, 1986; Meyer & Schvaneveldt, 1971; Rudman & Borgida, 1995; Sechrist & Stangor, 2001; Wittenbrink, Judd & Park, 1997). For example, Rudman and Borgida found that women who were primed to view women as sexual objects responded slower to nonsexual words than women who were not primed.

27 Filler application materials. After completing the lexical decision task, participants completed the remaining sections of the application packet. The application materials consisted of a background information sheet, a personal statement, and a series of scenarios asking the participant to make organizational decisions. The information sheet assessed basic demographic information, including the gender of the participant. I included the decision making scenarios, adopted from Bailey and Alexander (1993), to increase the amount of effort the participants put into applying for the role of comanager. Finally, completing the personal statement required participants to write a few paragraphs on why they would make a good co-manager. I also included the personal statement to increase participant effort and the fidelity of the application (see Appendix D). After 10 minutes the experimenter collected the application materials and indicated that the materials would be brought to the manager to be evaluated. While the waiting for feedback from the manager, the experimenter asked each participant to fill out a questionnaire assessing role preference and perceptions of performance on the application (see Appendix E). I included these measures to ensure the participants knew that the co-manager position was more desirable than the associate position and to measure perceptions of performance on the application materials. Ambiguous Negative Event. A few minutes later, the experimenter returned with the role assignment decision made by the manager, presented on a sheet of paper. The sheet included the name of the participant and two common white male names. Next to one of the fictitious names, the word “manager” was handwritten. Next to the other fictitious name, the word “co-manger” was handwritten. Next to the name of the actual participant, the word “associate” was handwritten. Also, handwritten comments appeared

28 on the bottom of the sheet indicating the manager felt that, “the girl did not come across well on the application. I [the manager] chose the guy for the role of co-manager because I don’t think I would work well with her [the participant].” Each participant received the same feedback (see Appendix F). Measurement of perceiving. After giving the participant a few minutes to read the feedback, the experimenter returned with a questionnaire (see Appendix G). The questionnaire stressed that all responses would be completely confidential. After each participant completed the questionnaire the experimenter instructed her to seal the questionnaires in an envelope. The questionnaire assessed the extent to which the participants thought the decision of the manager was due to sexism (Do you believe the
decision made by the manager to assign you to your current role was due to sexism?) on a five

point scale ranging from 1 = not at all to 5 = definitely. The questionnaire also asked
participants to indicate the extent to they attributed the manager’s decision to an external

source (the manager’s biases) versus an internal source (performance on the application materials). The questionnaire included five additional questions assessing perceptions of the manager and the fairness of the decision. I included the additional items to minimize suspicion that the study was about sexism. Measurements of reporting. The first measure of reporting discrimination followed the paradigm used in research on filing of organizational grievances (OlsonBuchanan, 1996). Participants received a packet of organizational forms to complete in their role as an associate. One of the forms gave the participants the option of filing a formal grievance with the company. The participants had the choice of either filing or not filing the grievance form. I embedded the reporting form within two other forms to

29 minimize suspicion. One of the filler forms assessed the participants’ attitudes toward RLK Consulting adopting a company wide charity and the other asked participants to indicate which of several benefits options they would prefer (see Appendix H). As a second measure of reporting, the manager (actually a male confederate) knocked on the door to check on each participant while the participant was working on the questionnaire packet. The manager asked the participants if they had any questions or comment. The manager recorded the response of each participant on a sheet of paper. Manipulation checks. Next participants received a final questionnaire (see Appendix I). This questionnaire assessed information about experiment including the climate manipulation (Generally speaking, what type of environment does this company provide for women?), the gender of the other participants (What was the name of the person assigned the role of manager?), and participants’ suspicions concerning the experiment (Do you believe that you were deceived in this study in any way?). After the participants completed the manipulation and suspicion checks, the experimenter provided a full debrief of the study. Pilot Testing Before collecting data, I conducted pilot testing. During the summer of 2003 10 undergraduate students at the same large Mid-Atlantic university participated in the study. I recruited participants by approaching instructors of psychology summer courses and asking if the instructors would be willing to offer extra credit to students for participating in research. After the instructors agreed, I went to the classes, explained the basics of the study to the students, and provided my contact information.

30 The main purposes of pilot testing included ensuring that the manipulation of CID was salient and that participants found the organizational simulation engaging. As measured by an item included in the final questionnaire of the study (Generally speaking, what type of environment does this company provide for women?), all participants correctly perceived the CID manipulation in pilot testing. In the negative CID condition all participants indicated that the climate was negative, and in the positive CID condition all participants indicated that the climate was either positive or neutral. I extensively interviewed each participant after the study. In addition to including the manipulation check concerning climate, during the interview I asked the participants about their impression of organizational climate. The interviews reinforced that each participant correctly interpreted the manipulation. Participants in the negative CID condition also indicated that the manipulation made it too obvious that the study was about gender and gender discrimination. In response to these comments, the strength of the manipulation was decreased. Originally the manipulation of CID included varying the number of females in top management positions, the tone of comments from employees in the organization, and whether the company was placed on a list of companies ranked as being among the best or worst places for women to work. After pilot testing, the list of best/worst companies for women to work for was omitted and the number of comments about the informal aspects of CID was reduced from four to one. During the post-study interview I also asked the pilot participants about the psychological fidelity of the study. When asked how engaging they found the organizational simulation, participants reported that they felt motivated to work for the

31 promotion to the role of co-manager and that they were disappointed when they were not promoted. Although the sample size was too small to conduct quantitative analyses, the pilot results were generally in the predicted direction. Consistent with H4 pilot participants in the negative CID condition had higher construct accessibility of discrimination (M = 718.52, SD = 226.13) than pilot participants in the positive CID condition (M = 765.60, SD = 299.23). Consistent with H5, pilot participants in the negative CID condition were more likely to perceive discrimination (M = 4.17, SD = 1.17) than pilot participants in the positive CID condition (M = 3.25, SD = .96). Inconsistent with H7 pilot participants in the negative CID condition were also more likely to report discrimination than pilot participants in the positive CID condition. I did not explore hypotheses concerning individual differences and the interaction between individual differences and CID because I did not measure individual differences in the pilot study. Results Suspicion and Manipulation Checks Deletion of participants due to suspicion. A review of the suspicion check questionnaires revealed that 11 of the 141 participants did not believe that they were competing with two other participants for the role of co-manger, that they ever had a chance at being promoted to the role of co-manager, or that the manager was a participant in the study instead of a confederate. Because it is doubtful that these participants were engaged in the organizational simulation, I deleted these participants before performing all subsequent analyses. The qualifications I used for deletion are consistent with those used in Major et al. (2002), who employed a similar design. Four of the suspicious

32 participants were in the positive condition and seven were in the negative condition. Therefore, I conducted the analyses with 130 participants, 66 in the positive CID condition and 64 in the negative CID condition. Manipulation checks. The final questionnaire included an item that assessed perceptions of CID (Generally speaking, what type of environment does this company provide for women?). In the positive CID condition 93.94% of participants correctly identified the CID as positive or neutral. In the negative CID condition, 84% of participants correctly identified the CID as negative. As a second manipulation check, the final questionnaire also assessed the number of women in top management at the company (How many women are in the top management of RLK Consulting?). In the positive condition 89.4% of participants correctly answered 50% and in the negative condition 96.9% of participants correctly answered none. Furthermore, 100% of participants in both conditions correctly indicated that the manager was male and that they competed with another male for the role of co-manager. Based on these results I concluded that the climate manipulation was successful and that the gender of the comanager and manager was salient. Participants did not differ between conditions in their evaluation of performance on the application or desire for the position of co-manager. As measured by a five-point scale ranging from very poorly to very well, on average participants considered their performance on the lexical decision task above the center point of the scale in the positive (M = 3.87, SD = .79) and negative (M = 3.79, SD = .88) conditions and perceptions of performance did not differ between conditions (t(119) = .53; p = .60). Similarly, on average participants considered their performance on the organizational decisions above

33 the center point of the scale in the positive (M = 4.00, SD = .53) and negative (M = 4.10, SD = .73) conditions and perceptions of performance did not differ between conditions (t(127) = -.85.; p = .40). Furthermore, on average participants considered their performance on the personal statement above the center point of the scale in the positive (M = 3.58, SD = .91) and negative (M = 3.44, SD = 1.15) conditions and perceptions of performance did not differ between conditions (t(127) = .72; p = .47). Finally, as measured by a 5-point scale ranging from strongly prefer associate to strongly prefer comanager, on average participants in the positive (M = 4.33, SD = .88) and negative (M = 4.38, SD = .85) conditions preferred the role of co-manager and the two conditions did not differ significantly (t(127) = -.31; p = .76). Descriptives Individual difference variables. Although all students recruited to participate in the study ostensibly participated in mass testing during the beginning of the semester, individual difference data was not available for 4 participants reducing the sample size for the individual difference analyses from 130 to 126. For each individual difference variable I conducted a factor analysis of the items and calculated the internal reliability. The six item measure of sensitivity to sexism (SS) included three items assessing sensitivity to sexism directed at the individual and three items assessing sensitivity to sexism directed at women as a group. The purpose of the sensitivity to sexism scale was to capture sensitivity toward sexism directed at individual women. I included the groupbased sexism items because previous research has shown all six items to be highly correlated. The six item scale had a reliability of = .76 and the correlations among the

items ranged from .08 to .65. An exploratory principal components analysis of the items

34 using oblimin rotation and the Kaiser’s extraction method, resulted in two factors. The three individual sensitivity to sexism items had high loadings (absolute value greater than .40) on the first factor (
1

= 2.58, 42.98% of variance, after rotation). The second two
2

sensitivity to group sexism items loaded highly on the second factor (

= 2.023, 33.73%

of variance, after rotation), and the first sensitivity to group sexism item had high loadings on both factors (see Table 2a). The emergence of two factors indicates that the individual and group sensitivity to sexism are separate factors. I examined the properties of a scale consisting of only the first three items, which measured individual sensitivity to sexism. The reliability of this scale was = .77 and the

correlations among the items ranged from .50 to .64. An exploratory principal components analysis using oblimin rotation and Kaiser’s extraction method resulted in one factor (
1

= 2.12). The loadings of the three items ranged from .79 to .87 (see Table

2b). I formed a composite by averaging individual scores on these three items and used the composite as a measure of individual sensitivity to sexism (ISS) in all subsequent analyses. I also examined the group based sensitivity to sexism items as an independent scale. Although the scale had good measurement properties, the composite was not related to any of the dependent variables in the study. The reliability of the four-item gender identity (GI) scale was = .57 and the

correlations among the items ranged from .17 to .62. The reliability of the items was lower than in previous research (e.g. Major et al., 2002; = .79). I also conducted an

exploratory principal components analysis using oblimin rotation and Kaiser’s extraction method on the GI items. The results indicated that two factors should be extracted (see Table 2c). The two items that loaded on the first factor (
1

= 1.70, 42.38% of variance,

35 after rotation) were reverse scored, whereas the two items loading on the second factor (
2

= 1.38, 34.55% of variance, after rotation) were not reverse scored. Because there was

no theoretical justification for separating this scale into two subscales I formed a composite by averaging all four items. Although this scale had poor measurement properties, this composite was used in all subsequent analyses as a measure of gender identity. The six item past experience with gender discrimination (PEGD) scale had a reliability of = .85 and the correlations among the items ranged from .24 to .77. An

exploratory principal components analysis of the items using oblimin rotation and Kaiser’s extraction method also resulted in one factor (
1

= 3.70) with the loadings of the

items on that factor ranging from .51 to .87 (see Table 2d). I formed a composite of past experiences with gender discrimination by averaging individuals’ scores on each of the six items. In all subsequent analyses the composite for each individual difference variable (ISS, GI, and PEGD) was centered and entered as a continuous variable. I conducted analyses to determine if participants in the two CID conditions differed in terms of the individual difference variables. Gender identity (t(123) = .22, p = .83) and past experiences with gender discrimination (t(124) = .43, p = .67) and did not differ by condition. There was a marginal difference between conditions for individual sensitivity to sexism (t(124) = 1.96, p = .052). However, the mean for ISS was higher in the positive climate condition (M = 3.05; SD = 1.0) than in the negative climate condition (M = 2.68; SD = .84). The data measuring individual sensitivity to sexism was collected several weeks before participants were exposed to the climate manipulation. Therefore it is impossible that climate influenced the individual sensitivity to sexism measure. The

36 marginal difference in ISS between the two conditions constitutes a failure of random assignment. Because this effect is in the opposite direction of my hypotheses regarding CID the marginal mean difference of individual sensitivity to sexism between conditions creates a stricter test of the hypotheses. For example, I hypothesize that individuals will be more likely to perceive discrimination in a negative CID than in a positive CID. Similarly, I predict that high ISS will increase perceptions of discrimination. If I confirm the hypothesis that perceptions of discrimination are higher in the negative CID condition than in the positive CID condition, the mean difference of ISS between conditions will not be a possible explanation of the effect because participants in the positive CID condition had higher ISS than participants in the negative CID condition. Climate for Intolerance for Discrimination. In all analyses, I dummy coded CID. I coded positive CID with a 0 and negative CID with a 1. Climate was considered a categorical variable. Construct Accessibility Data. I collected construct accessibility data from 120 participants. Due to problems with the computer application used to collect the construct accessibility data, 10 of the original 130 participants were not able to complete the task. Each participant responded to seven critical trials (i.e. discrimination related words), embedded within a total of 100 trials. Because response time data are often characterized by outliers and positive skew, I identified all outliers and incorrect responses in the data (Ratcliff, 1993; Wittenbrink et al., 1997). Outliers were defined as values greater than 150 ms and less than 1,500 (Wittenbrink et al., 1997). Outliers accounted for 3.81% of the data and incorrect responses accounted for 3.57% of the data. Consistent with

37 previous research I treated outliers and incorrect responses as missing data. Therefore, in sum, 7.38% of the data were recorded as missing. In cases where one of the 120 participants who completed the construct accessibility task were missing values for one or more of the seven critical response times (because the response was either incorrect or an outlier) mean substitution was used in calculating the composite score for that individual. Thirty-five subjects were missing one value, nine subjects were missing two values, and three subjects were missing three values. If mean substitution was not used and participants missing one or more response times for the critical trials were deleted casewise, the sample size for the construct accessibility measure would have been reduced from 120 to 73. I used mean substitution because there is not sufficient justification for excluding participants from the construct accessibility analyses on the basis of having one or more incorrect or outlying responses. I used the response times for the seven discrimination related words to form a composite score of construct accessibility of discrimination. The reliability of the seven response times was = .82 and the correlations among the response times for each of the

seven discrimination related words ranged from .24 to .51. An exploratory factor analysis with oblimin rotation and using Kaiser’s extraction method resulted in a one factor solution (
1

= 3.44) with all loadings ranging between .62 and .77. The seven response

times were averaged for each participant to form a composite measure of construct accessibility of gender discrimination. As is often the case with response time data, the composite construct accessibility data were significantly positively skewed (skewness = .949; t(119) = 4.29; p < .00). One method for dealing with positively skewed data is to conduct a natural log transformation

38 (Wittenbrink et al., 1997). I applied a natural log transformation to the individual scores for each of the seven discrimination related words and then averaged the transformed data across individuals. The log transformation reduced the amount of skew in the data (skewness = .431; t(119) = 1.95; p = .06). I also examined the properties of the transformed data. The reliability of the seven response times was = .83 and the

correlations ranged from .28 to .52. An exploratory factor analysis with oblimin rotation and using Kaiser’s extraction method resulted in a one factor ( loadings ranging from .63 to .75. I used the log transformed composite construct accessibility data in all subsequent analyses. For ease of interpretation, I report descriptive statistics for the untransformed data. Because response time data has a meaningful zero point, I did not center the construct accessibility data in subsequent analyses. Perceiving. Perceiving discrimination was measured by a single item (Do you believe the decision made by the manager to assign you to your current role was due to sexism?), measured on a five-point scale ranging from 1 = not at all to 5 = definitely. In all analyses I entered perceiving as a continuous variable. I only centered perceiving for analyses in which it was included as an independent variable. Reporting. Reporting discrimination was measured using two dummy coded dichotomous variables. The first reporting variable reflected whether or not the participant complained about sexism on the organizational grievance form. If a participant did file a complaint about sexism I coded the response as a 1 (N = 36). I coded responses as complaints about sexism if in the written explanation of the grievance the participant mentioned that the decision was based on gender. If the participant did not file
1

= 3.54) solution with all

39 a complaint about sexism I coded the response as a 0 (N = 94). The second reporting variable reflected whether or not a participant reported being discriminated against to the manager when the manager came to check on the participant. Of the 130 subjects, only 4 participants commented to the manager that they were unhappy about the decision he made (3 in the positive condition and 1 in the negative condition). Because only a small number of participants reported discrimination to the manager directly (3%) I only conducted analyses using the first reporting variable. Correlations among the variables. Table 3 reports the means, standard deviations, and correlations for the variables in the study including ISS, GI, PEGD, CID, construct accessibility, perceiving, and reporting. Analyses at Each of the Three Attribution to Discrimination Stages Stage 1: Construct accessibility as a dependent variable. I used hierarchical regression to assess the influence of individual difference variables and climate on construct accessibility of gender discrimination. I entered the three individual difference variables, individual sensitivity to sexism (ISS), gender identity (GI), and past experiences with gender discrimination (PEGD), in Step 1. I entered CID in Step 2. I entered the hypothesized interactions between CID and each of the individual difference variables in Step 3. These analyses tested Hypotheses 1a-c, 4, and 9a-c and the results are presented in Table 3. None of the hypotheses regarding the influence of individual differences (ISS, GI and PEGD), CID, and the individual differences by CID interaction on construct accessibility were supported. Stage 2: Perceiving as a dependent variable. I used hierarchical regression to assess the influence of the individual difference variables and CID on perceptions of

40 gender discrimination. Steps 1-3 were identical to those in the hierarchical regression examining construct accessibility as a dependent variable. In this analysis, however, I entered perceiving as the dependent variable. These analyses tested Hypotheses 2a-c, 5, and 10a-c and the results are presented in Table 3. As seen in Table 3, at Step 1 the model was not significant and none of the individual difference variables in Step 1 were significant. These results fail to support H2a-c, which predict that individual differences will influence perceiving discrimination. At Step 2 the model was significant (R2 = .15, p < .00). In support of H5, the change from Step 1 to Step 2 ( R2 = .11, p < .00) and the effect of CID ( = .34, p < .00) were significant indicating that CID predicted perceptions of discrimination such that individuals in the negative climate condition were more likely to perceive discrimination (M = 3.58; SD = 1.02) than individuals in the positive CID condition (M = 2.97; SD = .98). H10a-c predict that individual differences and CID will interact to predict perceiving discrimination. At Step 3 the model was significant (R2 = .20, p < .00), and the change from Step 2 to Step 3 was marginal ( R2 = .05, p = .096). In support of H10a, there was a significant interaction between ISS and CID ( = -.44, p = .02). Simple effects of the ISS by CID interaction indicated that the interaction was in the predicted direction. I dichotomized the sensitivity to sexism variable and conducted a Scheffé’s test to specify the nature of the interaction. In the positive condition, individuals who had high sensitivity to sexism were significantly more likely to perceive discrimination than individuals who had low sensitivity to sexism (F = 16.59; p < .00). In the negative condition, ISS did not predict perceiving (F = .34; p > .25). Therefore, participants with

41 high ISS perceived sexism in both conditions, but participants with low ISS only perceived sexism in the negative condition (see Figure 2). H10b (GI) and H10c (PEGD) were not supported. H6 predicts that construct accessibility will mediate the relationship between CID and perceiving discrimination. Results using construct accessibility as the dependent variable indicate that CID does not predict construct accessibility. A relationship between CID and construct accessibility is a necessary condition for construct accessibility to mediate the relationship between CID and perceiving. Therefore, H6 was not supported. Stage 3: Reporting as a dependent variable. I used hierarchical logistics regression to explore the influence of individual differences and CID on reporting discrimination. I used a coded variable indicating whether or not a participant used the organizational grievance form to indicate that the promotion decision made by the manager was sexist as the measure of reporting discrimination. I used logistics regression for these analyses because the reporting variable was dichotomous.1 As in the previous analyses, I entered all three individual difference variables in Step 1, CID in Step 2, and the three interactions in Step 3. This analysis provided a test of Hypotheses 3a-c, 7, and 11a-c. The results are reported in Table 4. The model was not significant at Step 1 and none of the individual difference variables predicted reporting. Therefore, H3a-c were not supported. At Step 2 the model was also not significant, but the change from Step 1 to Step 2 was significant (
2

= 4.05, p = .04; Cox & Snell R2 = .05) as was the effect of CID

(Wald = 3.91, p < .05; Exp(B) = 2.30), indicating that climate predicts reporting such that individuals in the negative CID condition were more likely to report discrimination than

42 individuals in the positive CID condition. H7 predicts that individuals will be more likely to report sexism in the positive than in the negative condition. Therefore, this finding was in the opposite direction of H7. H11a-c predict that individual differences will interact with CID to predict reporting of discrimination. At Step 3 the model was marginal ( was the change from Step 2 to Step3 (
2 2

= 13.07, p = .07) as

= 6.64, p = .08; Cox & Snell R2 = .10). H11a,

which predicts that CID will interact with ISS, was supported (Wald = 5.31, p = .02; Exp(B) = .22). I explored the ISS by CID interaction by splitting the sample by CID condition to test the simple effects. I did not use Scheffé’s test because the dependent variable was dichotomous. In the positive CID condition, individuals with high ISS were more likely to report discrimination than individuals with low ISS (
2

= 7.11, p = .01;

Cox & SnellR2 = .11, Wald = 6.20, p = .01; Exp(B) = 2.37). In the negative CID condition, ISS did not predict reporting of discrimination (
2

= .003, p = .96; Cox & Snell

R2 = .00; Wald = .00, p = .96; Exp(B) = .99). These results indicate that participants with high ISS reported sexism in both conditions, but participants with low ISS reported sexism in the negative condition, but not the positive condition (see Figure 2). H11b and c, which predict that gender identity and past experiences with discrimination will influence reporting of discrimination, were not supported. Mediation. H8 predicts that construct accessibility and perceiving will mediate the relationship between CID and reporting. Because construct accessibility was not significantly related to any variable in the study, it follows that construct accessibility should be removed from the model. Testing a model involving CID, perceiving, and reporting, however, would not explore all aspects of the data because ISS also interacted

43 with CID to predict perceiving and reporting. Based on the preceding analyses, I modified H8 to predict that perceiving will mediate the relationship between the ISS by CID interaction and reporting (see Figure 3). Researchers propose as four step process for testing if a variable mediates the relationship between a predictor and a criterion (Baron & Kenny, 1986; Judd & Kenny, 1981). I used these steps to determine if perceiving (mediating variable) mediates the relationship between the ISS by CID interaction (predictor variable) and reporting (criterion variable). The test for mediation includes showing that: (1) the ISS by CID interaction predicts reporting; (2) the ISS by CID interaction predicts perceiving; (3) controlling for the ISS by CID interaction, perceiving predicts reporting; and (4) controlling for perceiving, the ISS by CID interaction no longer predicts reporting (Baron & Kenny, 1986; Judd & Kenny, 1981). I used hierarchical logistics regression, entering ISS and CID at Step 1, the interaction at Step 2, and reporting as the dependent variable, to test the first part of the mediation process. Step 1 was marginal (
2

= 4.88, p = .09; Cox & Snell R2 = .04) as was

the effect of CID at Step 1 (Wald = 3.24, p = .07; Exp(B) = 2.11). At Step 2 the model was significant (
2

= 9.62, p = .02 as was the change from Step 1 to Step 2 (

2

= 4.74, p

= .03; Cox & Snell R2 = .07). The interaction between ISS and CID predicted reporting of discrimination (Wald = 4.41, p = .04; Exp(B) = .42) and therefore the first criteria for mediation was met. To test the second criteria for mediation, I used hierarchical regression, entering CID and ISS at Step 1, the interaction at Step 2, and perceiving as the dependent variable. Step 1 was significant (R2 = .16, p < .00) as were the effects of climate ( = .35, p < .00)

44 and ISS ( = .26, p = .00). At Step 2 the model was significant (R2 = .18, p < .00) and the change from Step 1 to Step 2 was marginal ( R2 = .02, p = .06). The interaction between ISS and CID marginally predicted perceiving of discrimination ( = -.24, p = .06), therefore marginally supporting the second criteria for mediation. To test the third and forth criteria for mediation, I used hierarchical logistics regression entering ISS and CID in Step 1, the interaction and perceiving in Step 2, and reporting as the dependent variable. Step 1 was marginal (
2

= 5.48, p = .07; Cox & Snell

R2 = .04) as was the effect of CID (Wald = 3.82, p = .05; Exp(B) = 2.28). The effect of ISS was not significant. At Step 2 the model was significant ( the change from Step 1 to Step 2 (
2 2

= 41.27, p < .00) as was

= 35.79, p < .00; Cox & Snell R2 = .28). In support

of the third criteria for mediation, in Step 2 perceiving predicted reporting (Wald = 21.10, p < .00; Exp(B) = 4.34). In support of the fourth requirement for full mediation, in Step 2 the ISS by CID interaction no longer predicted reporting of discrimination (Wald = 1.98, p = .16; Exp(B) = .49). Therefore, the results indicate that the relationship between the ISS by CID interaction and reporting was fully mediated by perceptions of discrimination. Discussion In sum, the results of this study provide support that individual differences and contextual variables interact to predict a multi-stage attribution to discrimination process (see Figure 3). Specifically, Climate for Intolerance for Discrimination interacted with individual sensitivity to sexism to predict perceptions of discrimination, which in turn predicted reporting of discrimination. Overall, women were more likely to perceive being passed over for a promotion as discrimination and report the incident to the company in

45 an environment in which women were not treated as equals both formally and informally (negative CID) than in an environment in which they were (positive CID). In the presence of a negative CID there was no difference in perceiving between women who had high and low sensitivity to sexism. In a positive CID, however, women who reported being sensitive to sexism were more likely to perceive discrimination than women who reported not being sensitive to sexism. Furthermore, perceptions of discrimination fully mediated the interactive effect of ISS and CID on reporting of discrimination. Previous research, that does not consider context, indicates that individual differences are an antecedent to the attribution to discrimination process (Stangor, et al., 1999; Swim & Hyers, 1999). Similarly I found individual differences predicted perceiving in the positive CID condition. In the negative CID condition, however, individual differences no longer predicted perceiving and reporting of discrimination. Therefore, this research adds to the literature by documenting the interactive effects of individual differences and context on attributions to discrimination. Support for perceiving discrimination as a mediator explains why the relationship between the ISS by CID interaction and reporting was not in the predicted direction. I predicted the interaction to influence reporting of discrimination in the opposite direction of the influence of the interaction on perceiving of discrimination. The full mediation of perceiving discrimination explains why the interaction influenced reporting in the same direction it influenced perceiving. As shown in the mediation steps, when controlling for perceiving the ISS by CID interaction did not explain a significant amount of the remaining variance in reporting.

46 Although 36 participants (27%) reported discrimination on the organizational grievance form, only 4 participants (3%) actually confronted the manager about the promotion decision. This result indicates women’s reluctance to directly address discrimination. An individual puts him or herself at greater risk by directly confronting a perpetrator of discrimination than by checking a box on a form. Women’s hesitancy to directly confront sexism has been supported in other research. For example, Swim & Hyers (1999) found that although 81% of women indicated that would confront a discriminatory individual, only 45% actually did when presented with the situation. In addition to contributing to the literature on attributions to discrimination, this study examines the influence of organizationally relevant variables on the attribution to discrimination process. Integrating the social psychology paradigm into the study of organizationally relevant variables adds a systematic approach to the study of attributions to discrimination that organizational psychology previously lacked. The current study also introduces the new construct of Climate for Intolerance for Discrimination. Although much research on climate for diversity and inclusion exists, the literature lacks a clear definition of the construct and a widely accepted scale. Although the concept is manipulated instead of measured, I presented CID as a well-defined facet of climate for diversity. Furthermore, this research contributes to organizational psychology by linking a macro-level organizational context variable to a micro-level individual cognitive process. Although researchers have developed theory concerning cross-level relationships, few empirical studies document their existence (e.g. Perry et al., 1994; Schneider et al., 2000). In this thesis, an organizational level variable, CID, influenced the individual cognitive

47 process of making an attribution to discrimination. The effect of a macro-level contextual variable on a micro-level cognitive variable suggests a need for researchers to explore other ways in which organizational variables influence not just the organization as a whole, but also individuals within the organization. Additionally, previous researchers examined the influence of perceptions of organizational level variables on perceptions discrimination (e.g. Goldman, 2001; Ragins & Cornwell, 2001). Such research, however, does not tap variables at the organizational level. Furthermore, these studies are afflicted by single source bias due assessing the independent and dependent variables from the same individuals at the same point in time. The current study minimized single source bias by manipulating one of the independent variables (ICD) and by measuring the other independent variables (individual differences) at a different time than the dependent variables (construct accessibility, perceiving, and reporting). Implications for Organizations The results of this study have implications for organizations seeking to combat discrimination. By creating a positive Climate for Intolerance for Discrimination, organizations can decrease both perceptions and reporting of discrimination. Unlike the individual differences that have been shown to influence the attribution to discrimination process, climate is in the control of an organization. Therefore, if perceptions of discrimination are problematic in a particular organization, that organization can take specific steps to create a climate that does not tolerate treatment of employees based on group membership. For example, an organization could make an effort to hire protected group members into top executive positions and institute training programs to encourage

48 managers to include all group members in informal organizational interactions. Decreased perceptions and reporting of discrimination are likely beneficial to both individuals and the organization as whole, as researchers have linked perceptions of discrimination to negative outcomes for individuals such as decreased psychological well being and career dissatisfaction (Foley & Kidder, 2002; Schmitt et al., 2002). In turn, these variables are likely to have negative effects on the work quality and productivity of employees. The discrepancy between the two measures of reporting discrimination also provides useful information to organizations seeking to combat perceptions of discrimination. Participants who perceived discrimination were more likely to report discrimination on an organizational form, than to directly confront the perpetrator. The reluctance of women, and other protected group members, to directly and openly confront discrimination indicates that less confrontational ways of reporting acts of discrimination are needed if organizations want to address instances of potential discrimination. This finding reinforces research suggesting that creating grievance filing systems in organizations that involve minimal confrontational and that are sensitive to workplace difficulties experienced specifically by women will increase the number of women willing to file a grievance (Gwartney-Gibbs & Lach, 1994). Limitations In spite of support of an interactive, multistage process of making attributions to discrimination, this study was not without limitations. I aimed to expand current understanding of the attribution to discrimination process by evaluating construct accessibility, perceiving and reporting of discrimination within a single study. The

49 hypotheses regarding construct accessibility, however, were not supported. It is unlikely that the lexical decision task captured construct accessibility of gender discrimination but construct accessibility was unrelated to perceiving and reporting of discrimination. The lack of findings for the construct accessibility data could be due to measurement issues. Construct accessibility by definition is a subconscious process, making measurement difficult. Moreover, previous research has shown construct accessibility to be related to individual difference variables including sensitivity to sexism, although the measure of construct accessibility used was not a word recognition task (Stangor et al., 1999). In the current study, construct accessibility was not related to sensitivity to sexism (r = .06, p = .54). Sensitivity to sexism did, however, did relate to perceiving and reporting of discrimination. Participants’ responses to the word recognition task provide another indication that the task did not tap the desired construct. Many participants indicated that they did not understand why the task was included as part of the application. Preoccupation with wondering how the task would be used in the promotion decision may have distracted students from concentrating on the task. Therefore, the lack of findings in this study does not necessarily indicate that construct accessibility is uninvolved in the attribution to discrimination process. The possibility remains that the relationship between the ISS by climate interaction and perceiving is fully or partially mediated by construct accessibility. Although I examined three individual difference variables, I failed to find a relationship between gender identity (GI) or past experiences with gender discrimination (PEGD) and the attribution to discrimination process. The scale I used to measure GI, however, had poor measurement qualities. The reliability of the GI scale was lower than

50 any of the other individual difference variables measured and was below the guideline of = .70. The reliability for this scale was also lower than in previous research where GI predicted perceptions of discrimination (e.g. in Major et al., 2002 = .79). The results of

a factor analysis indicated that the four items used to measure GI split into two factors with the reverse coded items loading on the first factor and the non-reverse coded items loading on the second factor. The split in the scale is problematic because all four items were designed to tap the same construct. The correlation between factor loadings and coding scheme of the items indicates that participants’ responses to the items reflected the scoring of the item and therefore were likely biased. Also, the GI scale was included in the same packet of questionnaires as the rest of the individual difference measures but was the last scale in the packet. It is possible that fatigue may have decreased the amount of accuracy used by participants when completing the questionnaire. Finally, recent research indicates that increased gender identity is a consequence of perceiving discrimination among women (Schmitt et al., 2002). Examination of GI as a consequence was not possible in this study due to the temporal order in which the variables were measured. However, it is possible that GI is a consequence of the attribution to discrimination process instead of an antecedent. Similarly, the results did not support any of my hypotheses regarding past experiences with gender discrimination. Unlike GI, however, the PEGD scale did have good measurement properties. Furthermore, PEGD was highly correlated with ISS (r = .70, p < .00) and the items for the PEGD and ISS scales loaded onto the same factor in a principal components analysis. In spite of evidence that PEGD and ISS could be combined into one scale, I kept the scales separate. Although closely related, I believe

51 that sensitivity to sexism and past experiences with gender discrimination are separate constructs. The insignificant results found for PEGD in comparison with the significant results found for ISS reinforce this decision. A causal relationship could be the reason for the high correlation of the two scales. For example, many past experiences with gender discrimination may increase sensitivity to sexism. A causal relationship offers a potential explanation for why PEGD and ISS are highly correlated, but only ISS interacts with climate to predict perceiving and reporting of discrimination. The power of detecting the effects in this study constitutes another limitation. Some participants were lost due to malfunctioning of the lexical decision task, failure to complete the individual difference measures, and suspicion regarding the organizational simulation. These factors reduced the original sample of 141 subjects, to a sample of 116 participants who completed all measures in the study. Ten of the original 141 participants completed all measures in the study with the exception of the construct accessibility task. Because construct accessibility was not included in the final model (see Figure 3) the sample size for significant analyses in the study was 126. The small sample size may have decreased my ability to find significant results for the individual difference variables. For example, the combined effect size of the three individual difference variables on perceiving discrimination was R2 = .041. This result was nonsignificant. I conducted a power analysis and determined that the probability of finding an effect of this size to be significant at the greater than 50% (L(2) = 5.25). Although the results indicate that perceiving fully mediates the relationship between the ISS by CID interaction and reporting of discrimination, the Baron and Kenny = .05 level given a sample size of 126 is only slightly

52 (1981) method for testing for mediation does not absolutely establish mediation. First, because the path from perceiving to reporting is large but the path from the ISS by CID interaction to perceiving is marginal, there is some evidence that perceiving is a distal mediator. In the ideal case of mediation, both paths would be large. Another concern in mediated models is the possibility of measurement error in the mediator increasing the size of the path from perceiving to reporting. The size of this path, however, is large and therefore it is unlikely that controlling for measurement error would make the path nonsignificant. Furthermore, it is possible that variables have been omitted in the mediated model presented in Figure 3. Although I measured many variables relevant to the attribution to discrimination process the possibility of other intervening variables cannot be ruled out. Multicollinearity and reverse causal effects, also potential pitfalls of test for mediation, are not an issue in the study. High multicollinearity between a predictor and a mediator can result in insufficient variance in left in the mediator to explain the outcome. In this study, however, the climate by ISS interaction and perceiving are not highly correlated. Finally, testing for mediation does not rule out the possibility of reverse causation between the mediator and the outcome. Due to the nature of the variables and the temporal order in which they were measured, however, it is not possible that reporting precedes perceiving. A final limitation of this study is the context in which it was conducted. Although the goal of the study was to provide meaningful information about diversity issues in organizations, I conducted the study in a laboratory setting using undergraduate students. A question remains as to whether or not the results of this study can be generalized from

53 undergraduates in a simulation to real employees in real organizations. A field study is needed to test the generalizability of the model to organizations. Furthermore, use of a laboratory study mandated that I manipulated climate instead of measuring it. An essential aspect of climate is that it is shared among members of a group or organization. In other words, if there is too much variance among employees’ perceptions of an organization’s policies, practices, and procedures, climate is no longer a meaningful construct. By manipulating climate, I ensured that participants within each condition were given the same information regarding CID. Agreement was then assessed using the manipulation checks. CID, however, should also be validated as a meaningful organizational construct using a field study. Future Directions The first step to validating this research is a field replication. Even if the results are supported in the field, however, additional organizational research on the attribution to discrimination process will still be needed. Although this thesis provides an initial investigation of an organizational model of the attribution to discrimination process, the picture is far from complete. Additional organizational antecedents should be examined (see Figure 4). Other multi-level organizational variables of interest should include team member diversity on the work group level, leader-member exchange on the dyadic level, and perceptions of organizational justice on the individual level. In addition to examining women as the protected group of interest, future research should examine whether climate and sensitivity have similar effects on ethnic minority group members. Finally, a comprehensive model of the attribution to discrimination process within organizations would not be complete without examining consequences in addition to antecedents.

54 Future research should explore organizational outcomes of the attribution to discrimination process including as job satisfaction, work withdrawal and turnover. Organizational consequences should be examined as they relate to multiple stages of the attribution to discrimination process. For example, individuals who perceive discrimination and report the incident may be less likely to experience a decrease in job satisfaction than individuals who perceive discrimination but do not report it. Future research should also explore the construct of Climate for Intolerance for Discrimination. In this study I assumed that the formal and informal organizational aspects of CID are highly interrelated. In reality, the two aspects of CID may differ within an organization. Due to the subtle nature of modern discrimination it is possible that some organizations have equitable formal policies, practices and procedures, yet protected group members are excluded from informal interactions. Future research should assess the differential effects of the facets of CID to determine if one aspect of CID has more influence over the attribution to discrimination process than the other. Conclusions The current study contributes to the current literature in several ways. This research expands the organizational literature by introducing the construct of Climate for Intolerance for Discrimination, documenting the ability of an organizational level variable to influence individual level processes, and studying the effect of organizationally relevant variables on the attribution to discrimination process. Furthermore, this project contributes to social psychology by examining the effects of antecedents at multiple stages in Stangor et al.’s (in press) model of attributions to discrimination and documenting the importance of individual difference by context

55 interactions. Finally, in addition to the academic contribution, this research provides information that can be used by organizations committed to combating the negative effects of perceptions of discrimination.

56 Table 1 The Dimensions of Climate for Diversity and Climate for Inclusion Authors Date Scale Title # of Formal Informal Factors Aspects Aspects Barak, 1998Diversity Cherin, & Perceptions Berkman Scale HicksClarke & Iles 2000Positive Climate for Diversity 2 Fairness factorInclusion Factor Policy support; Equity scale Equality of departmental support for women Value efforts to promote diversity; Attitudes toward qualifications of minorities; Attitudes toward qualifications of women Organizational values, norms, beliefs, and assumptions reflect the importance of diversity

Value Placed on Diversity

2

Kossek & 1993Climate for Zonia Diversity

4

Nishii & Raver

2001Climate for Diversity

3

Organizational Integration of policies, nontraditional practices, and employees procedures into the social fabric of the organization

Pelled, 1999Organizational Ledford, Inclusion & Mohrman

3

Decisionmaking influence; Access to sensitive information; Job security

Note. The scale from Barak et al. (1998) actually contains four dimensions instead of two. The other two dimensions, diversity value and personal comfort factors, however, measure personal instead of organizational attributes.

57 Table 2a Sensitivity to Sexism (SS) Item Factor 1 Loading 1 0.80 2 0.88 3 0.83 4 0.51 5 -0.03 6 -0.06 Table 2b Individual Sensitivity to Sexism (ISS) Item Factor 1 Loading 1 0.79 2 0.87 3 0.86 Table 2c Gender Identity (GI) Item Factor 1 Loading 1 0.01 2 0.91 3 -0.01 4 0.89 Table 2d Past Experiences with Gender Discrimination (PEGD) Item Factor 1 Loading 1 0.81 2 0.84 3 0.83 4 0.51 5 0.79 6 0.87

Factor 2 Loading -0.02 -0.09 0.01 0.52 0.85 0.87

Factor 2 Loading 0.80 -0.02 0.80 0.02

58 Table 3 Correlations among the Variables Mean SD 1. ISS 2. GI 3. PEGD 4. Climate 5. Construct Accessibility 6. Perceiving 7. Reporting Note. * p < .05 2.87 4.98 3.08 x

N

1 -0.10

2

3

4

5

6

1.09 126.00 1.03 125.00

--0.01 -0.02 -0.10 0.05 -0.03 --0.04 -0.04 --0.12 --0.05 --

1.47 126.00 0.70* x 130.00 -0.17 0.06

754.95 146.36 120.00 3.27 0.28

1.04 129.00 0.20* 0.45 130.00 0.11

0.15 0.29* 0.04 0.15

0.02 0.51*

59

Table 4 Hierarchical Regression Results Step R2 p p Effect R2 DV = Construct Accessibility 1.00 0.03 0.31 0.03 0.31 ISS GI PEGD 2.00 0.04 0.34 0.01 0.32 ISS GI PEGD CID 3.00 0.05 0.56 0.01 0.72 0.32 -0.15 -0.21 -0.09 -0.20 0.03 0.04 0.13 0.23 0.29 0.34 0.36 0.82 0.85 ISS GI PEGD CID ISS x CID GI x CID PEGD x CID DV = Perceiving 1.00 0.04 0.17 0.04 0.17 ISS GI PE 2.00 0.15 0.00 0.11 0.00 SS GI PEGD CID 3.00 0.20 0.00 0.05 0.10 ISS GI PEGD CID ISS x CID GI x CID PEGD x CID 0.61 -0.06 -0.21 0.35 -0.44 0.12 0.24 0.00 0.88 0.23 0.00 0.02 0.26 0.16 0.27 0.03 -0.02 0.34 0.03 0.75 0.86 0.00 0.17 0.03 0.04 0.18 0.73 0.78 0.17 -0.12 -0.18 -0.10 0.23 0.19 0.19 0.32 0.20 -0.13 -0.19 0.16 0.19 0.16

p

60

Table 5 Hierarchical Logistics Regression Step
2

p

2

Cox & p Effect Snell R2 DV = Reporting 0.02 ISS GI PEGD

Wald

p

Exp (B)

1.00

2.38

0.50 2.38 0.50

2.11 0.27 0.63 3.34 0.30 1.09 3.91 7.56 0.66 2.70 5.49 5.31 0.20 1.51

0.15 1.47 0.60 0.90 0.43 0.86 0.07 0.57 0.30 0.05 0.01 0.42 0.10 0.02 1.66 0.90 0.81 2.30 4.60 0.78 0.58 3.34

2.00

6.43

0.17 4.05 0.04

0.05 ISS GI PEGD CID

3.00

13.07 0.07 6.64 0.08

0.10 ISS GI PEGD CID ISS x CID GI x CID PEGD x CID

0.02 0.22 0.66 1.21 0.22 1.68

61 Figure 1. The effect of Climate for Intolerance for Discrimination on the attribution to discrimination process.

Individual Differences

H1-H3

Construct Accessibility
H9-H11 H4 H6

Climate for Intolerance for Discrimination

Perceiving
H5

H8

Reporting
H7

62 Figure 2. The effect of the individual sensitivity to sexism by Climate for Intolerance for Discrimination interaction on perceiving and reporting of discrimination.

Perceiving Discrimination

5 4 3 2 1 0 Low ISS Hi ISS Individual Sensitivity to Sexism Positive CID Negative CID

Reporting Discrimination

1 0.8 0.6 0.4 0.2 0 Low ISS Hi ISS Individual Sensitivity to Sexism Positive CID Negative CID

63 Figure 3. Revised model of the effect of individual sensitivity to sexism and Climate for Intolerance for Discrimination on perceiving and reporting of discrimination.

Climate for Intolerance for Discrimination Individual Sensitivity to Sexism

Perceiving

Reporting

64 Figure 4. A comprehensive model of organizational antecedents to the attribution to discrimination process.

Organizational Level Variables (e.g. Climate for Intolerance for Discrimination)

Sensitivity to Discrimination

Work Team Level Variables (e.g. Group Homogeneity) Three-Stage Model

Dyad Level Variables (e.g. Leader-Member Exchange)

Construct Accessibility

Perceiving Discrimination

Reporting Discrimination

Organizational Individual Level Variables (e.g. Perception of Organizational Justice)

Individual Level Outcomes (e.g. Withdrawal Behavior)

65 Appendix A Pre-measures Demographics Please circle the appropriate response. Gender: Female Male Asian Black/African-American Latino/a, Hispanic Middle Eastern/Arab Native American White/European-American/Caucasian International: ___________________ Biracial: _______________________ Other: _________________________

Race/Ethnicity:

Sensitivity to Gender Discrimination (Individual-/Group-Based Sexism Scale) Please read each statement carefully and circle the appropriate response ranging from not at all to very much. 1. How often do people discriminate against you on the basis of your gender? 1 2 3 4 5 6 7 not at all very much 2. How much does the gender discrimination you experience bother you? 1 2 3 4 5 6 7 not at all very much 3. How often do you think about being the victim of gender-based discrimination? 1 2 3 4 5 6 7 not at all very much 4. How often do people discriminate against other women on the basis of gender? 1 2 3 4 5 6 7 not at all very much 5. How much does the gender discrimination other women experience bother them? 1 2 3 4 5 6 7 not at all very much

66

6. How often do other women think about being the victim of gender-based discrimination? 1 2 3 4 5 6 7 not at all very much Gender Identity (Importance to Identity Subscale of the Collective Self-Esteem Scale)

INSTRUCTIONS: We are all members of different social groups or social categories. We would like you to consider your gender (e.g. male or female) in responding to the following statements. There are no right or wrong answers to any of these statements; we are interested in your honest reactions and opinions. Please read each statement carefully, and respond by using the following scale from 1 to 7:

1
Strongly disagree

2
Disagree

3
Disagree somewhat

4
Neutral

5
Agree somewhat

6
Agree

7
Strongly agree

____ 1. Overall, my gender has very little to do with how I feel about myself. ____ 2. The gender I belong to is an important reflection of who I am. ____ 3. My gender is unimportant to my sense of what kind of a person I am. ____ 4. In general, belonging to my gender is an important part of my self image Past Experiences with Gender Discrimination Scale Past Experience with Gender Discrimination Scale Please read each statement carefully and circle the appropriate response ranging from strongly agree to strongly disagree. 1. I have personally been a victim of sexual discrimination 1 2 3 4 5 6 7 8 strongly disagree strongly agree 2. I consider myself a person who has been deprived of opportunities because of my gender. 1 2 3 4 5 6 7 8 strongly disagree strongly agree

67

3. I feel like I am personally a victim of society because of my gender 1 2 3 4 5 6 7 8 strongly disagree strongly agree 4. I have personally been the victim of sexual harassment. 1 2 3 4 5 6 7 8 strongly disagree strongly agree 5. I regularly encounter sexism against my gender 1 2 3 4 5 6 strongly disagree

7 8 strongly agree

6. Prejudice against my gender group has affected me personally. 1 2 3 4 5 6 7 8 strongly disagree strongly agree

68 Appendix B Climate for Intolerance for Discrimination Manipulation and Fillers (Formal, Positive CID) WWW.THEINSIDERSCOOP.COM
Company: RLK Consulting Description: Company Flowchart Source: www.RLKConsult/employment/upper_management/hk-36sh.com

RLK Consulting
Flowchart of Top Executives

Rachel Collins Chief Executive Officer

Brian Forrester Chief Financial Officer

Jacob Fletcher Vice President

Diane Baxter Vice President

Michele Brown Vice President

Richard Stanton Vice President

Meghan Rogers Director

Charles Bower Director

Steven Peters Director

Jane Berger Director

Andrew Martin Director

Julie Phillips Director

Paul Fisher Director

Lauren Miller Director

69 (Formal, Negative CID) WWW.THEINSIDERSCOOP.COM
Company: RLK Consulting Description: Company Flowchart Source: www.RLKConsult.com

RLK Consulting
Flowchart of Top Executives

Robert Collins Chief Executive Officer

Brian Forrester Chief Financial Officer

Jacob Fletcher Vice President

David Baxter Vice President

Michael Brown Vice President

Richard Stanton Vice President

Markus Rogers Director

Charles Bower Director

Steven Peters Director

John Berger Director

Andrew Martin Director

Jason Phillips Director

Paul Fisher Director

Logan Miller Director

70 (Informal, Positive CID) WWW.THEINSIDERSCOOP.COM
Company: RLK Consulting Description: Business Biweekly recently conducted a series of interviews

on the atmosphere at a variety of management consulting companies. Below are some of the quotes from employees at RLK Consulting.
Source: Business Biweekly

“I’ve only worked for RLK Consulting for a few weeks, so I don’t have any strong opinions. I’ll be able to tell you more in a few months.”

“The most powerful people in the company are accessible to everyone. They give just as much extra help and advice to female employees as they do to male employees.”

“The building security in the DC office is kind of a pain. It can take 10 minutes just to get through the door in the mornings. As only one of several tenants in the building I know that RLK can’t change the policy, but it’s still annoying.”

“RLK Consulting has great facilities. The lobbies of our buildings look like they could be in a luxury hotel.”

71 (Informal, Negative CID) WWW.THEINSIDERSCOOP.COM
Company: RLK Consulting Description: Business Biweekly recently conducted a series of interviews

on the atmosphere at a variety of management consulting companies. Below are some of the quotes from employees at RLK Consulting.
Source: Business Biweekly

“I’ve only worked for RLK Consulting for a few weeks, so I don’t have any strong opinions. I’ll be able to tell you more in a few months.”

“The most powerful people in the company have formed a boys’ club. They only give extra help and advice to male employees.”

“The building security in the DC office is kind of a pain. It can take 10 minutes just to get through the door in the mornings. As only one of several tenants in the building I know that RLK can’t change the policy, but it’s still annoying.”

“RLK Consulting has great facilities. The lobbies of our buildings look like they could be in a luxury hotel.”

72 (Filler 1) WWW.THEINSIDERSCOOP.COM
Company: RLK Consulting Description: Salary Information Source: www.RLKConsult/reports/salarydistribution/bw-93hr.com

RLK Consulting

2002 Fiscal Year Annual Salaries for Employees of RLK Consulting:

CEO/CFO: $234,576 Vice President: $175,899 Director: $147,254 Manager: $103,465 Senior Associate: $81,532 Associate: $41,678 Administrative Assistant: $29,438

Note: Salaries do not include bonuses yearly and are rounded to the nearest dollar. The salary reported for each job title reflects the mean annual salary of all employees in that job.

73 (Filler 2) WWW.THEINSIDERSCOOP.COM
Company: RLK Consulting Description: Company self-description Source: www.RLKConsult/intopag/description/eh-49ei.com

RLK Consulting

RLK Consulting is a midsized consulting firm that specializes in management consulting. The company was founded in 1985 when the first office was opened in Washington, D.C. While the company has remained local, we have since opened offices in Baltimore, MD, Arlington, VA, and most recently in Chevy Chase, MD. The company began as a team of 20 consultants, but by the mid 1990s it reached the current size of over 400 employees. Although RLK Consulting accepts consulting projects that span a diverse range of topics the company specializes in strategy, organization, technology, and operations. All employees of RLK work in one of four departments. Each department primarily handles projects falling into only one of these four categories. RLK Consulting also gives back to the community by offering pro-bono services to some non-profit organizations. Roughly 5% of the company’s projects are dedicated to pro-bono work.

74 (Filler 3) WWW.THEINSIDERSCOOP.COM
Company: RLK Consulting Description: Profit Information Source: www.RLKConsult/market/profits/quarterly/fi-41-ql.com

RLK Consulting
Average Quarterly Profit:
1993 - $356,876 1994 - $326,387 1995 - $335,743 1996 - $583,923 1997 - $496,209 1998 - $593,399 1999 - $639,398 2000 - $693,298 2001 - $294,943 2002 - $523,293

Note: The numbers are reported as net profits. The numbers reflect the mean profit for each quarter for the year. All figures are rounded to the nearest dollar.

75 (Filler 4) WWW.THEINSIDERSCOOP.COM
Company: RLK Consulting Description: Departmental Breakdown - Each of the four departments within

RLK Consulting represents an area of expertise shared by the members of that department. Strategy is the largest department at RLK Consulting because most of the company’s clients seek strategy advice for management.
Source: www.RLKConsult/deparments/employees/bydivision/ke-35jr.com

RLK Consulting
Departmental Breakdown of RLK Consulting Employees:

Department

Strategy Organization Technology Operations

Number of Vice Pres. 1 1 1 1

Number of Directors 2 2 2 2

Number of Managers 27 17 8 8

Number of Senior Associates 45 28 11 11

Number of Associates 88 54 26 24

Number of Admin. Assistants 34 21 9 8

Total

197 123 57 54

76 Appendix C List of Words for Construct Accessibility Measure TARGET WORDS
SEXISM BIAS DISCRIMINATION PREJUDICE UNJUST INEQUALITY PARTIAL

TARGET NONWORDS
MEIXSS SIAB CIMINATIDRSION JUPRECIDE NUJTUS QUIANIELTY TRIAPAL

NONTARGET WORDS
CENTER LIKE CORRESPONDENCE DETERMINE EXCUSE MODIFIED CONNECT RABBIT HOPE COMMUNICATIONS HAPPINESS CAVEAT CREATIVITY CONCERT BOTTOM BIKE MULTIPLICATION SELECTION OUTLET SIMILARITY RESOLVE RELOAD MICE COMMERCIALIZED COMPLETED BONNET PURPOSEFUL PIANIST WONDER HEAT TRANSFORMATION REPLACING SIGNAL INNOVATION RECEIVE MISUSE BLUE MISCONCEPTIONS CARPETING REMOVE PACKAGING EMPATHY

NONTARGET NONWORDS
NERTCE EIKL ROCONCERSENPED DINTEMEER SECUXE FDOMEIDI NETCONC TRBIBA PEHO ONIUOSNMTCICAM ESSHPIPAN VEAACT YRCIVATIET OCNRCET TTBMOO KBIE ONPLACUTIITMIL SLCNEOTEI UEOTLT ITAIRSIMLY SVERLOE ELDROA ICEM MEDCEZILMRIOAC ECEDTLMOP NOETBN FRLPSUERPO TNPSIAI ERWDNO EAHT FTMRNOITAONRSA CIRALPENG NALSGI AONNTIVOIN EIREVEC SSMIUE UELB TSPNEOCISCNMIO PRCTEAIGN EOERVM NIGAAKCPG YHTMPAE

77 Appendix D Application

RLK Consulting
APPLICATION FOR CO-MANAGER POSITION PART I: BACKGROUN D INFORMATION

Fill in the blank or circle your response as appropriate. Please indicate your score on the word decision task: Name: Gender: Age: Year: ___________________ male female ______________

___________________ freshman sophomore junior senior

Expected Date of Graduation: Major:

___________________

___________________

78

RLK Consulting
APPLICATION FOR CO-MANAGER POSITION PART II: ORGANZIATIONAL DECISION MAKING Please read each of the following four scenarios carefully. At the end of each scenario you will be asked to make a decision. Please indicate your decision by checking your answer.

FROM: [email protected] RE: Location of new office I am writing to get your opinion. As you know, we are going to build a new office in order to expand our company. We can open the office in New York or Chicago. The office space costs less in Chicago, but the taxes will be higher for the first five years of business. After five years taxes will be the same in either city. In New York the office space costs more, but we get a five year tax break. The money really equals out overall if we figure an average level of business over the next five years. I would like your opinion on what we should do. R. Collins, CEO

I MUST NOW CHOOSE BETWEEN THE FOLLOWING TWO OPTIONS: ____ I will suggest we open the new office in New York. If we open an office in New York we get the tax break, and if we do more than average business then we gain by the reduced taxes. I will suggest we open an office in New York.

_____ I will suggest we open the new office in Chicago. If we open an office in Chicago we get the office at a better price, and if we do less than average business then we at least gain by purchasing the less expensive office space. I will suggest we open an office in Chicago.

79

RLK Consulting

FROM: [email protected] RE: Hiring Decision As you may know, one of the associates in our group had to resign last month because of medical problems. Due to the high volume of business our group has been handling recently, we need to fill the position immediately. Human resources has sent me the resumes of 72 applicants for the position. Due to the tight economy, it seems that we have an abundance of highly qualified applicants. We do not have time to interview all of these candidates. Realistically, we can only interview 20 people if we want to fill the position within the month. In determining which of these candidates will receive interviews we need to decide whether we want to emphasize either performance in business school or past experience in the consulting industry. Please let me know which strategy you recommend. J. Foley, Senior Associate

I MUST NOW CHOOSE BETWEEN THE FOLLOWING TWO OPTIONS

_____ I believe that individuals with a lot of experience in the field have a wider source of knowledge about the industry and know more about the day-to-day life of a consultant. We want to hire people that will be able to hit the ground running. It is my opinion that experience is the key to success as a consultant. I will recommend that we emphasize experience over performance in business school when ranking the candidates. _____ I believe that performance in business school is the best indication of pure intellect. While experience helps to develop a consultant, intelligence is what determines a consultant’s ultimate success. It is my opinion that in the long run the candidates who were the most successful in business school will make the best consultants. I will recommend that we emphasize performance in business school over experience when ranking the candidates.

80

RLK Consulting
FROM: [email protected] RE: Plan for Attracting New Business The company is currently debating what the best plan is for attracting new clients to our company. I have been placed in charge of developing different strategies and surveying employees’ opinions of these strategies. I would like your opinion on which of two general strategies you believe to be more effective. The first strategy involves seeking contracts to do large-scale projects. Developing presentations to use to solicit business will be fairly time intensive as large projects must be tailored to the needs of each specific company. If we do get clients to sign with us on big projects, each project will be hugely profitable. The second strategy involves focusing our efforts on contracts to do small-scale projects that address common problems in companies. We would not have to spend much time developing presentations for each company we want to solicit business from because many organizations can often benefit from the same or similar small-scale projects. The payoffs from smaller contracts, however, are not as profitable. Please let me know which of these strategies you think will be more profitable for RLK Consulting. T. Green, Senior Associate

I MUST NOW CHOOSE BETWEEN THE FOLLOWING TWO OPTIONS

_____ If we seek contracts for large projects, the company will get big payoffs. Big payoffs are the only way to have a really profitable business. Although it will take more time for us to develop presentations to pitch to specific companies, we will only need to sign a few contracts in order to be profitable. The effort required to sign enough small contracts to be as profitable as signing just a few large contracts would not be worth the effort. I will tell the senior associate that I think soliciting large contracts will be more profitable. _____ Spending a lot of time developing a specific presentation to pitch to a company in hopes of signing a contract for a large project is not a wise business move. If the contract is not signed, the company has wasted a lot of time and energy. It will not take long to develop a single presentation for a small project that we can pitch to lots of companies. Even if we only sign contracts with some of the companies we solicit business from, our efforts will not have been wasted. I will tell the senior associate that I think soliciting small contracts will be more profitable.

81

RLK Consulting
APPLICATION FOR CO-MANAGER POSITION PART III: PERSONAL STATEMENT Please use the following space to provide the manager with a few sentences describing why you are a good candidate for the position of co-manager.

82 Appendix E Pre-feedback Questionnaire CONFIDENTIAL QUESTIONNARE Your responses to this questionnaire will remain strictly confidential. This questionnaire will not be shown to other employees of RLK Consulting. When you finish this questionnaire please place it in the envelope marked CONFIDENTIAL.

1. How well do you think you performed on the computer task that asked you to indicate whether or not a string of letters formed a word? 1 very poorly 2 somewhat poorly 3 okay 4 somewhat well 5 very well

2. How well do you think you performed on the task that asked you to make four organizational decisions? 1 very poorly 2 somewhat poorly 3 okay 4 somewhat well 5 very well

3. How well do you think you performed on the personal statement part of the application for the role of co-manager? 1 very poorly 2 somewhat poorly 3 okay 4 somewhat well 5 very well

4. Would you prefer to stay in the role of associate or to be promoted to the role of comanager? 1 strongly prefer associate 2 somewhat prefer associate 3 4 strongly prefer 5

no preference somewhat prefer co-manager

co-manager

83 Appendix F Feedback Form

RLK Consulting
Form PD_07 Promotion Decision Form

Manager:

Kevin Bannister Brian MacDonald Participant’s Name

Co-Manager:

Associate:

Manager’s Comments: __The girl did not come across well on the application. I_____ chose the guy for the role of co-manager because I don’t think I would work well with her.__________________________________________________________________

84 Appendix G Measurement of Perceiving

CONFIDENTIAL QUESTIONNAIRE
Your responses to this questionnaire will remain strictly confidential. This questionnaire will not be shown to other employees of RLK Consulting. When you finish the questionnaire please place it in the envelope marked CONFIDENTIAL. Please use the following space to indicate why you think you were assigned to your current role. ________________________________________________________________ ________________________________________________________________ ________________________________________________________________ Please use the following scale to respond to questions 1 – 9. 1 not at all 2 probably not 3 unsure 4 probably 5 definitely

_____ 1. Do you think the manager will be effective in the role of manager? _____ 2. Do you think the manager will do a good job dealing with real world organizational dilemmas? _____ 3. If you met the manager, do you think you would like the manager? _____ 4. Do you think you are similar to the manager? _____ 5. Do you think that the decision made by the manager to assign you to your current role was fair? _____ 6. Do you think the decision made by the manager to assign you to your current role was due to your application? _____ 7. Do you think the decision made by the manager to assign you to your current role was due to the manager’s personal preferences? _____ 8. Do you believe the decision made by the manager to assign you to your current role was due to your gender? _____ 9. Do you believe the decision made by the manager to assign you to your current role was due to sexism?

85 Appendix H Measurement of Reporting (Reporting Measure)

RLK Consulting
Form OG_02 Organizational Grievance Form

This form is used whenever employees want to file a complaint or grievance with RLK Consulting. Employees file grievances when they think they have experienced unfair treatment by a member or client of RLK Consulting. Grievances are not reported to managers, but are handled one of the Vice Presidents. You are not required to fill out the form, but please feel free to do so if you have any complaints about the company.

1. Would you like to file a grievance with RLK Consulting at this time? Please check a box. Yes No

2. If you responded yes above, what is the nature of your complaint? _______________________________________________________ _______________________________________________________ _______________________________________________________ _______________________________________________________ 3. What action would you like to see taken as a result of filing a grievance? _______________________________________________________ _______________________________________________________ _______________________________________________________ When you are finished with this form, please place in your RLK Consulting outbasket.

86 (Filler item)

RLK Consulting
Form CC_01 Company Charity Questionnaire

RLK Consulting is thinking of selecting a company wide charity organization to increase the ways in which we give back to the community. RLK would make an annual contribution to this charity and also provide employees with the opportunity to make individual contributions. Employees would be able to vote on what type of charity RLK Consulting would adopt. Before implementing this plan, we want to assess support among employees to determine if it’s a worthwhile endeavor.

1. Would you support RLK Consulting adopting a company-wide charity organization? Yes No

2. Do you think it’s a good idea for RLK Consulting to make an annual contribution to a charity organization? Yes No

3. If you approved of the charity selected by the company, would you be willing to make an annual individual contribution to the charity? Yes No

4. If so, how much would you be likely to donate annually? ___ ___ ___ ___ ___ ___ $0 – 20 $21 – 40 $41 – 60 $61 – 80 $81 – 100 More than $100

When you are finished with this form, please place in your RLK Consulting outbasket.

87 (Filler item)

RLK Consulting
Form BP_03 Benefits Plan Change Form

RLK Consulting is thinking about revising the vacation and sick days aspects of the benefits plans. Currently, new employees initially receive 16 vacation days and 4 sick days annually. Any unused days roll over to the next year. Please vote which of the following changes to the current plan you would most favor. ___ ___ ___ ___ 10 vacation days, 4 sick days, and an extra $2,000 annually in salary 10 vacation days and 10 sick days 16 vacation days, 4 sick days, and employees receive monetary compensation for any unused days at the end of the year. No change

When you are finished with this form, please place in your RLK Consulting outbasket.

88 Appendix I Manipulation/Suspicion Check CONFIDENTIAL QUESTIONNARE Your responses to this packet of questionnaires will remain strictly confidential. These questionnaires will not be shown to other employees of RLK Consulting. When you finish the packet please place it in the envelope marked CONFIDENTIAL.
1. What do you think the purpose of this experiment is? _____________________________________________________ _____________________________________________________ 2. Did you find any aspect of the experiment odd or suspicious? _____________________________________________________ _____________________________________________________ 3. Did you find any aspect of the experiment disturbing? _____________________________________________________ _____________________________________________________ 4. Did you enjoy this experiment? _____________________________________________________ _____________________________________________________ 5. Do you have any other comments? _____________________________________________________ _____________________________________________________ 6. Do you believe that you were deceived in this study in any way? If so, please explicitly explain how you think you were deceived. _____________________________________________________ _____________________________________________________ _____________________________________________________ _____________________________________________________

Please circle your responses to the following questions.

89

1. Approximately how many employees work for RLK Consulting? 100 200 300 400 500 600 700 800 900 1000

2. How many departments are there at RLK Consulting? 1 2 3 4

5

3. Generally speaking, how would you characterize the quarterly profits of RLK Consulting? small medium large 4. Generally speaking, what type of environment does this company provide for women? negative neutral positive 5. How many Vice Presidents are there at RLK Consulting? 1 2 3 4 6. How many offices does RLK Consulting have? 1 2 3 4

5

5

7. How would you characterize the ethics of business conducted by RLK Consulting? Ethical Neutral Unethical 8. How many women are there in the top management of RLK Consulting None 50% women 100%

90

1. Were you assigned the role of manager?

Yes

No

2. If not, what was the name of the person assigned the role of manager? __________________ 3. If you cannot remember this person’s name, was the manager male or female? __________________ 4. Were you assigned the role of co-manager? Yes No

5. If not, what was the name of the person assigned the role of co-manager? _____________________________________________________ 6. If you cannot remember this person’s name, was the co-manager male or female? _____________________________________________________ 7. Were you assigned the role of associate? Yes No

8. If not, what was the name of the person assigned the role of associate? _____________________________________________________ 9. If you cannot remember this person’s name, was the associate male or female? _____________________________________________________

Please place this packet of questionnaires in the CONFIDENTIAL envelope

91 Footnote
1

The statistics reported for the logistics regression analyses include
2

2

, Cox and

Snell R2, the Wald statistic, and Exponential B. The

statistic provides a test of the

overall significance of the model and the change in significance of the model from step to step. The Cox and Snell R2 provides a measure of effect size, however the maximum value is .75 instead of 1. The Wald statistic and accompanying significance level provides a test of each effect included at each step and is calculated by dividing the squared regression coefficient by the squared standard error of the regression coefficient. Finally, Exponential B is a likelihood ratio calculated by raising e to the value of the regression coefficient. Exponential B estimates the change in the odds of the occurrence of the dependent variable in response to a one unit change in the independent variable. Therefore values greater than 1 indicate an increased likelihood of the dependent variable and values less than 1 indicate a decreased likelihood of the dependent variable.

92 References Bailey, J. J., & Alexander, R. A. (1993). Organizational social cues, framing, and justice: Effects on management’s ethical decisions. International Journal of Organizational Analysis, 1, 133-160. Barak, M. E. M., Cherin, D. A., & Berkman, S. (1998). Organizational and personal dimensions in diversity climate: Ethnic and gender differences in employee perceptions. Journal of Applied Behavioral Science, 34, 82-104. Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychology research: Conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182. Branscombe, N. R., Owen, S., Kobrynowica, D., (1993). Measuring identification with one’s gender group in women and men. Paper presented at the American Psychological Association, Toronto, Canada. Brief, A. P., Dietz, J., Cohen, R. R., Pugh, S. D., & Vaslow, J. B. (2000). Just doing business: Modern racism and obedience to authority as explanations for employment discrimination. Organizational Behavior and Human Decision Processes, 81, 72-97. Burke, M. J., Borucki, C. C., & Hurley, A. E. (1992). Reconceptualizing psychological climate in a retail service environment: A multiple stakeholder perspective. Journal of Applied Psychology, 77, 717-729. Chatman, J. A., Polzer, J. T., Barsade, S. G., & Neale, M. A. (1988). Being different yet feeling similar: The influence of demographic composition and organizational culture on work processes and outcomes. Administrative Science Quarterly, 43, 749-780. Cohen, L. L, & Swim, J. K. (1995). The differential impact of gender ratios on women and men: Tokenism, self-confidence, and expectations. Personality and Social Psychology Bulletin, 21, 876-884. Cox, T. (1993). Cultural Diversity in Organizations: Theory, Research and Practice. Berrett-Koehler Publishers: San Francisco. Crocker, J., Voelkl, K., Testa, M., & Major, B. (1991). Social stigma: The affective consequences of attributional ambiguity. Journal of Personality and Social Psychology, 60, 218-228. (check) Crosby, F., Bromley, S., & Saxe, L. (1980). Recent unobtrusive studies of Black and White discrimination and prejudice: A literature review. Psychological Bulletin, 87, 546-563.

93

Dovidio, J. F. Evans, N. & Tyler, R. B. (1986). Racial stereotypes: The contents of their cognitive representations. Journal of Experimental Social Psychology, 22, 22-37 Dovidio, J. F., & Gaertner, S. L. (1998). On the nature of contemporary prejudice: The causes, consequences and challenges of aversive racism. In J. L. Eberhardt and S. T. Fiske (Eds.) Confronting racism: The problem and the response (pp. 3-32). Thousand Oaks, CA: Sage Publications, Inc. Dovidio, J. F., Kawakami, K., & Gaertner, S. L. (2002). Implicit and explicit prejudice and interracial interaction. Journal of Personality and Social Psychology, 82, 6268. Fitzgerald, L. F., Drasgow, F., Hulin, C. L., Gelfand, M. J., & Magley, V. J. (1997). Antecedents and consequences of sexual harassment in organizations: An integrated model. Journal of Applied Psychology, 82, 578-589. Friedman, R., & Davidson, M. N. (1999). The Black-White gap in perceptions of discrimination: Its causes and consequences. In R. J. Bies, R. J. Lewicki, et al. (Eds.). Research in negotiation in organizations, 7 (pp. 203-228). Gwartney-Gibbs, P. A., & Lach, D. H. (1994). Gender and workplace dispute resolution: A conceptual and theoretical model. Law and Society Review, 28, 265-296. Hicks-Clarke, d., & Iles, P. (2000). Climate for diversity and its effects on career and organisational attitudes. Personnel Review, 3, 324-345. James, E. H., Brief, A. P., Dietz, J., & Cohen, R. R. (2001). Prejudice matters: Understanding the reactions of whites to affirmative actions programs targeted to benefit blacks. Journal of Applied Psychology, 86, 1120-1128. Judd, C. M., & Kenny, D. A. (1981). Process analysis: Estimating mediation in treatment evaluations. Evaluation Review, 5, 692-619. Kenny, D. A., Kashy, D. A., & Bolger, N. (1998). Data analysis in social psychology. In D. Gilbert, S. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (Vol. 1, 4th ed., pp. 233-265). Boston, MA: McGraw-Hill. Kossek, E. E., & Zonia, S. C. (1993). Assessing diversity climate: A field study of reactions to employer efforts to promote diversity. Journal of Organizational Behavior, 14, 61-81. Luhtanen, R., & Crocker, J. (1992). A collective self-esteem scale: Self-evaluation of One’s social identity. Personality and Social Psychology Bulletin, 18, 302-318. Major, B., Gramzow, R. H., McCoy, S. K., Levin, S., Schmader, T., & Sidanius, J.

94 (2002). Perceiving personal discrimination. Journal of Personality and Social Psychology, 82, 269-282. Mannix, E. A., Neale, M. A., & Norhtcraft, G. B. (1995). Equity, equality, or need? The effects of organizational culture on the allocation of benefits and burdens. Organizational Behavior and Human Decision Processes, 63, 276-286. Meyer, D. E., & Schvaneveldt, R. W. (1971). Facilitation in recognizing pairs of words: Evidence of a dependence between retrieval operations. Journal of Experimental Psychology, 90, 227-234. Nishii, L. H., & Raver, J. (2001, June). Examining contextual predictors of organizational conflict: Climate for diversity. In L. Nishii (chair), Conflict in Groups and Organizations. Symposium conducted at the International Association for Conflict management annual conference, Paris, France. Olson-Buchanan, J. B. (1996). Voicing discontent: What happens to the grievance filer after the grievance. Journal of Applied Psychology, 65, 96-102. O’Neil, J. M., Egan, J., Owen, S., & Murry, V. M. (1993). The gender role journey measure: Scale development and psychometric evaluation. Sex Roles, 38, 167185. Pelled, L. H., Ledford, G. E. Jr., & Mohrman, S. A. (1999). Demographic dissimilarity and workplace inclusion. Journal of Management Studies, 36, 10131031. Perry, E. L., Davis-Blake, A., Kulik, C. T., (1994). Explaining gender-based selection decision: A synthesis of contextual and cognitive approaches. Academy of Management Review, 19, 786-820. Ragins, B. R., & Cornwell, J. M. (2001). Pink triangles: Antecedents and consequences of perceived workplace discrimination against gay and lesbian employees. Journal of Applied Psychology, 86, 1244-1261. Ratcliff, R. (1993). Methods for dealing with reaction time outliers. Psychological Bulletin, 114, 510-532. Rudman, L. A., & Borgida, E. (1995). The afterglow of construct accessibility: The behavioral consequences of priming men to view women as sexual objects. Journal of Experimental Social Psychology, 31, 493-517. Schmitt, M. T., Branscombe, N. R., Kobrynowicz, D., & Owen, S. (2002). Perceiving discrimination against one’s gender group has different implications for wellbeing in women and men. Society for Personality and Social Psychology, 28, 197210.

95

Schneider, B. (1975). Organizational climates: An essay. Personnel Psychology, 28, 447479. Schneider, B., Bowen, D. E., Ehrhart, M. G., & Holcombe, K. M. (2000). The climate for service: Evolution of a construct. In N. M. Ashkanasy, C. P. M. Wilderom, & M. R. Peterson (Eds.), Handbook of organizational culture and climate (pp. 21-36). Thousand Oaks, CA: Sage. Schneider, K. T., Hitlan, R. T., & Radhakrishnan, P. (2000). An examination of the nature and correlates of ethnic harassment experiences in multiple contexts. Journal of Applied Psychology, 85, 3-12. Schneider, B., & Reichers, A. E. (1983). On the etiology of climates. Personnel Psychology, 36, 19-39. Schneider, B., Smith, D. B., Sipe, W. P. (2000). Personnel selection psychology: multilevel considerations. In K. J. Klein & S. W. J. Kozlowski (Eds.), Multilevel theory, research, and methods in organizations. San Francisco: Jossey-Bass. Sechrist, G. B., & Stangor, C. (2001). Perceived consensus influences intergroup behavior and stereotype accessibility. Journal of Personality and Social Psychology, 80, 645-654. Stangor, C., Sechrist, G. B., & Swim, J. K. (1999). Sensitivity to sexism and perceptions of reports about sexist events. Swiss Journal of Psychology, 58, 251-256. Stangor, C., Swim, J. K., Sechrist, G. B., Van Allen, K., & Ottenbreit, A. (in press). Perceiving prejudice. European Review of Social Psychology. Stangor, C., Swim, J. K., Van Allen, K. L., & Sechrist, G. B. (2002). Reporting discrimination in public and private contexts. Journal of Personality and Social Psychology, 82, 69-74. Swim, J. K., Aikin, K. J., Hall, W. S., & Hunter (1995). Sexism and racism; Oldfashioned and modern prejudices. Journal of Personality and Social Psychology, 68, 199-214. Swim, J., & Hyers, L. (1999). Excuse me--What did you just say?!: Women's public and private responses to sexist remarks. Journal of Experimental Social Psychology, 35, 68-88. Wittenbrink, B., Judd, C. M., & Park, B. (1997). Evidence for racial prejudice at The implicit level and its relationship with questionnaire measures. Journal of Personality and Social Psychology, 72, 262-274.



doc_281899471.pdf