Group Evaluations among Members of Disadvantaged Groups

Abstract

Members of disadvantaged groups sometimes display outgroup favoritism, an evaluative preference for an outgroup over the ingroup. Some scholars argue that such outgroup favoritism is common among members of disadvantaged groups, and that it is caused by an internalization of negative stereotypes. The present dissertation critically reflects upon these ideas. Specifically, this work argues that empirical findings regarding group evaluations among members of disadvantaged groups are more mixed than previously acknowledged, and that this heterogeneity is possibly due to moderators. The first study investigated moderators of group evaluations, derived from system justification theory (SJT; Jost, Banaji, & Nosek, 2004). We used a meta-analytic approach with large samples of online participants, spanning 8 social identities and 14 nations, and including Implicit Association Tests (IAT; Greenwald, McGhee, & Schwartz, 1998) and self-report measures as dependent variables. We observed that disadvantaged groups displayed outgroup favoritism on the IAT, but no group preference or ingroup favoritism on self-report measures. However, effects were highly heterogeneous and exploratory moderator analyses revealed that social identity moderated group evaluations in disadvantaged groups: Whereas some disadvantaged groups always displayed ingroup favoritism, other disadvantaged groups always displayed outgroup favoritism, and yet others displayed divergent patterns on IATs and self-report measures. Furthermore, group-based stigma and self-reported conservatism moderated group evaluations. A second set of studies investigated ingroup typicality as another potential moderator of group evaluations. We assumed that members of disadvantaged groups who perceive themselves as less typical for their ingroup may be more likely to demonstrate outgroup favoritism. In Study 1 and 2, Black participants with lighter skin tone more strongly preferred White relative to Black people. In Study 3, heavyweight participants with lower body weight more strongly preferred thin relative to heavyweight people. In Study 4, participants with less visible disabilities more strongly preferred non-disabled relative to disabled people. A meta-analysis across studies estimated an overall small effect of ingroup typicality on group evaluations. A third set of studies investigated procedural differences between measures as another potential moderator of group evaluations. In Study 1, Turkish-German participants displayed a preference for Turkish relative to German on two IAT variants, but no preference for either group on feeling thermometers. In Study 2, Muslim participants displayed preferences for Arabs and Muslims relative to Whites on two variants of the Affect Misattribution Procedure (AMP; Payne, Cheng, Govorun, & Stewart, 2005) and on self-report measures, suggesting ingroup favoritism. Muslim participants also created classification images of a typical Muslim via a reverse correlation task. Compared with classification images created by a control sample, those created by Muslim participants were rated higher in trustworthiness, but did not differ on other dimensions. Taken together, findings do not suggest ubiquitous patterns of outgroup favoritism, but highlight the importance of moderators for our understanding of group evaluations among members of disadvantaged groups.