The Effects of Group Affiliation Versus Individuating Information on Direct and Indirect Measures of the Evaluation of Novel Individual Group Members

Abstract

Automatic evaluation has emerged as a central concept in contemporary thinking about prejudice. The current research tested a quintessential aspect of prejudice: whether group affiliation dominates the automatic evaluation of individual group members even when diagnostic evaluative information about the individuals is available. Participants read a list of descriptions about the behaviors of two individuals: one from a typically liked group and one from a typically disliked group. The list portrayed one individual more positively than the other, and we manipulated the extremity and direction of that difference. We conducted six studies ( N = 11,572) with samples consisting of U.S. adults across different regions and group types (age, gender, and race) and two indirect measures that purportedly measure automatic evaluation: the implicit association test (IAT) and the evaluative priming task (EPT). Group affiliation (relative to personal characteristics) influenced the IAT and the EPT more than it influenced the self-reported evaluation. These results may suggest that the automatic evaluation of individuals is more prejudiced than nonautomatic evaluation.