"Intuitive Analysis of Covariance" and Group Stereotype Formation

Abstract

There exist actual correlations between group membership and behavior that are purely spurious, resulting from very different situational constraints on group members' behavior. Accurate group inferences require that people recognize and control for the differing situational constraints by using reasoning analogous to a statistical analysis of covariance. An initial study demonstrated that people often fail to engage in such reasoning and consequently form erroneous group impressions. A second experiment tested two hypotheses concerning the conditions that may facilitate more sophisticated covariation reasoning. Results indicated that (a) relaxed temporal constraints and (b) explicit attention to situational differences both facilitate 'intuitive analysis of covariance" and more accurate group inferences. Discussion focuses on implications for understanding the formation of group stereotypes and on factors that may inhibit stereotype formation.