Effect of moral stereotypes on the formation and persistence of group preferences

Do stereotypes have a stronger and more persistent effect on impressions when they are moral in tone? In two experiments (N = 187), participants interacted with members of two groups in an interactive social decision task. Prior to the task, participants were exposed to positive or negative group stereotypes that were moral or nonmoral in content. Although players from each group were, on average, equally likely to provide reward feedback, participants formed choice preferences for players from positively-stereotyped over negatively-stereotyped groups. Importantly, this effect was stronger and more resistant to change when stereotypes contained moral content. Computational modeling indicated that moral stereotypes induced more extreme initial reward expectancies and influenced how reward associations were updated over time. Additionally, moral stereotypes generalized more strongly to subsequent evaluations of novel group members, suggesting that the biasing effect of moral stereotypes on learning contributed to group-level prejudice.