False consensus and out-group homogeneity: a methodological note on their relationship

Two different lines of research in social psychology ask subjects to generate estimates of the percentage of group members who they think would espouse some position or make some behavioural choice. Research on the false consensus effect has shown that such estimates are influenced by the subject's own choice. Research on out-group homogeneity has shown that such estimates are higher for group stereotypic choices than counterstereotypic ones and that this difference in turn is more pronounced for out-groups than in-groups. In this article we explore the relationship between these two effects, both of which rely on the same dependent measure. We show that out-group homogeneity will be estimated with bias unless the subject's own choice is included as a factor in the analysis, whenever false consensus is present and whenever the two target groups differ in the prevalence of their actual choices. We demonstrate this confounding with both hypothetical and actual data.