Outgroup Homogeneity Effect

Background Research

In some of the first work on the outgroup homogeneity effect, men and women were asked to rate men and women on positive and negative [Page 632]dimensions that were stereotypically masculine or feminine. Results showed that over and above any overall ingroup preference (i.e., rating one's own group more positively than the outgroup), participants judged the outgroup more stereotypically than the ingroup. More specifically, when participants were asked what percentage of each group had attributes that were consistent with the group's stereotype and what percentage had attributes that were inconsistent with that stereotype, they reported that the outgroup had relatively more consistent group members and fewer inconsistent group members than did the ingroup. These results represent strong support for the outgroup homogeneity effect because ratings were collected from both groups, and the effect was found for the men's ratings and for the women's ratings. In addition, the effect was found on positive as well as on negative stereotypic attributes, indicating that it is independent of any tendency to see the outgroup as having relatively more members who are consistent with only negative attributes. And finally, the use of gender groups permits one to conclude that the effect exists even when familiarity is high for both the ingroup and the outgroup.

The outgroup homogeneity effect has been replicated with a wide variety of social groups in addition to gender-defined groups: sororities, experimentally created groups, and groups defined by age, nationality, and ethnicity. A meta-analysis of the effect across published studies concluded that the outgroup homogeneity effect is a small but nevertheless robust effect. In addition, it appears that the effect is smaller with experimentally created groups than with real groups.

One interesting research development has been the identification of two different forms of perceived group variability and, therefore, of the outgroup homogeneity effect. The first component of perceived variability is the degree to which a group is seen as having a relatively large number of people who confirm the stereotype, compared with the proportion of those who do not. This component, referred to as perceived stereotypicality, was originally examined when outgroup homogeneity was demonstrated with groups defined by gender, as described above. The other component is the perceived dispersion of a group, that is, the extent to which group members vary around what is perceived to be the group mean on an attribute dimension. To make clear this distinction, imagine that one person sees a group on average more stereotypically than another person does. It might still be the case that they both agree on the actual variability within the group around the group average. Stereotypicality refers to the extremity of the group mean on stereotypic attributes. Dispersion refers to the perceived variability around that group mean. It has been demonstrated that these two components of perceived group variability need not be highly related to each other. However, outgroup homogeneity has been shown for both components, although it appears to be larger for perceived stereotypicality than for perceived dispersion.

...