The impact of gender on perceptions of group homogeneity was assessed using a memory paradigm. In the main experimental conditions, male and female subjects were presented with 12 pairs of photographs and gendered silhouettes. The photographs portrayed a series of environments (e.g. a kitchen) in public settings (e.g. a restaurant) and in domestic ones (e.g. at home). The connection of settings and gender of silhouettes varied across subjects. In control conditions, all silhouettes were of the same sex or were geometrical forms (triangles and squares). The subjects memorized the pairs of photographs and targets, and then had to reconstruct them. The experiment addressed three hypotheses. The categorization effect: within-category errors would exceed betweencategory errors, especially when gender was used as a categorization criterion. The out-group homogeneity effect: within-category errors concerning the out-group would be greater than those concerning the in-group. The social status effect: men would display more out-group homogeneity than women. Results provided evidence for the categorization and the out-group homogeneity hypotheses. They also supported to some extent the social status hypothesis. Different styles of processing of information were suggested by men's and women's performances. Men consistently displayed the outgroup homogeneity bias, whereas women did not. Women took into account the kind of setting: they made more errors when matching the male or the female silhouettes to the domestic settings (the stereotypical in-group settings) than when matching these silhouettes to the public settings (the stereotypical out-group settings). Finally, sex-typing measures about the self were related to the subjects' behavioural performances, especially for women.Out-groups tend to be perceived as less variable and less differentiated than in-groups:They all look alike, but we don't. Several explanations of this out-group homogeneity effect have been proposed. Some of them stress the perceivers' familiarity with in-group members (Linville, Salovey & Fischer, 1986), others focus on the nature of the intergroup interactions (e.g. Wilder, 1984) and others on individuals' cognitive processes (e.g. Park & Rothbart, 1982; see Quattrone, 1986, for a review).Despite the abundance of arguments sustaining the pervasiveness of this effect, authors do not consistently argue in favour of the out-group homogeneity effect. Self-categorization theory, for instance, posits an equal balance of in-group and out-group homogeneities *Requests for reprints.
112Fabio Lorenzi-Cioldi when intergroup comparisons between oneself and others are concerned (Turner, Hogg, Oakes, Reicher & Wetherell, 1987). Moreover, inspection of the empirical literature reveals a puzzling pattern of results. Perceptions of out-group homogeneity vary greatly in magnitude and sometimes in-groups are more homogeneous than out-groups. Various mediating factors which account for these inconsistencies have been reported: interpersonal attraction (Stephan, 1977...