Power, Sex, and Social Identity Theory

Rochelle Louise Cole


Three research methods were employed to investigate the intergroup behavior of men and women using Social Identity Theory (SIT; Tajfel & Turner, 1986) as a conceptual framework. In the first study, an extensive survey was administered to 105 male and 105 female undergraduates. Among several important findings, subjects perceived power differentials in favour of the male group. However, both male and female subjects identified strongly with their gender group and had a very positive gender social identity. Other key findings demonstrated that group power was very important to both male and female undergraduates.

In the laboratory, a variant of the Minimal Group Paradigm was used to investigate the effect of power and sex on the behavior of undergraduates as members of the same-sex (N=346) and opposite-sex groups (N=341). The main dependant measure was subjects' allocations using the Tajifel matrices. As in a power study by Sachdev and Bourhis (1985) in which sex was not salient, both male and female group members with power generally discriminated against outgroup members, whereas group members without power, did not. These findings are in contrast to Williams’' (1984) notion that men have a more competitive orientation than women and would thus be more discriminatory. Furthermore, regardless of subjects' sex, power contributed towards a more positive social identity. Overall, although subtle effects of sex were obtained, power had a strong impact on intergroup behavior and subjects' social identity.

For the field study, 79 members of two sex-segregated labour federations were interviewed. As expected from SIT, female members of the dominant federation had a more positive social identity than did male members of the subordinate federation. Reasons for the behavior of female members of an intermediary group who were attempting to 'pass' from the female to the male group were investigated. Taken together, evidence from these studies demonstrated that power had a greater impact on intergroup behavior than did hypothesized sex-specific orientations, identification with the gender group, or attraction to opposite-sex group members.