| This study extended the research on Kanter's tokenism theory by addressing the influence of group proportion on task-specific self-evaluations and global self-esteem. One hundred eighty female and 1554 male Army captains working in small real groups of 12--14 rank ordered each other, and themselves, on leadership ability. The survey, taken at two intervals (6 weeks apart), also assessed their global self-esteem, perceptions of tokenism, commitment to the army and self-criticism. Results indicated that female Army captains in disproportionate groups reported lower self-evaluations than males at both T1 and T2. The number of women per staff group was not related to females' self-evaluations, peer evaluations, or global self-esteem at either T1 or T2. Additionally, there were no significant differences in peer evaluations or global self-esteem between males and females and no differences between solo and non-solo females on these variables. Females reported more tokenism at T1 and less commitment at T2 than males. Furthermore, there was a significant interaction between tokenism and commitment in predicting global self-esteem for all participants. Finally, the self-evaluations of solo females and females in the three women per staff group condition appeared to be more influenced by peer evaluations than the self-evaluations of females in the two women per staff group condition and males. |
