女性管理職に対する態度(WAMS)と女性リーダーシップの評価に関する研究

The present study was designed to examine effects of the sex of a leader and of a respondent on evaluation of leader behavior from the viewpoint of an implicit leadership theory and theories on sex-role stereotyping. Five hypotheses were derived based on the examination of the past studies as follows. (1) The leader sex does not affect factor structure of the leadership instrument. (2) The respondent sex does not affect factor structure of the leadership instrument. (3) The level of leader behavior evaluated by respondents differs according to the sex of a leader, which reflects a bias based on the sex-role stereotyping. (4) The sex of a respondent interacts with the leader sex in determining the level of leader behavior evaluation. (5) A favorable attitude toward WAM is in general related to the positive evaluation of female leasership. However, among male respondents female leadership will be appreciated only for feminine behaviors, whereas among female respondents hot feminine and masculine leader behaviors will show significant correlations with a favorable attitude toward women leaders (WAMS). The following set of instruments were administered to test the above hypotheses. (1) Leader behavior was measured by using a Leader Behavior Description Questionnaire (LBDQ) Form XII. For this study, LBDQ was modified to include 60 items out of the original 100-item instrument: 5 items from each of 12 subscales. Each subject was asked to rate, by using all 60 items, both male and female leaders who were presented as abstract and imagery figures. (2) An attitude toward female managers was measured by a Japanese version of Women as Managers Scale (WAMS) developed by Wakabayashi and Munekata (1985). The Japanese version of WAMS includes 19 bipollar scales with either a favorable or an unfavorable statement on each end of the scale. Based on the result of a factor analysis, 16 items were included into a single composite WAMS. (3) A self-image instrument consisted of 23 adjectives with a 7-point scale from which masculinity and femininity scales were derived. Subjects for this study consisted of 110 male and 698 female college students. Major findings of this study are summarized as follows.The first and the second hypotheses were concerned with an implicit leadership theory: no sex differences in cognitive structure of leadership. Factor structure of the LBDQ were compared among four sub-groups: (1) male leader - male respondent, (2) male leader - female respondent, (3) female leader - male respondent, and (4) female leader - female respondent. Results of the analyses revieled that factor structures were highly congruent among four sample groups, indicating that the sex of both leader and respondent did not affect factor structure of leader behavior description. This gave support to Hypothesis (1) and (2). Third and fourth hypotheses were concerned with the effects of leader and respondent sexes on the level of leader behavior evaluation. Mean LBDQ subscale scores were compared across four groups. It was found that the male leader was evaluated superior to the female leader in the masculine leader functions, while the female leader was evaluated in feminine leader functions. In general, however, male respondents tended to less recognize the female as a leader than did the female respondents. Therefore, Hypotheses (3) and (4) were supported. To test the Hypothesis (5), 12 LBDQ subscales were regressed on the WAMS score for male and female subjects separately. Results of the analysis indicated that among male subjects WAMS was found to be contributed only by a limitted, stereotyped LBDQ scales (e.g., a consideration scale), while among female respondents a variety of LBDQ subscales (both male and female-typed scales) showed significant contributions to WAMS. The result was supportive of Hypothesis (5). Finally, it was found that the self-image of masculinity among female respondents was significantly correlated with the evaluation of fema