女性管理職に対する態度(WAMS)の研究 : 日本語版WAMS開発の試み A STUDY ON ATTITUDES TOWARD WOMEN AS MANAGERS : An attempt to develop a japanese version women as managers scale (wams)
Recently, social, economic and legislative changes have provided increasing opportunities for women to seek business careers and to assume administrative and management positions. The integration of women into managerial positions, however, does not occur without conflicts and problems. Peters, Terborg and Taynor (1974) pointed out that the existence of sex role stereotypes is one of the barriers which inhibits the successful integration of women into male sex-typed jobs (e.g., management). Then, they developed the Women as Managers Scale (WAMS) in order to identify and measure stereotypic attitudes toward women as managers. The present study reports on development of 4 Japanese version of WAMS. For this purpose, 55 items of the American WAMS were translated into Japanese and examined by the item selection analysis and the factor analysis. Each of the 55 items was written with two stems. One stem expressed a favorable attitude toward women as managers, while the other stem expressed its logical reversal. Then, parallel forms of the questionnaire (Form A and Form B) which differed only in terms of an item direction were constructed. The items consisted of statements with seven response alternatives from "strongly disagree" to "strongly agree". Each response alternative was given a score from one to seven, where "1" represented a traditional, conservative attitude toward women as managers, and "7" a liberal attitude. In addition to WAMS, the questionnaire contained 4 other instruments stated as follows. (1) An attitude toward the Equal Employment Opportunity Law was asked with using a single-item question regarding the subject's opinion about this law by choosing one from four alternatives. This instrument was used as one of the criterion variables for selecting relevant WAMS items. (2) The social-role attitude includes 31 statements regarding roles of women in the society. Subjects were asked to rate each one by using a 7-point scale. This instrument was reduced to three composite scales by combining relevant items: equality in social participation, traditionalism on domestic roles, and equality in the division of household labor. (3) The image of successful women was measured based on the 6 adjective pairs that were presented with a semantic differential format with a 5-point scale. All items were integrated into a single composite scale, according to the result of a factor analysis. 4) The last instrument was presented in different forms in accordance with the sex of subjects. For the female, two items were used to ask her career choice: one regarding a subject's life plan, and another on subject's determination to continue occupational career. For the male, a single item asking a subject's preference for the spouse's (or future spouse's) career choice was employed. Subjects for the study consisted of 2577 students including the following subgroups: male college students (n=568), female college students (n=817), female junior college students (n=997) and female nurse professional school students (n=195). The data processing required as a first step the item selection analysis that was designed to determine which of the original 55 items could differentiate among persons holding favorable and unfavorable attitudes toward the Equal Employment Opportunity Law. The data was analyzed independently for each of the following four groups, i.e., females responding to Form A and Form B and males responding to Form A and Form B. The item selection analysis resulted in the retention of 32 items. These 32 items were then placed in a principle factor analysis for each Form separately. Then, 13 out of 32 items were deleted from the final form of the WAMS, because they either had factor loadings of less than .40 or had equally high loadings on more than one factor in either one of the two Forms. The factor analyses conducted independently produced two identical factors from both Forms. These factors were named (I) equality versus discrimination in business, and (II) women's aptitude as managers. Items contributing to each factor were combined into composite scales. Reliability estimates calculated based on the Cronbach's alpha poduced coefficients with satisfactorily high levels. Then, in order to examine characteristics of scales constructed, they were subject to the correlational analysis with other variables. Major findings of the correlational analysis are summarized as follows. (1) The equality in business factor showed high positive correlations with women's aptitude as managers in either Form A or Form B for both sexes. This may suggest that the WAMS can converge into a single dimension. (2) The image of successful women highly correlated with both WAMS subscales. This indicates that the person holding a positive image toward successful women tends to hold a liberal attitude toward women as managers. (3) In all four conditions, the WAM subscales positively correlated significantly with two subscales of the socialrole attitude: equality in social participation and equality in the division of household labor. (4) An attitude toward the Equal Employment Opportunity Law showed significant positive correlations with two subscales of WAM. Those who strongly supported this law tended to evaluate aptitude of women as managers highly. (5) For males, WAM sub scales correlated with a subject's preference for spouse's career. For females, WAM subscales correlated with a subject's life plan and her determination to continue occupational career. Mean scores for the WAM subscales were compared across subgroups by the ANOVA method. The analysis was carried our for each Form separately. Major results of the analysis are summarized as follows. (1) Among 4 subgroups (male college students, female college students, female junior college students, and female nurse professional school students), it was found that male students have more traditional and conservative attitudes than the other 3 female groups do toward WAM and also toward social-roles of women. (2) A comparison of means among 3 female subgroups based on the college major (literature, home economics and management) failed in identifying clear, significant group differences. However, home economic students tended to be more traditional and conservative than did the literature students as a whole. (3) A comparison of means among 3 female groups based on the junior college major (literature, child education, human relations) revieled that a significant group difference exists in the social-role attitude. On the other hand, no group difference was found on attitudes toward women as managers. In conclusion, compared to the original American WAMS, the Japanese version WAMS was found to have different factor-analytic dimensions, and thus different items to be used for the scale construction. Although the two Japanese WAM subscales maintained high internal consistency and meaningful relationships with other related variables, they had one problem. The mean scores were significantly different between two forms of sentence stems: between Form A and Form B. Therefore, instead of choosing one statement (stem) from either Form A of B, both statements were used as either a favorable or an unfavorable end of a bipolar scale for the final form of the Japanese WAMS.