Mexicano de Relaciones Grupales y Organizacionales
Mexican Institute of Group and Organizational Relations
SUBORDINATES' REACTIONS TO FEMALE LEADERS
Virginia Commonwealth University
Paper presented at the annual meeting of the Eastern
Psychological Association, Baltimore, MD. April 12 - 15, 1984
Although recent reviews of the effectiveness of male leaders versus female leaders reveal little evidence of male superiority (e.g., Bass, 1981; Brown, 1979), evaluative and perceptual biases among group members persist. For example, surveys (e.g., Ferber, Huber, Spitze, 1979) indicate that both men and women overwhelmingly prefer men as their bosses, that members of small groups generally select male rather than female leaders (e.g., Eskilson & Wiley, 1976), and that females receive lower evaluationsand fewer promotions than males even when actual performance data are held constant (Rosen & Jerdee, 1973). Despite recent changes in the public's overall attitude toward women as leaders (Bass, 1981), group members continue to be biased against female leaders.
Recent analyses of categorization processes underlying social cognition suggest that these biases against female leaders stem from two sources: beliefs about what makes for good leadership and stereotypic beliefs about men and women. As is consistent with an attributional approach to leadership (Calder, 1976), many group members' personal "leadership theories" presume that good leaders are "active," "competent," "outspoken," "self-confident," and "ambitious." In other words, and in contrast to research which has shown over and over again that successful leadership depends upon both task and socioemotional abilities, many group members overemphasize the task side of leadership. Although maintaining smooth internal relationships is viewed as a positive aspect of leadership, many group members consider it to be a goal of only minor importance (Forsyth, Schlenker, Leary, & McGown, in press).
In addition to assuming leaders should be active and task-oriented, group members also characteristically attribute the qualities need for leadership to men rather than women (Bass, 1981). Furthermore, the attributes stereotypically attributed to women-- emotionality, consideration, love of children, warmth, nonaggressiveness--work to further disqualify women as suitable for leadership. Applied to leadership, these stereotypical beliefs suggest that women not only possess the qualities that will make them "bad" leaders, but they also lack those qualities needed for "good" leadership.
This attributional explanation for biases against female leaders was tested in the current research by arranging for men and women--who adopted either a stereotyped (conservative attitude) or nonstereotyped (liberal attitude) view of women--to work on tasks under the guidance of a female leader who adopted either a task or relationship leadership style. Overall, we predicted that conservative subordinates would be less satisfied with a female leader than liberal subordinates, and that this dissatisfaction would be particularly pronounced with the female leader violated sex-role expectations by behaving in a task-oriented manner.
Subjects. A total of 85 (41 males and 44 females) subjects recruited from introductory psychology classes participated in mixed sex groups led by one of three female confederates. All sessions were conducted by one of three male experimenters. All groups--except in several cases when one of the subjects failed to keep his or her appointment--included two individuals who were conservative in the attitudes toward the role of women in contemporary society, and two more liberal-minded individuals (as measured by the Attitudes Towards Women Scale).
Procedures. At the start of the session the leader was selected from among the group members on the basis of scores on a "leadership test". In actuality, the appointed leader of the group was a confederate who had been trained to behave in either a task or interpersonal manner. In the task condition the members were told that the female had been selected to be leader because her scores on the leadership quiz supposedly indicated that she would succeed in helping the group reach its task goals, and during the actual group experience she emphasized the importance of succeeding at the task and outperforming other comparable groups. In the interpersonal condition the group members were told that the leader was selected because she would be able to ensure smooth intermember relations, and during the task itself the leader repeatedly emphasized the socioemotional needs of the group members. (The three female confederates ran a balanced number of groups across the two experimental conditions).
After the group worked on several group and individual exercises, the leader was apparently dismissed and the remaining group members completed questionnaire measures of the leader's abilities, their perceptions of the leader, and checklist adapted from the SYMLOG method of group observation (Bales, 1980).
Manipulation check. Subjects' responses to the item "The person who was chosen to be the group leader was most concerned with" could range from "creating positive interpersonal relations" (9) to "accomplishing the group task" (1). A 2 X 2 X 2 (sex X attitudes toward women X style of leader) ANOVA revealed only a significant main effect of leader on this item; F (1,84) = 6.27, p <.05. The means for the task- and relationship-oriented confederates were 4.61 and 3.16, respectively. Perceived leadership effectiveness. A multivariate analysis of three 9-point items assessing satisfaction with the leader, willingness to accept the leader in other group settings, and perceived leadership effectiveness revealed a significant main effect of sex and an interaction of leader and attitudes toward women; Pillai's Trace F approximation (3,74)="2.93" and 2.60, p < .05. The sex main effect reached univariate significance on only the satisfaction item; males were less satisfied than females. The means, with higher scores indicating greater satisfaction, were 7.12 and 7.77, respectively. The means for the two-way interactions of leader's style and attitudes toward women on these three items are shown in Table 1. All three items reveal a similar pattern. For conservative subjects, the relationship-oriented leader was perceived to be less effective than the task-oriented leader. In contrast, the liberal subjects felt that the task-oriented leader was less effective than the relationship-oriented leader.
Table 1. Effects of attitudes toward women and leadership style on ratings of the leader's effectiveness. How satisfied are you with the performance of the group's leader? Leadership Style Attitude Toward Women Task Interpersonal Conservative 7.70 6.61 Liberal 7.30 7.95 Would you be willing to accept this person as your leader in other group settings? Leadership Style Attitude Toward Women Task Interpersonal Conservative 6.96 6.68 Liberal 6.55 7.26 In your opinion, was your group leader effective or ineffective? Leadership Style Attitude Toward Women Task Interpersonal Conservative 7.17 6.82 Liberal 7.05 7.42 Note. Higher scores indicate more positive ratings. Liking for the leader. Multivariate analysis of three 9-point items assessing attraction to the leader revealed only a marginally signficant main effect of attitudes toward women; Pillai's Trace F approximation (3,75)="2.22" p < .10
Relative to liberal subjects, conservative individuals liked the leader less; F (1,77)="4.99," p < .03. The respective means were 6.90 and 6.28. In addition, the liberal subjects, moreso than the conservative subjects, felt that the leader would be "easier to get along with"; F (1,77) = "4.34" p < .05. The respective means were 7.97 and 7.46. SYMLOG ratings. A complex pattern of results was obtained on the SYMLOG dimensions. On the instrumental dimension, only the main effects of sex and leader reached significance; Fs (1,75)="4.47" and 6.45 ps < .05. Females rated the confederate as higher in instrumental control relative to the males; the means were 2.36 and 1.12, respectively. In addition, the task- oriented leader was viewed as more instrumental than the relationship-oriented leader; the means were 2.50 and 0.98. The interaction of leadership style and attitudes toward women on the friendliness dimension, F (1,75)="4.12," p < .05, is shown in Table 2. Although the the two types of leaders were rated equivalently by liberal group members, the conservative subjects felt that the interpersonal leader was far more friendly than the task leader.
Table 2. Effects of attitudes toward women and leadership style on ratings of friendliness. Leadership Style Attitude Toward Women Task Interpersonal Conservative 6.21 8.95 Liberal 8.00 7.95 Note. Higher scores indicate more friendly than unfriendly. Lastly, the three way interaction of sex, attitude, and leadership style on the dominance dimension of SYMLOG (see Table 3), F (1,75)="4.14," p < .05, indicates that conservative males felt that the female leader who used an interpersonal leadership style was judged to be more dominant than a task-oriented leader.
Table 3. Effects of attitudes toward women, gender, and leadership style on ratings of dominance. Gender Conservative Attitude Liberal Attitude Task Interpersonal Task Interpersonal Males 0.00 1.90 1.13 0.63 Females 2.20 1.82 2.50 3.27 Note: Higher (and more positive) scores indicate more dominant than submissive.
An analysis of biases against female leaders based on subordinates intuitive prescriptive "theories" about leadership and their stereotyped expectations concerning men and women was largely supported. In general, more conservative group members rejected the task-oriented female leader, while more liberal group members preferred a leader with a strong focus on the task.
Bales, R. F. (1980). SYMLOG case study kit. New York: Free Press.
Bass, B. M. (1981). Stogdill's handbook of leadership. New York: Free Press.
Brown, S. M. (1979). Male versus female leaders: A comparison of empirical studies. Sex Roles, 5, 595-611.
Calder, B. J. (1977). An attribution theory of leadership. In B. M. Staw & G. R. Salancik (Eds.), New directions in organizational behavior. Chicago: Clarion Press.
Eskilson, A., & Wiley, M. G. (1976). Sex composition and leadership in small groups. Sociometry, 39, 183-194.
Ferber, M., Huber, J., & Spitze, G. (1979). Preference for men as bosses and professionals. Social Forces, 58, 466-476.
Forsyth, D. R., Schlenker, B. R., Leary, M. R., & McGown, N. (In press). Self-presentational determinants of sex differences in leadership. Small Group Behavior.
Rosen, B., & Jerdee, T. H. (1973). The influence of sex role stereotypes on evaluations of male and female supervisory behavior. Journal of Applied Psychology, 57, 44-48.