study had no significant effect. The significant effect that type of study produced when all the effect sizes were analyzed thus reflects primarily the trends observed for the interpersonal and task styles.
Table 5 also reports categorical models that were based on classifying the effect sizes into the four types of style and were computed within each type of study (i.e., organizational, assessment, laboratory). The significant between-styles effect for the organizational studies primarily reflects the relatively large mean for the democratic versus autocratic style, and the significant effect for the assessment studies primarily reflects the relatively large means for the interpersonal and the democratic versus autocratic styles. The non significance of the between-styles effect for the laboratory studies suggest that leadership styles were stereotypic in laboratory studies regardless of the type of style assessed. Although the number of laboratory studies on leadership style is unfortunately quite small, the relative consistency of this stereotypic trend across the types of style lends confidence to our generalization that leaders' behavior is somewhat gender stereotypic in experimental settings.