When and why group gender composition affects group members' evaluations of their group-mates: Perception, behavior, and outcome interdependence

Interdependent work, where men and women work together in groups, is becoming an increasingly common part of today’s workplace. In these interdependent settings, gender is not just an attribute of an individual (target gender), but also an attribute of the group as a whole (group gender composition). Recent research suggests that, in these interdependent contexts, it is group gender composition, rather than a target’s gender, that affects group members’ evaluations of their group-mates (West, Heilman, Gullett, Moss-Racusin, & Magee, 2012). The current research is the first to explore when and why group gender composition influences intragroup evaluations. Across three studies, I tested two hypotheses. First, I hypothesized that group gender composition would influence intragroup evaluations via one of two routes—either via the target’s behavior or via the perceiver’s biased evaluations of the target. My second hypothesis was that increasing the amount of outcome interdependence (i.e. the extent to which group members are rewarded based on the group’s performance instead of their own individual performance) experienced by a group would improve evaluations in female relative to male dominant groups. Consistent with past research, I expected that under conditions of low outcome interdependence intragroup evaluations would be more negative in female dominant than male dominant groups. However, under conditions of high outcome interdependence, I argue that a task may appear more female gender-typed (i.e. emphasize traits typically associated with women, like cooperation), and as a result, the influence of group gender composition on intragroup evaluations should dissipate. Findings suggest that group gender composition biases perceivers’ evaluations of their group-mates. In Studies 1 and 2, there was no evidence that target behavior mediated the relationship between group gender composition and intragroup evaluations; moreover, in Study 3, group gender composition influenced intragroup evaluations even when targets’ behavior was held constant. Consistent with my second hypothesis, level of outcome interdependence moderated the relationship between group gender composition and intragroup evaluations. For Studies 1 and 3, I found the expected interaction between group gender composition and level of outcome interdependence. When groups experienced low outcome interdependence, members of male dominant groups evaluated each other more positively than members of female dominant groups. Results reversed under conditions of high outcome interdependence, such that members of female dominant groups evaluated each other more positively than members of male dominant groups, albeit not significantly so. However, there was no evidence that moderation by outcome interdependence was due to changes in the perceived gender type of the task. Findings from Study 3 suggest that participants who experienced conditions of high outcome interdependence did not believe that the task was more female gender-typed than participants who experienced conditions of low outcome interdependence. Moreover, other methods for making a task appear more female gender-typed (using female gender-typed materials and framing a task as requiring female gender-typed skills) did not moderate the relationship between group gender composition and intragroup evaluations. Results from these studies are the first to provide insight into when and why gender composition affects intragroup evaluations in interdependent task groups. The current research suggests that it is possible to improve intragroup evaluations for female dominant groups, relative to male dominant groups, and reduce bias based on group gender composition by rewarding group members based on group rather than individual performance. Additionally, the current research suggests making a task appear more female gender-typed (e.g. using traditionally female materials) may not be effective at reducing gender bias in group contexts