Avatar Sex Moderates Aggression in Violent Video Games, But Only for Women

Three studies tested findings reported by Yang, Huesmann, and Bushman (2014) that playing a male avatar in a violent video game leads to greater aggression than playing a female avatar in the same game. The male avatar effect was confirmed in Study 1 (N=79) for post-game aggression: compared to playing a female character, participants who had played the male fighter in a violent mixed martials arts game chose more Hot Sauce for another participant who allegedly disliked spicy food. In contrast to Yang et al. (2014), however, the male avatar effect was qualified by participant sex, indicating that the effect was more strongly pronounced and only significant for female participants. A similar interaction effect was observed in Study 2 (N=76) and Study 3 (N=70) for in-game aggression: only female participants playing a male avatar showed a greater hit ratio in a mixed martials arts game (Study 2) or a greater number of attacks in a brawler game (Study 3) than their colleagues who played a female avatar. At this stage, the reason for this cross-gender effect is unclear. Given that games allow for behavior (i.e., aggression) independent of socially shared gender norms, we may speculate that for women, male avatars may provide the opportunity to “step out” of prevailing social norms regarding non-aggressive female behavior and adopt the role of the (hyper-)aggressive male. However, this hypothesis needs to be tested in future studies.
All three studies additionally tested the mediating effect of male gender stereotype activation that was hypothesized by Yang et al. (2014). In addition to priming violent behavior, and in line with the General Aggression Model, the authors had speculated that playing the male avatar automatically activated male gender stereotypes (i.e., aggressive thoughts and behavior) which then caused aggressive behavior. In order to address this activation hypothesis, we designed an indirect cognitive measure of gender role identity using the Positive-Negative Sex-Role Inventory (PN-SRI: Berger & Krahé, 2013). After participants played the violent game, positive and negative aspects of masculinity and femininity were presented as word fragments in a five-minute response window in Study 1 and 2. Fragment completion rates served as indicators of cognitive activation of male stereotypes. In Study 3, participants used the intact PN-SRI gender attributes to rate the avatar after playing the game. However, both direct and indirect measures failed to corroborate the stereotype activation hypothesis in the present studies: word fragments related to male stereotypes were not completed more often than fragments related to female stereotypes (Study 1 and 2). Also, neither in-game aggression nor success in the game was associated with how masculine participants perceived their fighter (Study 3). At the present stage, thus, the mechanisms underlying the gender effect that participants respond differently when playing a male or female avatar in a violent video game remain unclear.