Acting like a Tough GuyViolent-Sexist Video Games, Identification with Game Characters, Masculine Beliefs, & Empathy for Female Violence Victims

A lab in Italy. One hundred and fifty-four teenagers sit down at computers and play video games for twenty-five minutes. This seems ordinary enough. But which game they play — determined randomly — will shape how much empathy they feel for a girl who has been beaten. That's the finding. It's not a correlational hunch or a survey of self-reported habits. It's a controlled experiment. The question Gabbiadini, Riva, Andrighetto, Volpato, and Bushman were asking wasn't whether violent games make people violent. It was something quieter and more specific: who changes, how much, and why. The paper begins with a definition that serves as a stake in the ground. Empathy is described as an emotional response that corresponds to the feelings of another person — feeling distress when you see someone in distress — and the authors call it one of the best predictors of prosocial behavior. Meta-analytic research has already shown that violent video games reduce empathic feelings and can numb players to the pain of others. That's the baseline. But this study didn’t stop there. It aimed for a more precise question: do violent-sexist games specifically erode empathy for female victims of violence? And if they do, who is most affected, and through what psychological process? The content of games like Grand Theft Auto makes that question urgent. In those games, female characters are frequently depicted as prostitutes or lap-dancers — peripheral to the narrative, existing mainly to entertain the male protagonist. The paper provides a sharp example: after paying a prostitute for sex, a player in Grand Theft Auto can kill her and get their money back. The game doesn't punish this behavior. It may even reward it with extra health points. Women are not simply background scenery here; they are part of what players actively interact with. To explain how that content might change a player, the researchers built a causal chain with two key links. The first link is identification with the game character — a psychological merging in which players imagine themselves to be that character, adopting the character's perspective and potentially its values. The second link is masculine beliefs: the cluster of norms that expects men to be aggressive, dominant, competitive, and emotionally restricted. Fear and empathy are discouraged, while anger and dominance are rewarded. Previous work has linked these norms to a greater acceptance of male dominance and aggression toward women. If you combine these two links, you get the model the study was designed to test. Exposure to a violent-sexist game increases masculine beliefs — but only for male players who strongly identify with the hyper-masculine character. Those heightened masculine beliefs, in turn, reduce empathy for female victims of violence. The statistical framework for testing this was a moderated mediation model — a method that examines whether the effect runs through a specific pathway and whether it depends on who the player is. The researchers used the PROCESS macro with one thousand bootstrap samples to test this across all three experimental conditions. The design was clean. One hundred and fifty-four Italian high school students, ages fifteen to twenty, were randomly assigned to play one of three types of game for twenty-five minutes: a violent-sexist game — Grand Theft Auto: San Andreas or Grand Theft Auto: Vice City; a violent-only game — Half Life One or Half Life Two; or a non-violent game — Dream Pinball 3D or Q.U.B.E. 2. Each participant watched about one minute of the game's intro video, practiced controls for five minutes, then played a preassigned mission alone. The violent-sexist condition included the hyper-masculine protagonist and the sexualized women. The violent-only condition had comparable action but no sexualized content. The non-violent condition served as the neutral baseline. Empathy was measured afterward by showing participants a photo of an adolescent girl who had been physically beaten by a boy and asking them to rate how sympathetic, compassionate, moved, and warm they felt toward her on a one-to-seven scale. That scale had a Cronbach's alpha — a measure of internal consistency — of point eighty-three, which is solid. Manipulation checks confirmed that the conditions worked as intended: players in the Grand Theft Auto condition rated female characters as far more sexualized than players in the Half Life condition, with an effect size of 2.72, which is large by any standard. Perceived violence also differed substantially across groups. Game involvement and excitement, meanwhile, did not differ — ruling out the possibility that one game felt more engaging overall. Now to the findings, which are specific enough to be worth sitting with. The headline result is this: masculine beliefs were negatively and reliably related to empathy for female victims — a regression coefficient of negative point forty-seven, with a p-value below 0.001. More masculine beliefs meant less empathy. That relationship held. The critical question was what drove masculine beliefs up in the first place. A three-way interaction — game type, participant gender, and identification with the character — on masculine beliefs was significant, with a coefficient of point twenty-seven and a p-value of point zero one one. When the researchers unpacked that interaction for male participants specifically, a clear pattern emerged. For boys who played the violent-sexist game, identification with the character predicted higher masculine beliefs — a slope of point thirty-two, with a p-value of point zero zero nine. For boys in the violent-only condition, there was no such relationship. For boys in the non-violent condition, there was no relationship. For girls in any condition, there was no relationship. What this means is that identification was the hinge. Boys who psychologically merged with the Grand Theft Auto character walked away endorsing masculine norms more strongly. Boys who played an equally violent game without the sexist content did not. The violence alone was not enough. It was the combination — a hyper-masculine, dominant male character in a game that depicts women as objects — that impacted masculine beliefs. The mediation tests then confirmed that this increase in masculine beliefs carried over to empathy. The indirect effect of violent-sexist gameplay on empathy, running through masculine beliefs, was significant for males: an indirect effect of approximately negative point eighteen, with a ninety-five percent confidence interval spanning negative point forty to negative point zero four. For females in the same condition, it was not significant. For males or females in the violent-only condition, it was not significant. The effect was specific to sexist content and specific to male players. One more precision worth noting: the conditional indirect effect became significant only above an identification score of approximately four point three on the identification scale. Below that threshold, even male players of the violent-sexist game didn’t show the pathway. The effect required both the content and a sufficiently strong psychological bond with the character. Who is most affected? Boys who see themselves in the guy on screen. Now, what should we make of this? The authors are careful about what their experiment can and cannot support. This involved just twenty-five minutes of gameplay, one session, in a lab. The long-term consequences are unknown. Gabbiadini and colleagues note that empathy was measured via self-report, that only two moderators and one mediator were tested, and that Grand Theft Auto's cultural reputation might have primed masculine schemas independent of the gameplay itself. They also didn’t test whether reduced empathy translates into actual aggression toward women. But what the study does establish is a plausible mechanism, documented with precision. If masculine beliefs are the operative lever — the thing that connects sexist game content to reduced empathy for female victims — then that has implications beyond screen-time arguments. Interventions might be better aimed at those norms themselves rather than simply at exposure. The paper's framing is worth returning to at the close: empathy has been described as the social glue that holds society together. This experiment shows that for particular players, under particular conditions, certain content can loosen that bond — not through violence alone, but through violence wrapped in a very specific set of messages about what men are supposed to be. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.