Abstract
The effect of exposure to violent video game content on aggression is intensely debated. Meta-analyses have produced widely varying estimates as to the effect (or non-effect) of violent video games on subsequent aggressive thoughts and behaviors. Recent work suggests that interactivity and player skill may play key roles in moderating the effects of violent content in video games on aggression. This study investigates the effects of violence, interactivity, and player skill on mild aggressive behavior using a custom-developed first-person shooter game allowing for high levels of experimental control. We conduct effect and equivalence tests with effect size assumptions drawn from prominent meta-analyses in the video game violence literature, finding that aggressive behavior following violent video game play is statistically equivalent to that observed following non-violent game play. We also observe an interaction between violent game content, player skill, and interactivity. When player skill matched the interactivity of the game, violent content led to an increase in aggressive behavior, whereas when player skill did not match the interactivity of the game, violent content decreased aggressive behavior. This interaction is probed using a multiverse analysis incorporating both classical significance testing and Bayesian analyses.
