Abstract
The article presents a meta-analysis of studies—yielding 26 independent samples of subjects—on the relationship between exposure to media violence and violent aggression. Mean effect sizes from aggregate and experimental studies do not suggest that media violence and criminal aggression are positively associated, but findings from prospective longitudinal studies are more ambiguous. Summary statistics based on models strictly conforming to each study's original design were compared to post hoc models (in which measures or model specification were not consistent with the original description). Overall, “original” models, controlling for “trait,” did not suggest that exposure to media violence is associated with criminal aggression. The summary statistic for boys reached statistical significance, but the effect size was small, and conclusions based on this finding are attenuated because of known biases in the coefficients estimated.
Get full access to this article
View all access options for this article.
