Effect of violence exposure on aggressive intervention in Chinese college students: A moderated mediation model

•Violence exposure positively predicted aggressive intervention.•Normative beliefs about aggression partially mediated the effect of violence exposure on aggressive intervention.•The combination of anxiety and number of bystanders moderated the mediating effect. Specifically, when the number of bystanders was small, the mediating effect enhanced as the level of anxiety increased. Meanwhile, when the number of bystanders was large, the mediating effect did not change as the level of anxiety increased. The present study was mainly based on the General Aggression Model and combined it with the Bystander Intervention Model to explore the mechanism of aggressive intervention. A total of 1018 college students (mean age 19.59 years) from five universities in China were recruited, who completed a series of questionnaires, mainly concluding violence environment exposure questionnaire, the normative beliefs about aggression scale, subscale of the aggressive intervention, trait anxiety scale and number of bystanders. The results showed that after controlling for social desirability and gender, violence exposure positively predicted aggressive intervention, and normative beliefs about aggression partially mediated between the two variables. The three-way interaction of normative beliefs about aggression, anxiety, and the number of bystanders negatively predicted aggressive intervention. Moreover, anxiety moderated the mediating effect. Specifically, when the level of anxiety was higher, the mediating effect was larger. The number of bystanders played as a secondary moderator. The moderating effect of anxiety decreased as the number of bystanders increased. The results suggest that the General Aggression Model remains a theoretical model for explaining aggressive intervention, and the combination of anxiety and the number of bystanders influences the process of GAM.