Self-Persuasion: An Experimental Evaluation of a Sexual Aggression Preventive Intervention for U.S. College Men

Grounded in the self-persuasion paradigm (an indirect persuasion approach, which places people in situations that motivate them to change their behavior), this study evaluated a brief, online intervention to reduce sexual aggression perpetration and increase prosocial bystander behaviors among heterosexual male college students (N = 241) in the United States. Students were randomly assigned to three conditions: (a) a self-persuasion intervention, (b) a social norms control condition, and (c) a control condition focusing on sense of belongingness. The self-persuasion intervention integrated three social psychological theoretical perspectives on attitudinal and behavioral change—cognitive dissonance (e.g., creating a personalized video message for incoming male college freshmen to explain the importance of consent in sexual contact), self-affirmation (e.g., reflecting on one’s core values and how they are congruent with sexual consent), and personal relevance (e.g., writing about personally relevant reasons to always seek consent when having sexual contact). Participants in the self-persuasion condition reported greater prosocial bystander behaviors (e.g., intervening in situations to prevent sexual aggression) 6 months after the intervention as compared with those in the other two conditions; however, there were no significant difference in the rate of self-reported sexual aggression perpetration across conditions. The positive effect of the self-persuasion intervention on prosocial bystander behaviors was mediated by reduced self-perceived likelihood to commit sexual aggression and moderated by in-group solidarity with other college students. That is, the intervention had the most positive effect on prosocial bystander behaviors among participants with a lower sense of in-group solidarity. These findings are discussed in light of the promise of self-persuasion for future sexual aggression prevention work.