Journal of Interpersonal Violence, Ahead of Print.
Responding to high rates of interpersonal victimization and perpetration among adolescents, schools have implemented bystander intervention (BI) training to educate students to intervene to prevent or stop violence. These trainings function much like an application of scripts for guardianship in action. The current study builds on the overlapping and complementary bodies of BI and routine activities research by testing whether participation in BI training, namely Green Dot (GD), influences individuals’ underlying ability to intervene. Using four years of survey data collected from high school students (N = 2,374–3,443), we use item response theory to model the difficulty of engaging in different BI behaviors. We then estimate multivariate ordinary least squares regression models, one for each year, to estimate the effect of GD training on students’ ability to intervene. The item response theory results show that BI behaviors differ in terms of how “difficult” they are for respondents to engage in. Findings show that in each year, GD training increased students’ underlying ability to intervene. Our findings suggest BI training and guardianship in action scripts should take into account this varying difficulty of intervention behaviors to best train individuals for successful intervention to prevent victimization.