Influence of Campus- and Individual-Level Factors on Risk of Violence Perpetration

Background and Purpose: There is a rich body of literature exploring sexual victimization prevalence and risk factors among college students and at the campus-level. Research on the rates and risk factors for sexual misconduct perpetration is more limited, as is research on the unique role of campus-level factors in shaping the risk for perpetration. This paper examined risk for perpetration using individual-level and campus-level factors to understand how variations in campus culture and environment might shape individual-level risk of perpetration.

Methods: We conducted a secondary data analysis of data collected from students at 12 universities that participated in the Multi-College Bystander Efficacy Evaluation. The data set included information from 23,389 students who completed program evaluation surveys from 2016 through 2019. At the individual level, we included variables describing individual identity (e.g., gender, sexual identity) and behaviors (e.g., association with risky peers, binge drinking). We also included campus-level variables describing campus characteristics (e.g., school size, region) and culture (e.g., perceived institutional intolerance for sexual misconduct, bystander self-efficacy). The primary outcome was any perpetration of sexual harassment, dating violence, or sexual assault. Since students in the dataset were nested within individual campuses, we analyzed data using multilevel logistic regression. In building our multilevel model, variables were removed in a step-wise fashion that was guided by theory, our data, and model performance. We used receiver operating characteristic (ROC) curves to select a final model that performed the best in terms of fit.

Results: We found several individual- and institution-level variables to be associated with having perpetrated violence during the current academic year. At the individual-level, age was associated with decreased odds of perpetration, while being a cisgender man, being in a relationship, associating with risky peers, and having previous drinking problems, were associated with higher rates of perpetration. There was also significant interaction effects between gender identity and sexual orientation, such that being a heterosexual, cisgender man was associated with an additional increase in likelihood of perpetrating violence on top of the increased odds of perpetration reported by cisgender men of any sexual orientation.

At the campus level, increased ethnic diversity, higher perceptions of institutional intolerance for sexual misconduct, and being on a campus where students reported more awareness of or exposure to sexual violence programming was associated with a reduced risk of perpetration. A cross-level interaction term suggested that the increased risk of perpetration connected to associating with risky peers was slightly reduced on campuses where students had more awareness of or exposure to sexual violence programming.

Conclusions and Implications: This study is one of the first analyses of campus variation in violence perpetration. These findings are important because they contribute to the nascent literature on campus violence perpetration while highlighting how campus-level factors contribute to one’s risk of violence perpetration. We recommend that campuses continue to develop prevention interventions that go beyond individual behavior change, and, instead target campus-level structures, attitudes, and norms that may encourage violence perpetration.