Risk and Protective Factors for Bullying, Cyberbullying, and Sexual Harassment in U.S. Schools: A Socio-Ecological Analysis Beyond Individual Level Factors

This study identifies the socio-ecological predictors at each level of the ecosystem that serve as risk or protective factors for the frequency of bullying, cyberbullying, and sexual harassment in U.S. schools, and examines how these predictors vary across the three behaviors, highlighting both common and distinct predictors. Using data from two waves (2018 and 2020) of the School Survey on Crime and Safety (SSOCS), a nationally representative cross-sectional survey of 5132 U.S. public elementary and secondary schools, we conduct a Generalized Ordinal Logistic Regression in hierarchical blocks of mesosystem-, exosystem-, and macrosystem-level factors to examine the relative contribution of each ecological level. We find that both mesosystem (teacher training, school size, school grade) and exosystem (neighborhood crime, parental involvement, and involvement of mental health & social services)-level factors play significant roles in predicting the frequency of bullying, sexual harassment, and cyberbullying. Urbanicity at the macrosystem-level is not a significant predictor of bullying and sexual harassment; however, schools in urban areas are less likely to have frequent and monthly cyberbullying and more likely to have occasional and no cyberbullying compared to schools in non-urban areas. We argue for the importance of multi-level interventions to address peer harassment in schools through evidence-based policies.