Predicting cyberbullying victimization among college students with household dysfunctional experiences: Using a decision tree model

Abstract Cyberbullying is a growing concern among college students, yet few studies have examined how household dysfunction predicts victimization in digital spaces. Guided by lifestyles and routine activities theory (LRAT), this study explored the extent to which various forms of childhood household dysfunction (substance abuse, mental illness, parental incarceration, separation, or death) and demographic factors (gender, race, and sexual orientation) are associated with cyberbullying victimization among college students. Survey data were collected from 550 undergraduates at four US universities. A decision tree model for binary classification was employed to identify the most influential predictors and their interactions. The result indicated that students from dysfunctional family environments—whether or not these involved substance use, incarceration, or parental separation—as well as those identifying with certain races or sexual orientations, are likely to be victimized in cyberspace later in life. These findings support the implementation of targeted prevention strategies through early intervention at the secondary and higher education settings.