Bullying and Aggression in Private Versus Public Schools: A Matched Sample Contrast

Introduction: Several studies have shown that private schools generally have lower rates of discipline problems and issues with school safety than public schools.  Yet, there has been relatively limited research examining bullying and related concerns in private vs. public schools. There are likely differences in the way that students in public vs. private schools experience and respond to bullying. The current study aimed to examine this contrast directly using propensity score matching, which allowed for a more direct comparison across school types. Specifically, the propensity scores allowed us to match students who were similarly likely to experience the “treatment” (i.e., school type), thereby reducing selection bias; this approach is novel with regard to its application to the study of bullying in public vs. private schools.   

Methods: Data come from a larger study of 26,245 students in 58 public and 5 private high schools participating in a statewide project focused on improving school climate. All students in private school were matched with similar students in public schools on the basis of 16 school and student background variables using nearest neighbor matching in R (MatchIt; Ho et al., 2011), resulting in a matched subsample of 2,154 public and 2,154 private school students.

Results: Youth in private schools were significantly less likely to report physical victimization (AOR = .856) yet almost 3 times more likely to report cyber perpetration (AOR = 2.60) as compared to youth in public schools.  Youth in private schools were less likely to report being bullied in the gym (OR = .81), walking to and from school (AOR = .67), at the bus stop or in the bus (AOR = .59), and within their neighborhood (AOR = .67). There were no significant differences in being bullied in the bathroom, after school activities, hallway/lockers, cafeteria, locker room, class, or on school grounds.  With regard to bystander responses to witnessing bullying, those in private schools were significantly less likely to encourage the victim to tell a teacher (AOR = .879) or to tell an adult (AOR = .75).

Conclusions:  The results based on the matched sample of public and private school students indicated that being in private schools appears to influence the forms of bullying experienced, the location where it occurs, and the likelihood that bystanders of bullying will tell an adult.  Additional analyses will further examine what individual characteristics and aspects of the environment are associated with these differences.