Bullying in Schools: How School and Student Characteristics Predict Bullying Behaviors Among Boys in American Secondary Schools

This dissertation argues that bullying is a fundamental response to bullies’ feelings of insecurity. Past research has found factors associated with bullying to include socioeconomic status and propensities towards violent behavior. Contextual factors posited here that produce the feelings of insecurity, which lead to bullying, include peer group dynamics, school climates, and teaching. In relationship to peer groups, the theoretical framework of this dissertation draws primarily from the theories of Robert E. Crosnoe and Dorte M. Sondergaard. The assumption is that students are socially embedded in peer groups in which they struggle for social status (Crosnoe 2011) and in many cases experience the threat of social marginalization (Sondergaard 2012). Sondergaard, in particular, theorizes that the more insecure students feel about their social status in peer groups, the more likely they are to resort to bullying behavior. All multivariate analyses in this dissertation are limited to white, black, and Latino boys. The resulting sample comprises N=6,491 student observations nested within 153 schools. The nested sampling structure requires multi-level modeling (MLM) for the calculation of unbiased estimates. I find that individual-level student background characteristics are stronger predictors of bully identification than the school context, as measured by student body composition and teaching style factors. In addition, social status insecurity is a mediating factor for many of the student- and school-level predictors of bullying. The dissertation distinguishes four types of schools, each of which is above or below average on two major dimensions. The first dimension is academic support (i.e., how caring and responsive teachers are), while the other is academic press (i.e., how strict and demanding they are). I find that black male students are more likely to self-identify as bullies in schools that are below average on both academic support and academic press, compared to those that are above average on both. The pattern for Latino boys is different. For them, self-reported bullying is higher when the school rates high on academic support, but low on academic press. I find no statistically significant role for teaching styles in predicting the amount of bully identification among white males