This dissertation presents an effort to assess the impact of bully victimization on a range of 10 delinquency outcomes measured over a six-year period.
The author of this dissertation presents data gathered from a research project on the nature and significance of school bullying. In the first chapter, the author discusses the following: previous research on the topic; the definition of bullying and the three types of bullying behaviors; the prevalence of bullying and victimization; how bullying rates are measured; the four participative roles of bullies, victims, bully-victims, and bystanders; and theoretical models that seek to explain why bullying occurs. In the second chapter, the author presents a research study on the effect of bully victimization on delinquency, in which the author used a large, nationally representative panel dataset, the NLSY97, and a propensity score matching technique to assess the impact of bully victimization on a range of 10 delinquency outcomes measured over a six-year period. That analytic strategy considered the effect of baseline group differences by matching bullied and non-bullied subjects on propensity scores, and allowed observable covariates to be eliminated as potential cofounders of the estimated treatment effect. Results showed that victimization before the age of 12 years was significantly predictive of the development of several delinquent behaviors, including running away from home, selling drugs, vandalism, theft and other property crimes, and assault. In the third and final chapter, the author assesses the overall effectiveness of school-based programs for preventing bullying and victimization. Results suggested that prevention programs, in general, are significantly effective at reducing victimization in schools but only marginally successful at reducing bullying. The author explored the possibility of systematic between-study heterogeneity via moderator analysis and identified several significant moderators of treatment impact on victimization.