Research That Matters (January 17 - 20, 2008)

Background and Purpose: Childhood bullying behavior is considered to be a common precursor of youth violence (Orpinas & Horne, 2006). Bullying is a form of aggression in which one or more children repeatedly and purposely harm peers. Bullying behavior can be physical (e.g., hitting), or relational (e.g., social exclusion), and involves a power differential between the bully and victim (Fekkes, Pijpers, & Verloove-Vanhorick, 2005). This power imbalance and a lack of regard for others might be an important risk factor for violence perpetration in adolescence. Although some studies have indicated bullies are at risk for youth violence (Kim et al., 2006; Orpinas & Horne, 2006), little research has examined what risk and protective factors might mediate this association. The social development model (SDM; Catalano & Hawkins, 1996) provides an explanatory framework to explain initiation of and persistence in a broad range of problem behaviors in childhood and adolescence. The SDM integrates key features of control, social learning, and differential association theories to specify the role of parental, school, peer, and individual influences predictive of problem behavior. The aim of this study is to examine whether the causal linkage specified by the SDM mediates the association between childhood bullying behavior and later youth violence.

Method: This study utilizes a sample of 941students (male=503, female=438) drawn from the Raising Healthy Children Project, a 13-year longitudinal study of the etiology of adolescent problem behavior. Childhood bullying is measured at grades 4 and 5 by student self-reports on past-year frequencies of bullying behaviors (e.g., hitting and pushing, spreading rumors, and social exclusion). SDM constructs are measured by annual parental, teacher, and student data while students were in grades 6 to 8. Violence is measured, during grades 10 to 12, by student self-report of violent behaviors (e.g., hitting to hurt other people and physical fights). Structural Equation Modeling (SEM) techniques are used to examine the extent to which SDM constructs mediate childhood bullying and youth violence.

Results: Results from SEM analyses support the causal relationships hypothesized by the SDM. Childhood bullying behavior has a positive overall association with later youth violence (standardized coefficient= 0.33, p< 0.05). This relationship is partially mediated by SDM constructs (e.g., opportunities, involvement, rewards, and bonding), after controlling for gender (standardized coefficient after taking account for the SDM mediators and gender= 0.20, p< 0.05). Pro-social rewards, pro-social and antisocial bonding, and pro-social belief measured at middle school have significant direct effects on violence during high school (range of the standardized coefficients= -0.22 to 0.31, p< 0.05).

Implications: Findings indicate that bullying behavior in elementary school is predictive of youth violence. Further, findings underscore that early intervention to reduce bullying might contribute to preventing more serious forms of violence in adolescence. The SDM provides a useful tool in developing intervention targets that might interrupt the link between childhood bullying and youth violence. Prevention programs that enhance recognition for pro-social activities, increase bonding to positive peers, teachers and parents, and strengthen anti-violence norms in families and schools might be promising.