Moderating Effects of School Climate On Outcomes for a Universal Violence Prevention Program

Introduction: Whereas a handful of studies have addressed the way that ecological characteristics (e.g., school, community factors) affect implementation of school-based prevention efforts (Farrell, Henry, & Bettencourt, 2011; Ozer, 2006), few studies have analyzed how ecological variables influence student intervention outcomes. This study examines the extent to which the impacts of a universal school-based approach to reduce violence were moderated by perceptions of school climate. The current study uses data collected from the Multisite Violence Prevention Project (MVPP) which was one of the largest and most comprehensive violence prevention programs ever implemented (Henry, Farrell, & MVPP, 2004). It involved a 2 x 2 research design wherein 37 schools were randomly assigned to four conditions: a universal intervention (i.e., a student social-cognitive curriculum and teacher training), a selective intervention for high-risk students, both the universal and selective interventions (combined condition), and a intervention-as-usual control condition (MVPP, 2009).

Method: School climate at baseline was assessed by teacher ratings on: the quality of relationships among school members (e.g., relationships between teachers and students), and the seriousness of school problems (e.g., problems with gangs). Intervention outcomes included student-rated norms for aggression and nonviolence, beliefs supporting aggression and nonviolent responses, overt victimization, and student-rated and teacher-rated aggression. A total of 396 teachers and 5,256 students completed these ratings with an average of 8 teachers and 80 students per school.

Results:Mixed effects regressions of data suggested limited positive program impacts for students in schools participating in the universal or combined intervention versus students in schools participating in a selective or control condition. There were, however, some differences in findings for students participating in the universal and combined intervention in schools with conducive climate versus those students in schools with distressed climate. For example, in schools with Conducive Climate, students in schools assigned to the universal intervention had greater endorsement of norms for nonviolence from pre-to-post than students in schools assigned to control.  Among schools with Distressed Climate, students in schools assigned to the universal intervention had greater increases in overt victimization and teacher reported aggression from pre-to-post compared to students in schools assigned to control.

Conclusion: These findings have important implications for identifying school characteristics that researchers and practitioners should consider in planning universal school-based interventions.