Denise C. Gottfredson

Student misbehavior, which includes school violence and bullying, interfere with academic achievements and is one of the major sources of teacher turnover in American schools. This article shows how schools can be organized and managed in order to minimize and prevent student misbehavior. It identifies the kinds of practices and policies they can adopt to decrease the levels of victimization, violence, substance use, and crime both in and out of schools. It considers the research conducted on the effectiveness of curriculum-based prevention programs. It then presents evidence that shows that how the school is organized and managed also influences school behavior and problem behavior. This article also discusses discipline management and school culture and explains how these can be used to reduce crime and other related problem behaviors.

This study assesses the effects of three aspects of school organization—student enrollment, student-teacher ratio, and the number of different students taught—on the property and personal victimization experiences of students. It hypothesizes that smaller schools, schools with lower ratios of students to adults, and schools in which the number of different students taught by the typical teacher is lower will produce less victimization because of the increased social capital to which students in these schools are exposed. Using data from the National Study of Delinquency Prevention in Schools, the authors find that, net of individual-level risk factors and confounding characteristics of schools and their surrounding communities, each of the three aspects of school organization is related to student victimization but that these effects vary across victimization type. Their research suggests that reducing the ratio of students to teachers and reducing the number of different students taught by the average teacher are likely to reduce student victimization. Reducing school size is not. The authors also find evidence that higher levels of social capital, as measured by student consensus about normative beliefs, partially mediate the effects of student-teacher ratio on personal victimization.

This study tested the effectiveness of a prevention curriculum, All Stars, as implemented in a year-long school-based after school program and provides an independent replication of the effects of All Stars on targeted mediators and problem behaviors using an experimental methodology. Middle school students (N = 447) who registered for the after school program were randomly assigned to the experimental or control condition. The sample included approximately equal proportions of males and females, was 70% African American, and 59% of the students received subsidized meals at school. All Stars was delivered with reasonable integrity to the program design, although with lower quality than reported in earlier efficacy trials. However, actual student exposure to the program was lower than expected due to low levels of attendance in the after school program. Students who ever attended received an average of 16 h of All Stars instruction. Results showed no differences between the treatment and control students at post-test on any of the outcomes or mediators. Further, no positive effects were found for youths receiving higher dosage, higher quality program delivery, or both. Insufficient time to achieve high quality implementation in the after school context and potential deviancy training are suggested as reasons for the failure to replicate positive program effects.

The Strengthening Washington DC Families Project (SWFP) examined implementation fidelity and effectiveness when a selective, evidence-based prevention program was implemented with a sample of 715 predominantly African American families across multiple settings in an urban area. Using a true experimental design, this study reports on the differential effectiveness of four conditions (child skills training only, parent skills training only, parent and child skills training plus family skills training, and minimal treatment controls) in reducing child antisocial behavior and its precursors. Major challenges with recruitment and retention of participants and uneven program coverage were documented. No statistically significant positive effects for any of the program conditions were observed, and a statistically significant negative effect on child reports of Negative Peer Associations was observed for children of families assigned to the family skills training condition. Two marginally significant findings were observed: Child's positive adjustment favored families assigned to family skills training condition relative to minimal treatment and child training only, and family supervision and bonding was lower for children in family skills training than in the other three conditions. Hypotheses about potential explanations for the weaker than expected effects of this program are offered, as are thoughts about the infrastructure necessary to successfully implement family strengthening programs and the future of prevention science.