ECPN poster contestant: Perceptions of School Climate Among Subgroups of Aggressive and Victimized Youth: A Latent Class Analysis

Introduction: Although support has been found for differentiating subgroups of adolescents who differ in their patterns of aggression and victimization, further work is needed to determine how these subgroups differ in protective factors that influence their behavior. This information is needed to inform the development of more effective interventions that are sensitive to the needs of these distinct groups. School-based prevention programs often focus on addressing factors within the school climate. This suggests the need for further work to establish how subgroups of adolescents might differ in their perceptions of school climate. The purpose of this study was to identify subgroups of 6th grade urban middle school students that differ in their patterns of victimization and aggression, and to investigate differences in perceptions of school climate across these subgroups.

Method: This study examined data collected during an efficacy trial of an anti-bullying program. Participants were 426 6th grade students (81% African American, 53% female) from three middle schools in an urban public school system in the Southeastern United States. Four cohorts of students completed the Problem Behavior Frequency Scale and School Climate Inventory in the Fall of 6th grade.

Results: Latent class analysis of overt and relational aggression and victimization items was conducted with Mplus (Muthen & Muthen, 2015). Model fit statistics, class size considerations, and theory were used to identify the optimal number of classes. Consistent with previous research (e.g., Bettencourt & Farrell, 2013), a four-class solution was found to best fit the data. Item endorsement patterns revealed the following classes: (a) nonvictimized aggressors (11%), (b) aggressive victims (14%), (c) predominantly victimized (23%), and (d) well-adjusted youth (52%). Chi square tests will be conducted to examine demographic differences in class membership. Analyses of covariance will be conducted to examine class differences in perceptions of school climate while controlling for demographic differences (e.g., gender, race, school). We hypothesize that youth in the well-adjusted class will have the most positive perceptions and aggressive victims will have the most negative perceptions of school climate.

Conclusions: Findings provide further evidence of the heterogeneity among aggressive and victimized youth. The development of interventions that are sensitive to the differences among subgroups of aggressive and victimized youth are warranted, and may be more effective than universal interventions. Findings and implications related to class differences in school climate will be discussed.