Methods. Secondary data analyses were conducted using data from Neighborhood Matters (Henry et al., 2014), which included 233 adolescents (Mage = 15.48, 50% male, 55% African American, 45% Hispanic) who had dated in the past year and were from 30 urban neighborhoods, characterized by concentrated disadvantage. Using latent class analysis, subgroups were identified based on peer involvement in dating violence and aggression, friend involvement in a gang, fighting within friendships, peer aggression and victimization, and delinquency. Associations between the membership in these classes, exposure to violence, and dating violence were assessed.
Results. Four classes emerged: 1) a highly violent peer context (n = 51, 22%), 2) a physically violent peer context with high levels of all physical forms of violence (n = 39, 15%), 3) a high self-reported aggressive behavior and victimization context, with low levels of peer involvement in aggressive behaviors (n = 72, 34%), and 4) a low violence peer context with low or minimal levels of violence (n = 71, 30%). Exposure to community violence predicted membership in the highly violent peer context in comparison to the other three classes. Finally, the three classes with average to high levels of violence showed significantly higher rates of dating violence victimization and perpetration in comparison to the low violence peer context class.
Conclusions. Adolescents who experienced high levels of community violence were at the greatest risk for a highly violent peer context and were at the greatest risk for dating violence victimization and perpetration. The heterogeneity of classes suggests that we need to consider these peer context variables together to better incorporate aspects of peer factors within dating violence prevention programs.