Adolescent Social Networks and Violence in Rural Colombia

Abstract

This dissertation examines violence in adolescent social networks in the context of a rural and resource-limited community in the Caribbean region of Colombia. Utilizing mixed methods data (focus groups and surveys) from 242 school-enrolled adolescents, three empirical studies explored adolescent violence experiences in their community, school, and intimate partner violence relationships. Study one utilizes a social complexity framework and mixed methods design to address victimization. Social network data showed that adolescents' psychological and physical violence victimizations occurred across their community, household, school, and emotionally adverse relationships; and were more likely in girl-nominated relationships, a relationship that shared more alters or included an adult. Qualitative findings highlight cycles of violence that were salient across inter-generational and community settings. Cultural beliefs about violence emerged as critical to consider, particularly as participants perceived differential effects and values of victimization by gender and age. The second study builds on the previous findings by concentrating on school peer violence perpetration. Results show that psychological and physical violence perpetration status is not related to being socially connected to similar perpetration status peers. Adolescent victims of school peer violence were more likely to engage in perpetration (both psychological and physical) than non-victims controlling for non-school victimization (e.g., community, household), age, and academic standing, ethnicity, and gender. Adolescents who engaged in physical violence perpetration were more likely to live with non-parental caregivers (than those living with one or both parents) and report multiple cross-gender friendships at school. Study three centers on school peers' social network influence on physical intimate partner violence (IPV) engagement. Results show that adolescents with higher proportions of socially connected peers who perpetrate IPV are more likely to report perpetration, controlling for gender, age group, social network position, and school victimization. Conversely, adolescents with a higher proportion of IPV victim friends are less likely to report victimization. Bidirectional violence in the partnership was associated with the opposite status, such that victims were more likely to be perpetrators in the same relationship and vice versa. IPV engaged adolescents were not more likely to be socially connected to school peers of similar status, than expected by chance. Finally, the last paper describes how Community-Engaged Research (CEnR) approaches were used to develop, implement, and guide the dissemination and protocols for my study.