Disentangling the Link Between Victimization and Violent Behaviors: The Effects of Developmental Timing and Types of Victimization Experience

Introduction: Violent victimization remains one of the most critical public health concerns for young people in the United States (CDC, 2006). Although prior victimization experiences are an established risk factor for adolescent and adulthood violent behaviors (Maas et al., 2008), few studies provide longitudinal assessment of violence exposure across the lifespan or distinguish between differing forms of victimization and cumulative exposure. The current NIMH-funded study both addresses these gaps and provides access to the underrepresented population of youth who are at-risk for violent behaviors, yet still accessible through community-based venues. Specifically, this paper targets in-depth victimization assessment to address two central research questions: What is the relative importance of victimization across three lifespan points (childhood, adolescence, early adulthood) in predicting subsequent violent behavior? What types of victimization experiences contribute most? These questions contribute to much-needed theory-building that is essential to the effective prevention of violence (Farrell & Flannery, 2006).

Methods: This sample (N=851) was initially surveyed in high school on the basis of school drop-out risk (age M=16) and subsequently as young adults (n= 739; age M=20.6). High school drop-out risk is established as a proxy for other risks and developmental difficulties (Herting, 1990). At baseline 45% of the sample were female, 60% were racial minority, and SES was generally low to moderate. A recently completed follow-up survey assessed victimization (17 items) for multiple forms of witnessed and direct experiences (physical, emotional, sexual, property assault) spanning childhood, adolescence, and young adulthood (Finkelhor et al., 2005). The violent behaviors measure assessed injuring others, harming a partner, damaging others' property, and hitting when angry. All measures had good to excellent psychometric properties.

Results: Regression analyses were performed predicting high school and early adulthood violent behaviors. Both childhood and adolescent victimization experiences provided significant, unique prediction of adolescent and violent behaviors (R2 values ranged=0.27-0.58, all p values<.01). Cumulatively, all five victimization forms provided significant prediction of adolescent violence; witnessing was the strongest contributor, followed by physical maltreatment and childhood sexual abuse. In predicting adulthood violence, witnessing effects were replaced by earlier life exposure to property assault (Standardized betas=0.66 in childhood, 0.72 in adulthood, p<.001) followed by adult emotional abuse.

Conclusions: These results extend prior research, providing life-course insights regarding linkages between victimization and violent behaviors spanning multiple types of victimization experiences. All five domains of victimization experience provided unique as well as cumulative contribution to later violent behaviors, underscoring the importance of multi-form assessment. Findings underscore the social learning pathway between prior witnessing and later perpetrating violence, compounded by effects of direct victimization. The prominence of property assault as a unique predictor of adulthood violent behaviors argues for attention to anger and retribution-based pathways. Implications of findings for prevention and intervention development and targeted attention to community-based, emotionally vulnerable youth are discussed.