Examining childhood risk and protective factors for youth violence

Background and Purpose: Numerous studies have examined predictors of youth violence associated with the individual child (e.g., ADHD/attention problems); the family (e.g., family conflict); school (e.g., academic failure and low bonding to school), and surrounding neighborhood (e.g., neighborhood disadvantage) (Herrenkohl et al., 2001; Loeber et al., 2008). However, few studies examine predictors comprehensively using a systematic approach to differentiate and compare risk and promotive influences (i.e., factors that increase and lessen the risk of violence in youth). Additionally, there have been relatively few longitudinal studies of risk and promotive factors that span the periods of childhood and adolescence. More common are cross-sectional studies that do not allow tests of the temporal sequencing among variables. A thorough understanding of role and function of hypothesized predictors is needed to deepen understanding of the etiology of violence, and provide a strong foundation for prevention and intervention efforts.

Methods: Data are from the Seattle Social Developmental Project (SSDP). SSDP is an on-going longitudinal panel study of the etiology of positive and antisocial behaviors that began in 1985 with 808 5th students from elementary schools in Seattle, Washington. The study includes data from multiple sources, including youth participants, parents, and teachers. Predictors in the current study cover the individual, family, school, peer, and neighborhood factors domains, all measured when youth were 10-12 years of age. Violence perpetration was measured at two time points that coincide with youth ages 13-14 and 15-18. All predictor variables were coded to examine risk and promotive effects of each variable within a single model. Bivariate and sequential multivariate regression models controlled for child gender and ethnicity.

Results: Bivariate models of risk and promotive variables revealed a slightly different set of statistically significant predictors for violence perpetration at ages 13-14 and 15-18, although some similar predictors did emerge. For example, attention problems measured at ages 10-12 had risk and promotive effect on violence at ages 15-18, but only a risk effect on violence at ages 13-14.. Availability of drugs at ages 10-12 was a risk and promotive variable for violence in early and later adolescence. Multivariate models showed that, for the violence in early adolescence, Native American ethnicity, attention problems, neighborhood kids in trouble, and availability of drugs had statistically significant risk effects. Multivariate regressions showed neighborhood factors to be among the most salient and consistent predictors of violence in earlier and later adolescence after controlling other variables in the tested models.

Conclusions and Implications: Methods advance understanding of the role and function of hypothesized predictors of violence in youth. Findings about risk and promotive effects can assist practitioners and policy makers in developing viable prevention and intervention efforts that can lessen the incidence and prevalence of violence among young people. Social work scholars and practitioners should be at the center of efforts to develop and promote empirically-supported prevention programs, as well as studies on the etiology of youth violence.