Neighbor Support, Parent Support, and Aggression Among Elementary Students: Are There Differences Across Grade-Levels?

Background and Purpose: Aggressive behaviors among elementary school children are a major public health concern, particularly because early aggression is one of the strongest predictors of criminality in adolescence and adulthood. To date, most research on childhood aggression has focused on parenting practices, peer behavior, and individual-level factors (e.g., genetics). Existing studies of neighborhood effects often focus on adolescents and negative neighborhood characteristics. What is not clear is if positive neighborhood effects begin to be important predictors of aggression before adolescence, and by what mechanisms they exert their effects on behavior during the pre-adolescent developmental stage. The current study was designed to address this gap in the literature by examining three questions: (1) Does neighborhood support have a direct effect on childhood aggression?  (2) Does neighborhood support have an indirect effect on childhood aggression through parent support?  (3) Do the effects of neighbor support change (e.g., grow larger) during the later years of middle childhood?

Methods: Data were collected about 1251 children from children themselves, parents, and teachers who completed the Elementary School Success Profile (ESSP). Evidence of the validity of ESSP data has been published. The sample for the current study was restricted to cases with data from all three sources (N=627).  About half of the sample was male (48%), and 28%, 48%, and 24% were in third, fourth, and fifth grade, respectively.

Measures. A parent-report measure of supportive neighbors was included as a predictor of child aggression in a model which included parent support to children as a mediator and controls for friend support, gender, poverty, and race/ethnicity. The analysis was structural equation modeling (SEM) using the WLSMV estimator in Mplus 6.11. We first identified the best-fitting structural model for the full sample. We then constrained structural parameters to be equal across grade levels and determined if fit deteriorated significantly. We also corrected for the nesting of students in 13 schools. Missing data were handled with FIML.  

Results: The final model fit had excellent fit: chi-square = 821.042, df = 685, p = .0003; CFI= .984, TLI= .984, and RMSEA=.031. Differences in neighborhood effects were not found across grades 3, 4, and 5, but the presence of supportive neighbors was found to have an impact on childhood aggression. Neighbor support had both a direct effect on child aggression, and an indirect effect through its relationship to parent support. Results also showed, compared to Caucasians, African Americans had significantly more reported aggressive behaviors, but only in third grade; being African American (.392) was the strongest predictor of aggression, followed by low parent support (-.273).

Conclusions and Implications: Although this study used cross-sectional data and causal inferences cannot be drawn, results suggest that neighborhoods affect child development during the years before children typically become more active and independent outside the home. Strategies for preventing aggressive behaviors among young children should consider ways to enhance neighborhood support. Further, the results suggest that African Americans in third grade may be in particular need of intervention.