School context and genetic influences on aggression in adolescence

Genetic and environmental contributions to variation in aggression were examined using adolescents' self-reports of aggressive behavior The National Longitudinal Study of Adolescent Health provided a sample of 1,515 pairs of adolescents in five genetically informative groups (i.e., monozygotic twins, dizygotic twins, full siblings, half siblings, and unrelated siblings reared together). The analysis a DeFries-Fulker regression within a hierarchical linear model, yielded findings on individual-level heritability (h(2)), shared environmental effects (c(2)), school-level effects, and school-level moderation of h(2) and c(2). The estimate of h(2) for aggression in the full sample was .32, and c(2) was .05. In the moderating effect, h(2) increased and c(2) decreased with greater school-level family warmth. Two effects on school means were Sound: Those schools with greater ethnic-racial heterogeneity had higher mean levels elf aggression, and schools where students perceived greater family warmth had lower mean levels of aggression.