School contextual effects and adolescent deviance: A test of the “other causes” part of self-control theory

The current study tested the self-control-deviance link at both the individual and the school contextual levels, to examine whether adolescent deviance varied across classrooms, whether low self-control and other individual-level variables were associated with deviance, whether the associations varied across classrooms, and whether school contextual variables explained the variation. Anonymous self-report data were collected from N = 8348 Swiss adolescents (Mage = 17.95 years, SD = 1.42; 48.5 % females) from 585 classrooms, part of a randomly selected national probability sample. Individual-level variables included background variables, immigrant status, low self-control, and deviance. School-level variables included classroom-level SES, immigrant composition, and low self-control, class size, school educational track, and school locale. Findings from multilevel model tests indicated that individual- and classroom-level variables accounted for unique variance in deviance. Low self-control and all other individual-level variables (except for SES) predicted deviance, and the associations varied across classrooms. Classroom immigrant composition and educational track predicted deviance. Classroom low self-control and immigrant composition, in particular, explained variation in the associations between individual factors and deviance. A higher proportion of immigrant youth in classrooms was associated with a higher level of deviance among male but not female adolescents, a novel immigrant paradox. Study findings provide important evidence for school contextual effects in understanding adolescent deviance.