A search of the Criminal Justice Abstracts and PsycINFO databases was conducted. We combined search terms (using “and” statements) related to potential independent variables with a list of outcome terms. In Criminal Justice Abstracts, the outcome terms were: aggression, delinquency, crime, violence, violent, property, theft, status, nonviolent and non-violent. In PsycINFO, the outcome terms were aggression, delinquency, crime, violence, violent, property, theft, status, aggression, conduct disorder, conduct problems, externalizing, behavioral problems, antisocial, nonviolent, and non-violent.
We used broad search terms, attempting to capture all educational and learning measures. The final set of search terms was: education, educational, attainment, academic, school, grades, learning disability, dyslexia, and reading. When choosing among studies, we selected for further evaluation those with measures of academic achievement; learning; school attachment (and related constructs); parent education; educational attainment; and school problems (including suspension, expulsion, truancy, and indices of school problems). We excluded measures of school involvement or specific forms of school participation such as sports participation, club participation, or “extracurricular activities.” We also excluded measures of academic expectations by child or parent. We excluded school environment variables such as “feel unsafe at school” or “unsafe school environment.” In addition, we added studies that came to our attention through other means. For example, some were added to the master list based on the search done for the chapter on intelligence. Other items turned up due to the inclusion of GPA, for example, as a control variable in studies of other topics.
The master list was 162 pages long and we culled it for “tier 1” items which we deemed had the highest likelihood of reporting associations between education factors and measures of either violent behavior/physical aggression or nonviolent- only offending (excluding drug-only offending) and tier 2 items which were more ambiguously titled. Ultimately, we coded 92 studies for our final full reference table. Thus, the final tables for this chapter are broken down into eight categories (see appendix for the list):
Measures of School Bonding
Learning and Learning Disability