Intelligence as a protective factor against offending: A meta-analytic review of prospective longitudinal studies

Purpose: To synthesize results from major prospective longitudinal studies that investigated the extent to which intelligence may function as a protective factor against offending and violence. Methods: Results are based on systematic searches of the literature across 18 databases. Papers are included in the meta-analyses if results are based on longitudinal data. Results: Fifteen longitudinal studies investigate the extent to which an above-average intelligence may function as a protective factor. Meta-analytic results of studies on interactive protective factors suggest that a higher level of intelligence is a factor which can predict low levels of offending differentially within the high-risk (random effects model OR = 2.32; 95% CI: 1.49-3.63; p = 0.0001) and the low-risk (random effects model OR = 133; 95% CI: 0.88-2.01; p = 0.18) groups. A high intelligence level is differentially protective against offending within different levels of risk. In agreement with an interaction effect, the high-risk and low-risk effect sizes were significantly different (mixed effects meta-regression: point estimate = 0.509; SE = 0.175; p = 0.004). Meta-analytic synthesis of studies that looked at risk-based protective factors (i.e. analyses based only on high-risk individuals) is also presented and results are consistent with initial hypotheses. Conclusions: This methodological demonstration paper confirms the variability in conceptualizations, theoretical approaches and methodological strategies used to investigate the protective effects of intelligence against offending. Intelligence can function as a protective factor for offending. Implications for policy and practice are highlighted. (C) 2016 Elsevier Ltd. All rights reserved.