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 = 1.33; 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.
- Prospective longitudinal studies
- Protective factor