Deviation Scores (PIQ > VIQ)
Although reviewers have commonly written that a difference in IQ scores favoring performance IQ over verbal IQ is associated with violence, and although the idea has been around for a long time, there are, in actuality, few studies that have tested this hypothesis. We uncovered only two studies with violent crime as an outcome, no studies with nonviolent-only crime, and five studies where violent and nonviolent offenders were compared. In those studies, the findings were mixed and only barely leaning in the direction of the predicted association. In light of our findings on PIQ and violence discussed above and the mixed evidence here, we cannot conclude that the evidence supports the hypothesis that a discrepancy between PIQ and VIQ scores is an important predictor of violence.
In Tables 4.3 and 4.4, we display the tallies from studies of executive functioning. As we can see, there are few studies in most categories, but there is definitely a pattern across categories in favor of the hypothesis that violent offenders have executive deficits. This is not evident in studies of memory, and only marginal in studies of attention. In studies using measures of planning, cognitive impulsivity, other tests of cognitive abilities or EFs (such as set switching), a strong majority indicate an inverse association with violent behavior. There is little support to be found that nonviolent-only offenders have executive deficits.
Studies comparing violent to nonviolent offenders frequently report that violent offenders have lower executive abilities. Many studies report nonsignificant findings, but most of those have very low sample sizes and low statistical power. The pattern is fairly clear, however: 7 out of 10 studies of memory, 7 out of 10 studies of problem solving ability, 5 out of 6 studies of planning, 3 out of 4 studies of attention, 5 out of 7 studies of cognitive control, 7 out of 10 studies of other cognitive measures and 12 out of 16 studies of other measures of executive function report a PoC where violent offenders have lower scores than nonviolent offenders. By contrast, few studies reported a preponderance in the opposite direction to that predicted. We cannot add the total of studies because the categories are not mutually exclusive, but if we tally all the comparisons, we estimate that 70% of estimates comparing violent to nonviolent offenders are in the expected direction, with 33.8% statistically significant, which is much more than expected by chance, particularly given the low statistical power. There are too few studies estimating the association between EFs and nonviolent-only offending to draw any conclusions for particular indicators, but if we combine all of the categories of EFs, we see that, so far, out of 14 total studies, only 2 are in the expected direction, 5 report null or ambiguous findings, and 7 report findings in the opposite direction to that predicted. This bolsters the conclusion that executive deficits are more strongly associated with violent than nonviolent offending.