Testing the construct validity of the PICTS proactive and reactive scores against six putative measures of proactive and reactive criminal thinking

ABSTRACT

This study tested the construct validity of the Psychological Inventory of Criminal Thinking Styles (PICTS) Proactive (P) and Reactive (R) scores. The layperson version of the PICTS was administered to 277 (65 male, 212 female) undergraduates and correlated with putative measures of proactive and reactive criminal thinking. The hypothesis that P and the proactive scales would correlate ≥.30 in zero-order correlations and regression equations controlling for R, whereas R and the reactive scales would correlate ≥.30 in zero-order correlations and regression equations controlling for P found support in this study. This corroborates the construct validity of the PICTS P and R scores and indicates that self-report measures of moral disengagement and neutralization, on the one hand, and impulsivity and risk taking, on the other hand, may serve as effective proxies for proactive and reactive criminal thinking, respectively.

KEYWORDS:

• Psychological Inventory of Criminal Thinking Styles
• construct validity
• proactive
• reactive

Disclosure statement

Glenn D. Walters is the author of the Psychological Inventory of Criminal Thinking Styles (PICTS) and has received small remuneration for sales of the PICTS manual in the past.

ORCiD

Glenn D. Walters http://orcid.org/0000-0002-7219-1542

Notes

1. In male participants (n = 65), 5 out of 6 zero-order correlations between the four proactive measures and all 6 zero-order correlations between the four reactive measures were >.30, compared to 10 out of 16 of the proactive–reactive cross-correlations. Overall, the mean proactive inter-correlation in the male subsample (r = .48) was 30% higher and the mean reactive inter-correlation (r = .53) was 43% higher than the proactive-reactive cross-correlation (r = .37). In female participants (n = 212), 5 out of 6 zero-order correlations between the four proactive measures and all 6 zero-order correlations between the four reactive measures were >.30, compared to 8 out of 16 in the proactive–reactive cross-correlations. Overall, the mean proactive inter-correlation in the female subsample (r = .50) was 47% higher and the mean reactive inter-correlation (r = .56) was 65% higher than the mean proactive–reactive cross-correlation (r = .34).