Relationship Between Elements and Manifestations of Low Self-control in a General Theory of Crime: Two Comments and a Test

• NCJ Number: 197624
• Journal: Deviant Behavior Volume: 23 Issue: 6 Dated: November/December 2002 Pages: 531-557
• Author(s): Stelios Stylianou
• Date Published: November 2002
• Length: 27 pages
• Publication Format: Article
• Publication Type: Report (Study/Research)
• Language: English
• Country: United States of America

Annotation

This article discusses the internal validity of crime causal models and the distinction between statistical and substantive significance.

Abstract

The general theory of crime states that, controlling for opportunity and age, a single social psychological dimension – self-control – determines an individual’s likelihood of involvement in deviant behavior. Low self-control causes deviant behavior. The conceptualizations of self-control are impulsivity, preference for simple tasks, risk seeking, physicality, self-centeredness, and temper. The failure to make the distinction between independent and dependent variables is present in numerous tests that report support for the general theory of crime. A solution to this problem is the strict distinction between attitudes as elements and behaviors as manifestations of low self-control. In testing the general theory of crime, effects have been characterized based solely on the statistical significance of regression coefficients. A more informative approach is the substantive interpretation of these effects. This allows researchers to know more precisely how much low self-control it takes to predict certain behaviors. In a test of the causal hypothesis, attitudes were used as elements of self-control and self-reported behaviors were used as manifestations. Conclusions were drawn based on both statistical and substantive criteria. It was shown that, although many of the effects were statistically significant and in the predicted direction, they were typically very weak in substantive terms. Substantive interpretation of effects must be made available in published research. Careful examination should be made of the existing support for the general theory of crime. These considerations and the results of the present test suggest that the issue of effect size is still open to substantive evaluation. These concerns become even more important as low self-control is gaining an increasingly secure position among the correlates of deviant behavior. Tables, references