NCJ Number
201074
Journal
Justice Quarterly Volume: 20 Issue: 2 Dated: June 2003 Pages: 241-263
Date Published
June 2003
Length
23 pages
Annotation
This study examined the validity and reliability of the Grasmick et al. scale (1993), which contains 24 attitudinal items, 4 for each of the 6 characteristics of self-control described by Gottfredson and Hirschi (1990).
Abstract
The six characteristics associated with level of self-control are impulsivity, simple tasks, risk-seeking, physical activities, self-centeredness, and temper. This study focused on whether self-control works differently for different populations; whether self-control is equally salient among low-risk samples and high-risk samples; the dimensionality of the scale; whether the scale is a valid and reliable indicator of self-control; and if so, whether it can be effectively used to predict assorted outcomes among diverse samples. The study first replicated the original scale and applied it to a sample of formerly incarcerated men (n=208), which extended the applicability of the scale to a criminal sample. This is an important step in determining the generality of self-control and its ability to explain offending among populations of offenders that are of most concern to society. The study also conducted exploratory and confirmatory factor analyses to assess the dimensions of the scale. Finally, the study used a measurement model refined by modification indexes in a structural equation model to examine the direct effects of self-control on two crime-based outcome variables. The findings indicate that the Grasmick et al. scale is not unidimensional, and more complex theoretical iterations failed to meet most goodness-of-fit statistics. Further, significant refinement through modification indexes was required before a measurement model that fit the data could be found. The study concluded that further refinement is required to justify the Grasmick et al. scale as the quintessential measure of self-control. 5 tables, 1 figure, and 64 references