NCJ Number
217460
Journal
Criminal Justice and Behavior Volume: 34 Issue: 2 Dated: February 2007 Pages: 157-178
Date Published
February 2007
Length
22 pages
Annotation
This study sought to validate the original Grasmick et al. (1993) self-control scale using the Rasch model, which transforms raw data into equal interval distances that are free from the tests and samples that were used to collect the data.
Abstract
The study found that although the original Grasmick et al. self-control scale provided criminologists with an excellent starting point for developing a reliable and valid self-control scale, the findings from the current study showed that the original 24-item scale with the four-point Likert-type response categories did not have validity under the Rasch model. Sixteen items from the pool of 24, however, were shown to fit the Rasch model and provide an important advance in measuring low self-control. These findings will be useful in identifying the items that should be removed or revised in the Grasmick scale. The findings do not indicate that attitude measures should be abandoned in studying the self-control theory of crime (Gottfredson and Hirschi (1990). Data were collected in the spring 2003 semester at one university. The study was conducted in five courses open to all majors and five courses that were only open to criminal justice majors. A total of 303 students participated in the study. Most of the sample was White, with an average age of 23. 8 tables, 2 figures, and 58 references