NCJ Number
192381
Journal
Journal of Criminal Justice Volume: 29 Issue: 6 Dated: 2001 Pages: 493-505
Date Published
2001
Length
13 pages
Annotation
This article examines measurement issues related to crime and deviance and applies the Rasch model to a scale designed to measure Tittle’s control balance ratio related to the analysis of deviant behavior.
Abstract
The discussion notes that classical test theory approaches to measurement rest on fairly restrictive assumptions. The result is that scales intended to examine latent constructs give little attention to the interaction between the human participant and the items that compose the measurement scale. The Rasch model is a competing measurement approach intended to improve upon the limitations of classical test measure. This study administered a control balance questionnaire to 165 undergraduates taking introductory criminal justice courses at a large public institution in the spring of 1998. The instrument contained questions intended to assess control balance in general and the control balance ratio in particular. Application of the Rasch model to a scale designed to measure Tittle’s control balance ration revealed that several items exhibited misfit; these items were removed from the scale. In addition, response category options were reduced from 11 to 4. Moreover, the scales that formed the control balance ratio exhibited very little gender bias. An alternative control balance scale that retained the original items no longer summarized the control balance ratio as one overall score. Findings suggested the need for further empirical research to aid understanding of the merits of each measurement strategy and how each approach relates to deviance. Figures, tables, notes, and 29 references (Author abstract modified)