NCJ Number
137286
Journal
Journal of Quantitative Criminology Volume: 8 Issue: 2 Dated: (June 1992) Pages: 233-245
Date Published
1992
Length
13 pages
Annotation
This study examines the reduction of the number of items in the scales of the Child Behavior Checklist, while maintaining the reliability and predictability of its original scales.
Abstract
The original scales were replicated with data from the Pittsburgh Youth Study. New, abbreviated scales were constructed using the most potent items. This process yielded trimmed scales that were highly correlated with the full scales and that still reflect the original meanings of the scales. The trimmed scales used approximately one-third (45 items) of the 118 items in the full Child Behavior Checklist. The trimmed scales showed a good internal consistency through the use of data from the Pittsburgh Youth Study; this was replicated with data from the Rochester Youth Development Study. The trimming resulted in a modest reduction of the variance of the factors that explain the variables. Externalizing scales and, to a lesser extent, internalizing scales predicted self-reported delinquency equally well whether they were saturated or trimmed. The trimmed scales worked equally well across research sites despite sizable differences in mean levels of delinquency between the two samples. Thus, for delinquency research, the trimmed scales may be a cost-effective alternative to the full scales. 2 tables and 4 references