NCJ Number
92338
Editor(s)
G P Waldo
Date Published
1983
Length
133 pages
Annotation
These seven papers address measurement issues related to the most appropriate operational definitions to use in measuring concepts; the proper techniques to use in selecting samples, items, and the data collection process; and decisions relating to the measurement process.
Abstract
Several chapters deal with the measurement of crime. The development of a perceptually-based offense seriousness scale is examined, with emphasis on the assumption of prior researchers that these scales provide a more refined measure of offense seriousness. The usefulness of a perceptually-based measure in predicting sentencing decisions made by judges is explored. The use of randomized response techniques to increase truthfulness in self-reporting is examined, with an application in which colored dice were used in the randomization process. The extent to which interviewers' characteristics and attitudes affect the results of self-report studies are discussed, using a national probability sample of youths. A comparison of self-reports and official data in terms of the relationship between social status and delinquency focuses on differential perceptions of police and juveniles. Substantive versus statistical significance in evaluating research results is discussed, emphasizing the importance of statistical power in evaluating research findings. The comparability of different data sources that supposedly measure the same phenomenon (homicide) is examined. The use of ridge regression and the principal components procedure to deal with independent but related variables and to reduce the number of dimensions is explained, with an application to the relationship between crime and economic factors. Figures, data tables, chapter reference lists, and author biographies are provided. For individual articles, see NCJ-92239-45.