NCJ Number
139427
Journal
Journal of Offender Rehabilitation Volume: 18 Issue: 1/2 Dated: (1992) Pages: 1-80)
Date Published
1992
Length
80 pages
Annotation
A new scaling method for describing crime statistics reported in Uniform Crime Reports (UCR), based on the scaling of intelligence used in differential psychology, is used to rescale reported crime occurrences in 76 large U.S. cities.
Abstract
One of the greatest contributions of psychology to understanding human behavior has been the development of psychometrically reliable and valid methods for assessing intelligence. To provide a scaling system that has equal standard deviations, psychometricians employ a relatively simple transformation methodology known as z-scaling or standard score scaling. The primary purpose of this transformation is to measure how far away from the distribution mean a score is in standard distance units so that comparisons across distributions with different raw score values, means, and dispersions can be made. Although several early psychological tests used z-scores or transformations of them, performance reporting on the Stanford-Binet Intelligence Scale is based on ratio IQ concepts developed by Binet and Stern. The Wechsler Adult Intelligence Scale, the Wechsler Intelligence Scale for Children, and the Wechsler Preschool Intelligence Tests are comprised of six verbal and five performance subtests that yield a global IQ score. Wechsler's verbal and performance subtests were employed to evaluate UCR data for violent and property crimes during 1989. The number of reported criminal events was obtained in 76 cities for murder, forcible rape, robbery, assault, larceny, burglary, motor vehicle theft, and arson. Data on demographic and fiscal characteristics of the 76 cities were obtained, and z-scores for the Crime Index Ratio, the Modified Crime Ratio, and ratios for specific crime indexes were converted to Wechsler-like scales. Crime quotients were examined in relation to four structural characteristics of the 76 cities (municipal expenditures, police service expenditures, minority percent of the population, and population density). The results of converting crime rate statistics to interval scales patterned after intelligence test scaling indicate the new method's feasibility and value for crime analysis. Psychometric and practical issues associated with the method are discussed. 16 references, 18 tables, and 3 figures