Analysis of Environmental Attribute/Crime Incident Characteristic Interrelationships (A Methodological Review) (From Link Between Crime and the Built Environment, Volume 2, P C351-C360, 1980, by Tetsuro Motoyama et al - See NCJ-79544)

• NCJ Number: 79573
• Author(s): T Motoyama; S Shore; H Rubenstein; P Hartjens
• Date Published: 1980
• Length: 10 pages
• Corporate Author: American Institutes for Research
• Sale Source: National Institute of Justice/
• Language: English
• Country: United States of America
• Note: Available on microfiche from NCJRS as NCJ-79544.

Annotation

This review assesses Christopher Dunn's analysis of crime incident data on assault, robbery, burglary, and auto theft in interrelationship with the social attributes of the areas where the crimes occurred.

Abstract

The study analyzed data on the type and location (census tract) of crime in Westchester County, New York. Since each unit of observation consisted of many social and crime-related variables, the study used cumulative communality key-cluster analysis to limit the number of variables under consideration at any one time. The variables were grouped into clusters through consolidating the 30 census tract variables into 4 clusters dealing with size and structure of household, sex and age of population, variables related to income and poverty, and other socioeconomic characteristics. The study then analyzed the 202 census tracts along the 4 dimensions and developed 9 groupings of census tracts termed 'social area.' The distribution of assault, robbery, burglary, and auto theft rates were then analyzed in relation to the clusters of social areas to determine whether the mean crime rate of each social area was significantly higher or lower than would be expected from chance distribution. The study concluded that crime rates for the four types of crime are significantly different in various types of social areas and that specific types of assault, burglary, robbery, and auto theft occur more frequently in certain types of social areas. Since the study probably used the cluster analysis incorrectly, however, study findings are inconclusive. More importantly, due to the nature of the study, consistent relationships between crime-related behavior and particular physical characteristics of the built environment cannot be determined. For the original report, see NCJ 25714. (Author summary modified)