U.S. flag

An official website of the United States government, Department of Justice.

Exploratory and Confirmatory Spatial Data Analysis Approaches to Studying the Correlates of Juvenile Violent Crimes, Volume II Final Report

NCJ Number
Date Published
57 pages
This study used spatial methodologies to study variations in juvenile violent crimes in Virginia, with the goals of determining whether counties with high juvenile violent crime rates tended to cluster together and identifying the risk factors that were the primary determinants of juvenile violent crime at the county level.
The first chapter reports on the implementation of exploratory spatial data analysis methods (ESDA) to study both the local and global patterns of juvenile violence, with a focus on the applications of ESDA as a tool for juvenile justice planning. The tools used were geographical information system (GIS) software and spatial statistics software. The GIS software was used for defining the spatial weights and for the visualization of the results. The spatial statistics software was used for both the exploratory and confirmatory analysis. As a first step, basic descriptive and exploratory methods were used to identify mean trends in juvenile violent crimes. In step 2, the Moran scatterplot map was used to develop a typology of counties with varying spatial patterns of juvenile violent crimes. In step 3, the Local Moran statistics, the Moran Scatterplot, and the Moran significance map were used to identify spatial clusters of juvenile violent crimes, as well as counties that were significantly different from their neighbors. This process was implemented for both the levels of juvenile violent crime rates and for changes in juvenile violent crime rates. Results show that ESDA can be used by State juvenile justice planning staff to monitor spatial and temporal variations in juvenile violent crimes and other key measures; ESDA can also assist in identifying "problem counties/cities" that might be difficult to identify otherwise. The second chapter of this report discusses the implementation of confirmatory spatial data analysis techniques to identify the correlates of juvenile violent crimes. Following Anselin et al. (2000), the study explored both the spatial interaction (spatial dependence) and the spatial structure (spatial heterogeneity) that linked risk factors to juvenile violent crime rates. The focus on spatial heterogeneity permitted the examination of the linkages between risk factors and juvenile violent crimes across the various regions in Virginia. The diffusive nature of juvenile violent crimes was also examined at the county level. With the dependent measure being juvenile violent crime, the independent measures were community health risk, risky behavior among teens (teen birth rates), risk of violence in school, school performance, family risks, resource deprivation and affluence component, and population structure component. Statistically significant relationships were observed between community health risks, teen risks, population dimension, and juvenile violent crime rates. Extensive tables and 55 references

Date Published: January 1, 2001