NCJ Number
218384
Journal
Policing: An International Journal of Police Strategies & Management Volume: 30 Issue: 1 Dated: 2007 Pages: 32-60
Date Published
2007
Length
29 pages
Annotation
Using 3 datasets of 100 cases each, this study examined data quality in analyzing geographic crime patterns through software applications in geographic information systems (GIS).
Abstract
The study found that in mapping geographic patterns for crime, street directional indicators (north/south and east/west) and zip codes were more important than previously thought. The findings reveal the need for crime mapping to provide complete descriptions of research methodology. All geographic crime analyses must include information about the "hit" rate (percent of cases plotted), characteristics and capabilities of the software being used, and the process used to input tabular crime data. Other information critical to crime mapping includes the software parameters set for importing data (geocoding preferences), reference information about the street file used, and information on the missing cases so as to identify some of the sampling error. When predicting geographic crime patterns or identifying "hot spots" (geographic areas where crime is concentrated), analysts must be aware of any bias inherent in the data imported and the way it is processed by the GIS software. In order to determine which street address components have the most impact on the crime mapping process and the nature of the sampling error typically introduced, this study adjusted user-defined settings while geocoding three different datasets in ArcView 3.3 and ArcGIS. Each dataset was geocoded at different preference settings in order to determine the accuracy of the plotted points. Plotted points were then compared to actual locations in order to determine the accuracy of the position. Seventy-six tests were conducted. 7 tables, 26 notes, and 32 references