Procedures for cleaning, geocoding, and aggregating crime incident data

This technical document describes the geocoding and cleaning procedures used when geocoding crime data in Southern California cities for the project, Crime in Metropolitan America: Patterns and Trends across the Southern California Landscape, a project funded by the National Institute of Justice (NIJ). A key component of this project entailed the collection of incident crime data for cities located in the greater Southern California region. Cooperating police agencies reported incident crime data with geographic information for as many years of the study period as possible (2000-2013). Crime events were geocoded through a procedure using a geographic information system (ArcGIS 10), as well as two other procedures to attempt to geocode cases not geocoded by ArcGIS.

Synthetic Estimation To Impute Values For Small Units Of Analysis

The synthetic estimation approach relies on the assumption that the relationship between variables at one level of analysis is similar at a different level of analysis, which is certainly not ideal. Nonetheless, whereas researchers often simply impute values from the larger units to the smaller units assuming homogeneity within the larger units, the synthetic estimation approach is more principled in attempting to build a model to predict such values.

Boessen, A., & Hipp, J. R. (2015). Close‐ups and the scale of ecology: Land uses and the geography of social context and crime. Criminology, 53(3), 399-426. doi:10.1111/1745-9125.12074