The Application of New Spatial Statistical Methods to the Detection of Geographical Patterns of Crime

The evolution of geographic information systems has spawned a resurgence in the development of spatial statistical methods. In this chapter, several recent developments in the field of spatial statistical analysis are summarised and applied to burglary data for census tracts in Buffalo, New York. The methods employed here are concerned with the detection of spatial patterns in geographic data. Some of the methods are concerned with finding regions with rates that are significantly higher than those that could be expected under the null hypothesis of no geographic pattern, relative to prior expectations. Others have as their goal the detection of temporal change in spatial data.