Exploratory spatial data analysis of missing and found persons in Vancouver

This thesis explored methods that can be used to improve the spatial analysis of crime. The emphasis was on exploratory spatial data analysis (ESDA) in the study of missing and found persons calls for service in Vancouver in 1996. Traditionally, spatial crime analysis has been limited to the cartographic display of clustering or hot spots of criminal events. Research in spatial clustering typically uses limited statistical analysis. This thesis addressed the value of spatial statistics in understanding crime patterns by using global and local methods of analysis on non-standardized and standardized missing and found persons data. Local analysis indicated that clustering exists within both the missing and found persons calls for service. The found persons point data indicate stronger clustering and less dispersion; whereas the missing persons point data illustrate a more dispersed pattern highlighting multiple clusters existing within Vancouver. Findings on a global level indicate that the non-standardized data demonstrate spatial autocorrelation and association within both the missing and found persons datasets; whereas analysis conducted on the standardized data indicate no spatial autocorrelation within the missing persons data. Spatial autocorrelation was present when analysis was conducted on found persons data, and high concentrations (clusters) were present specifically in the downtown east side of Vancouver.