Causality Analysis of Point Events in GIS Environments

Naresh Kumar, Brown University

Almost social and natural events/occurrences happen in geographic context(s) that can be captured as points in GIS environments. Nearest neighborhood analysis (NNA) and autocorrelation methods have been used to evaluate spatial dependency of point events. Lack of geo-referenced data availability hinders causality analysis of point events, since most socio-economic and demographic data are not available for point locations. Advances in GIS support proximity analysis and spatial joins to create area-aggregated social, economic and demographic data for point events that can be directly imported in statistical packages to evaluate causality of the spatial dependency. In this paper causality of crime locations is analyzed in relation to proximity to alcohol services, urban density, income level and ethnic segregation in Savannah City, Georgia in 2000. Similar approach can be used to analyze other point events/incidences including disease patterns. Keywords: GIS, Causality, Point Pattern, Nearest Neighborhood Analysis.