ABSTRACT

This chapter suggests a simple tool to visualise and examines the degree of spatial instability in spatial association by means of Moran’s I. It discusses the salient characteristics of techniques for exploratory spatial data analysis. Exploratory spatial data analysis should focus explicitly on the spatial aspects of the data, in the sense of spatial dependence and spatial heterogeneity. The indicators of spatial association from either view that are most relevant for an exploratory approach to spatial data analysis are those that show local patterns and allow for local instabilities. The chapter reviews some methods that have been suggested to deal with local instability in spatial association. It outlines of the ideas behind the Moran scatter plot and a discussion of its properties and potential use. The implementation of a Moran scatterplot is straightforward, since most statistical and many geographic information systems software packages include a scatterplot function and an associated linear regression smoother and indication of fit.