Abstract

This entry introduces some key principles around spatial data (i.e., mappable data) in the social sciences and geostatistical methods for their analysis. It takes as its main focus areal data. An example of such data is counts of people within census areas. The results of analysis of such data are dependent on the size and shape of the areas, and this entry considers the implications of the choice of areas for spatial analysis. It also introduces key concepts such as spatial dependence—the tendency for neighbouring data values (e.g., unemployment rates in census areas) to be similar. Methods for analysing spatial dependence are outlined next. The analysis of spatially varying relationships is then discussed, allowing for the possibility that relationships between variables (e.g., poor health and deprivation) may not be the same at all locations. Next, the entry details geostatistical methods for analysing the spatial structure of variables (e.g., the scales over which unemployment rates are concentrated). Finally, methods for overcoming scale effects by reallocating data from one set of areal units to another are detailed.