8.1 INTRODUCTION

As a tool, GIS is well suited to storing and manipulating large volumes of complex spatially referenced data. The challenge for the researcher is to turn these data into understanding and meaning. This usually involves simplifying the data sufficiently to make them comprehensible, while avoiding making them over-simplistic. Mapping is one approach to doing this, but, as Chapter 5 described, it has some important limitations. Spatial analysis is another approach that complements mapping. The spatial component of data in a GIS are fundamentally quantitative, as they consist of co-ordinate pairs – numbers that can be manipulated mathematically. Spatial analysis makes use of this to allow us to either summarise the patterns within spatial data, or to ask how attribute data are arranged in space. Having said this, the quantitative nature of spatial data does not mean that spatial analysis is necessarily positivist. In many ways, it can be the reverse, as rather than trying to produce aggregate summaries of an entire study area, it allows us to explore how different parts of the map behave differently. Therefore, rather than searching for similarities, much of spatial analysis is concerned with establishing how different places behave differently.

There are three broad approaches to spatial analysis. The first merely simplifies the map, using forms of smoothing to attempt to enhance the spatial patterns in the data. For example, smoothing a point pattern representing cases of a disease into a raster surface may help detect whether the pattern clusters, and if so, where.