This chapter of the Guide focuses on a range of exploratory data analysis techniques and statistical methods that have been implemented in widely available GIS and GIS-related software packages. Initially we describe a range of simple statistical facilities that are provided within GIS packages, notably those supporting descriptive statistical measures and special facilities relating to spatial sampling.
Section 5.2, Exploratory Spatial Data Analysis, then addresses the question of exploratory data analysis in an explicitly spatial context and describes a range of tools currently available that support such analysis as a precursor to further investigation and modeling. Aspects of these methods have been touched upon in elsewhere in this Guide, notably in Section 4.3, Queries, Computations and Density.
Sections 5.3, 5.4 and 5.5 then examine three of the main areas of spatial pattern analysis: Grid-based Statistics and Metrics; Point Sets and Distance Statistics; and Spatial Autocorrelation. Each area is covered in some detail, although a truly comprehensive coverage of tools and techniques lies beyond the space available here. Readers are recommended to refer to the many references and toolsets cited in the text for a more complete picture of each area, especially in relation to software that has been for specific disciplines such as epidemiology, spatial econometrics or ecological analysis.
Finally, in Section 5.6, Spatial Regression, building on the material in Section 5.5, we make a brief foray into the field of statistical modeling using regression techniques. We describe and illustrate some of the main approaches that have been developed to tackle the specific difficulties that arise with spatial datasets.
Readers are also recommended to refer to the Afterword on "Big Data and GIS" and the associated presentation resource: http://www.spatialanalysisonline.com/PPTS/BigData.pdf as this is becoming of increasing importance in any consideration of approaches to spatial data analysis.