ABSTRACT
MANY OF THE STATISTICAL techniques that geographers commonly use tend to be of a non-spatial type that can just as easily be applied to any other scientific discipline. Indeed, the majority of the techniques mentioned so far in this book fall into this category, where location is but one of many variables under examination. However, there also exists a set of spatial statistics which allow the geographer to summarise and describe numerically a variety of spatial patterns ranging from simple dot distributions through to contour maps. It is possible to recognise a typology of maps that loosely correspond to the nominal, ordinal, interval and ratio scales of measurement that we discussed in Chapter 4. In this way a picture can be built up of the types of data we are likely to want to analyse using different spatial statistics. To carry out such spatial analysis three major groups of techniques are available, namely: centrographic techniques, spatial indices and point pattern analysis. It has, however, to be observed that while such methods may fall clearly within the realm of geographical research, they are poorly served by commercially available packages such as MINITAB and SPSS that are designed for a wide range of discipline-based uses.
