ABSTRACT
This chapter introduces the concept of a statistical map, i.e., the use of a thematic map to highlight special features in the spatial distribution of the data, especially the location of outlying observations. A special emphasis is on how these techniques are implemented in GeoDa.
For continuous variables, three types of extreme value maps are considered, employing different criteria to identify outliers. A percentile map focuses on the lowest and highest 1% of the observations. It is a special case of a quantile map. This is also the case for the box map, which is a quartile map with separate categories for the lower and upper outliers (if present), defined in the same manner as for a box plot. A standard deviation map is based on a transformation of the variable into standard deviational units, with values larger than 2 typically considered to be outliers.
Two additional map types are the unique values map, for categorical variables, and the co-location map, which shows the overlap of categories for two variables.
The chapter closes with a brief introduction to the circular cartogram and a simple form of map animation, in the form of Geoda's map movie.
