ABSTRACT
The chapter on maps for rates treats the important concept of variance instability associated with rates or proportions for which the denominator is a small value, the so-called small area problem. The difference between spatially extensive and spatially intensive variables is explained and illustrated. The GeoDa interface allows rates to be calculated on the fly and visualized by means of the usual range of thematic map classifications.
A special map category is the excess risk map, which highlights locations where the rate exceeds a reference, such as the rate computed for a larger area. This concept is the same as the standard mortality rate used in demography and the location quotient used in economic geography. This is another example of an extreme value map.
The important topic of rate smoothing is introduced, and its conceptual basis in the Bayesian concept of borrowing strength explained. The application of Bayes Law to correct for the variance instability in rates is illustrated as well as its visualization in Empirical Bayes smoothed rate maps.
