Nested-Means Map Classes for Statistical Maps

ABSTRACT A general, objective method is presented for the calculation of class intervals for statistical maps. The arithmetic mean divides a numerical array into two classes and the means of each of these two map classes yield four map classes with smaller intervals. Repeating the process yields additional means and additional classes with smaller intervals. The result is a series of map classes derived from a nested hierarchy of means; the means are the class limits and also the points of minimum variance, that is, the most representative points for the data values which they classify. The class intervals vary according to the degree of uniformity (in the univariate sense) of the numerical array which is classified for mapping. Thus, class intervals are small in the modal portions of a distribution and large in the tails. The classification technique is replicable and it is based on ordinary univariate statistical procedures.