Comparing Traditional Methods for Selecting Class Intervals on Choropleth Maps

The most common goal when classing data for choropleth maps is to create homogeneous classes which contain similar data values. None of the four traditional data classing methods examined here (quartile, equal interval, standard deviation, and natural breaks) consistently generalized the experimental data sets into homogeneous classes. These methods were most accurate for data sets with specific distributional characteristics, but none classed all of any type of distribution accurately. Only the optimization method produced reliable and accurate results for all of the experimental data.