The estimation of graduated symbols in thematic maps

Abstract

This study deals with the problems which originate from the presentation of data using graduated symbols.Within the last 35 years a large number of cartographical studies concerning this problem have been published. Their results, however, are most often based on statistical parameters, especially on medians, which of course characterize them best by a single number. Too little attention is paid, however, to the range of the results. In this study efforts are made to visualize the whole range of estimates to find the most typical errors which dominate the process of estimation. Only a full understanding of these errors provides a sound foundation for drawing up thematic maps and their legends so that mistakes in interpretation can be kept within acceptable limits.