Predicting Visual Clusters on Graduated Circle Maps

Abstract

Abstract One of the basic challenges facing geographic authors is to determine how effectively the maps in their papers and books communicate information. Such a challenge can be met if the geographer has a method available for predicting the spatial patterns that readers are likely to perceive. By comparing these predicted patterns with their own, geographers can determine the communicative effectiveness of maps they use. A logit model can predict map readers' responses to one particular type of pattern feature, the visual cluster, on graduated circle maps. The model, based on measured map characteristics, correctly predicts 92 percent of all circles as clustered or unclustered and generally predicts the correct number of clusters. Although the model does not predict clusters equally well on all maps, it does provide a useful starting point for further research on this topic. Keywords: PERCEPTIONPATTERNSGRADUATED CIRCLESLOGIT MODEL