Classifying graphical information

The research lays the groundwork for a taxonomy of visual representations by establishing a methodology for determining the kinds of knowledge conveyed by different graphical representations. In the first of two experiments, the basic categories and dimensions of a set of graphics were established using a sorting procedure. Five principal categories emerged: graphs/tables, maps, diagrams, networks, and icons. Furthermore, two principal dimensions characterize these groups: amount of spatial information and amount of cognitive processing effort. The second experiment validated and extended this understanding of the cognitive structure of visual representations. In that experiment, similarity among items was assessed using pairwise similarity judgements. The results confirmed the original categories and revealed distinct differences between subjects who did or did not have graphic arts training.