ABSTRACT
This paper examines the ways in which people organise their knowledge of external representations (ERs). These include diagrams, graphs, charts, textual forms such as lists and notations, maps, tables, pictures etc. We focus upon the kinds of ERs that are used in reasoning and problem solving. Twenty-eight subjects were given an untimed card-sorting task. Cluster analysis of the card-sort data revealed that the cards were classified into 9 major categories of ER – maps, set diagrams, logic/math notations, tables & graphs, lists, music, pictures of objects and scientific diagrams. The card sorts of two sub-groups of subjects were compared. The first group was selected on the basis of skilled (upper quartile) performance on ER reasoning tasks and the second was a group of subjects who performed in the lower quartile on ER reasoning tasks. The card-sort behaviour of the better performing subjects differed from that of their poorer performing peers in that they had fewer but more distinct categories. Their ER naming accuracy was also significantly better. The upper-quartile subjects tended to use structural characteristics and ER semantics as a basis for classifying representations to a greater extent than their lower-quartile peers. The results are discussed in relation to theories of category formation and the graphic-linguistic distinction.
