Date of this Version
From The Influence of Cognitive Psychology on Testing, edited by Royce R. Ronning, John A. Glover, Jane C. Conoley, and Joseph C. Witt (Hillsdale, NJ: Lawrence Erlbaum Associates, 1987)
The goal of cognitive psychology is to provide a general understanding of human cognitive processes through the development of general, formal models of cognition. Although it is clearly true that some areas (such as memory) have been more highly developed than others, it is undeniable that cognitive psychology has witnessed a proliferation of models in the past decade. Perhaps researchers are finding it increasingly difficult to discriminate among competing memory models because the constraints are so weak. One possibility that will be explored in this chapter is the prospect of using multidimensional scaling (MDS) and related procedures as a means of providing constraint for theorizing.
In this chapter, we initially provide a brief description of the problem of the inability to distinguish among models. Subsequently, we sketch some scaling and clustering procedures. We then discuss a number of applications of MDS and related procedures to domains of interest to cognitive psychologists. Particular attention is given to the constraint provided by these techniques on cognitive theorizing. Subsequently, we outline how one might choose the correct procedure and how one might circumvent some problems raised by using these procedures to study cognitive domains . Next , we provide a brief application of these procedures to the domain of cognitive psychology models. Finally, we attempt to provide an assessment of the utility of MDS and related procedures in cognitive psychology.
One particularly salient example of the difficulty in telling seemingly contradictory theories apart is the recent dispute over the viability of the semantic/episodic distinction in human memory. Briefly, Tulving (1983) has proposed that the human memory system can profitably be divided into memory for general world knowledge (semantic) and memory for personal events (episodic). In contrast, other theorists have claimed that a unitary theory of memory provides a better account (Anderson & Ross, 1980; McKoon, Ratcliff, & Dell, in press).
It would certainly seem that two theoretical viewpoints that differ in the desirability of partitioning the memory system along such fundamental lines should be easy to tell apart. In fact, this goal has proven elusive. To date, the most conclusive kind of evidence on this issue is the dissociation experiment in which one examines the effects of an independent variable on an episodic memory task and a semantic one. If we find that the variable has different effects on the two tasks, then according to Tulving (1983) we have evidence for the distinction.