DELAYED MATCHING-TO-SAMPLE PERFORMANCE AS A MEASURE OF HUMAN VISUOSPATIAL WORKING MEMORY

The characteristics of the cognitive processes or systems involved in temporary maintenance of visual information are as yet unclear. In the present delayed matching procedure, subjects judged whether two visual patterns were identical under conditions in which patterns could vary in size and delay-interval length varied. To address the suggestion that a component of working memory is specialized for short-term visual storage, the type of distractor activity interpolated in the delay interval varied in terms of apparent differences in demands for working memory resources. Analyses of errors indicated that judgments of size disparity between two successively presented visual patterns were influenced by the length of the delay interval, by the specific processing activity interpolated in the delay, and by the temporal arrangement relating distractor activity and memory testing. Visual recognition was interfered with when memory testing occurred concurrently with the terminal processing of distractor activity, but not when testing occurred following termination of such distractor-activity processing. This result supports the view that a passive, visuospatial slave system-rather than central processing resources-is responsible for temporary maintenance of a visuospatial stimulus (Logie, Zucco, & Baddeley, 1990). Data from delayed matching tasks may contribute to the theoretical development of a working memory system.