Sharing verbal and visuospatial resources in working memory

A digit memory task and a delayed visual aiming task were used in a test of resource capacity theories. These tasks were run either singly, with short- or long-retention intervals, or in pairs. In the dual-task conditions, the short version of one was inserted into the longer version of the other. Delay by itself had no effect on retention, but the dual-task combinations showed asymmetrical interference: Inserting verbal memory into the visual-motor delay had little effect on either task, but inserting the motor task into the verbal retention interval disrupted both. These results thus support neither single nor multiple resource models, but can be explained with reference to component processes in working memory.