The effect of a subsidiary task on iconic memory.

Using a subsidiary task technique, Doest and Turvey (1971) concluded that iconic memory was independent of the central processing system. However, they did not control the timing between the short-term memory and the iconic-memory tasks in their first experiment and they used a rather long stimulus duration in their second experiment. These procedural difficulties were rectified here in Experiments I and II. It was found that memory load reduced partial report at all interstimulus intervals and there was no interaction. The results of Experiment I were replicated with auditory presentation in Experiment III, ruling out a masking interpretation. Experiment IV ruled out an interpretation in terms of rehearsal or response competition. It was concluded that iconic memory, like short-term memory, is dependent on the central processing system in the sense that it will suffer in a subsidiary-task situation. The locus of the interference effect appears to be in the encoding stage.