Physical versus semantic class recall in audio-visual split-span memory

Abstract

Class recall grouping based on physical attributes of verbal stimuli was compared with grouping based on semantic attributes, in an audio-visual split-span memory task. It was found that on lists which required no alternation between modalities, semantic recall was superior to physical recall, whereas on lists which required subjects to alternate between modalities, physical recall was more efficient than semantic recall. The results are discussed in terms of input processes and output processes. It is suggested that the existing hypotheses about split-span grouping need to be extended to account for differences between physical and semantic recall.