CHUNKS AS INTERFERENCE UNITS IN FREE RECALL

The experiment tested the hypothesis that free recall is limited by the number of chunks that can be reproduced, not the number of words. Recall of a given word should be higher the fewer chunks are required to encompass the remaining words in the list. To test this, a subset of 12 critical words was embedded in a larger list containing 12 fillers composed either of single words, or three-word cliches, or unrelated word triplets. As predicted, recall of the critical words was the same in the single-word and cliche lists, but poorer in the triplet list. Cliches and single words were equally recalled. Total units recalled in the three lists were the same, with only the selection of units differing in recall. The results confirm the chunking hypothesis and have implications for Murdock's (1960) total-time hypothesis about free recall. © 1969 Academic Press, Inc. All rights resered.