Cued and uncued free recall of unrelated words following interpolated learning.

Abstract

.1 The 5s were given free-recall training to criterion with a list of 30 unrelated words, original learning (OL), were given seven trials with a second list, and were then tested for OL retention. Half of the 5s were cued with list members taken from alternate positions of the final recall of OL, half were not cued. Cued 5s retained significantly more critical items than did control 5s. Although similar results have been offered in support of the trace-dependent storage view, it was argued that such studies do not provide a differential test of independent vs. dependent storage. In two recent papers (Slamecka, 1968, 1969), it has been argued that in free-recall learning, memory traces for the words of the list are stored independently. Organization appearing in recall order is said to result from the use of an organized retrieval plan. In addition, the relative efficiency of such organized retrieval plans is assumed to produce the heightened recall usually associated with increases in recall organization. The alternative position is that organization takes place at the time of storage; i.e., that traces of list members are stored in some dependent fashion. According to this view, recall of one member of a set of relationally stored items (S unit) is assumed to cue recall of other items from the set. Thus, this trace-dependent storage hypothesis attributes both organization in recall order and concomitant increases in recall to organization in storage. Slamecka (1968, 1969) has argued that if, as indicated by the trace-dependent storage hypothesis, recall of one member of an S unit cues recall of the remaining members of that unit, then providing a member of that unit at the time of recall should have a similar effect. In a series of six experiments, Slamecka (1968) compared recall when some of the list members were provided as cues with recall when no such cues were provided. In none of these was cued recall found to be superior to uncued recall. In a subsequent series of experiments, Slamecka (1969) presented lists of unrelated words in constant serial order, cuing half of the 5s with words taken from odd-numbered serial positions in the presentation order. He again found the recall of cued 5s to be no greater than that of uncued controls. These findings were taken as evidence that item traces are not stored dependently. Subsequent investigators have pointed out that