The influence of relationships among items to be recalled upon short-term retention

Abstract

Summary Four experiments are reported in which Ss were required to remember familiar English words until their recall was requested. While remembering these words, Ss were presented other words or asked to recall previously presented items. In all four experiments, the recall of more than one item at a time was typically required. The studies are concerned with the manner in which short-term retention is influenced by the relationships among words whose recall is requested at the same time. Retention was found to be facilitated by the opportunity to group words of the same generic class (e.g., automobiles, trees, sports, etc.) at recall. On the other hand, if words from the same class had to be simultaneously stored in different groupings for recall, the resulting interference led to poorer retention than was the case when Ss were required to remember only unrelated words at any one time. It was found to make little difference whether the same or different Ss were responding under the various conditions of opportunity to group. Moreover, if the opportunity to group is varied in a stepwise fashion from the condition in which maximum grouping is possible to the condition in which S is required to remember only unrelated words, retention scores are found to be ordered with respect to this variation. However, it is interesting to note that under the task conditions of these experiments, retention is benefited by multiple-item recall, i.e., increased average load reduction, even when the items being recalled are unrelated. While it is possible that these benefits may have been derived in part from the grouping of unrelated items at recall, it seems more likely that the improvement was due to a decrease in retroactive inhibition that resulted from the concomitant decrease in number of interpolated recall points with increases in average load reduction.