Factors affecting memorial access to analogical similarity

This research explored the hypothesis that repetition of similarities, primarily analogical similarities, promotes the common, underlying relational structure (the "emergent structure hypothesis"). The following predictions were made. Based on prior research, superficially similar items should be recalled better than analogically similar items without repetition. Repetition should improve recall performance for both types of similarity. Support for the emergent structure hypothesis would be garnered if (1) recall of analogies improved more with repetition than recall of superficial similarities, (2) recall of analogies (but not recall of superficial similarities) became superadditive with repetition, and (3) the repetition gain for analogy was more resistant to the deliterious effects of competing similarities than that for superficial matches (as discussed below).The general method across the first three experiments was to present a large study list of unfamiliar proverbs (e.g., 50), containing sets of one to five analogically similar and superficially similar items. The subjects were then given a cued-recall task using, as cues, proverbs that were either analogically similar or superficially similar to the study proverbs. Further predictions were based on the intuition that competition between superficial similarity and analogical similarity may contribute to the low retrievability of prior analogs; in other words, when both superficial and analogical similarities were presented on a study list, recall of superficial similarities will interfere more with analogy recall than the reverse. To test this possibility, dissimilar proverbs (one analogous to the cue, and one superficially similar to the cue) competed for retrieval given the same cue item. The first three experiments tested this similarity-based retrieval competition, as well as repetition, on the retrievability of prior similar events. The results showed repetition increased recall for both analogical and superficial similarities. At one presentation, superficially similar proverbs were retrieved significantly better than analogically similar proverbs, but at five repetitions the difference in retrievability between analogical and superficial similarity decreased substantially (i.e., it was not statistically significant). None of the other predictions for the emergent structure hypothesis were borne out, however, leaving little overall support for the hypothesis. Competition reduced recall overall, but there was no evidence that competition affected retrieval of analogies more than retrieval based on superficial features. Thus, there was also no support for asymmetric similarity competition.The results of Experiments 4a and 4b tested whether subjects were using an implicit recall strategy by directing subjects to recall only analogical similarities. These experiments also tested whether the low recall of analogy in the first experiments was due to the presence of superficial similarities. Directing subjects to recall analogies did not improve recall over neutral instructions; furthermore, it did not appear that the low recall of analogy was due to the presence of superficial similarities.U of I OnlyETDs are only available to UIUC Users without author permissio