Superiority of Free Recall to Cued Recall with "Strong" Cues

Subjects encoded homographs (e.g., iron) in one of two contexts, produced by preceding them with words from two different categories to which the item belonged (e.g., items that were metals or utensils). Later, subjects were tested either under free-recall conditions, or with one or the other of the category names that were either congruous or incongruous with the original encoding of the targets. Recall with cues that were congruous with the encoding of the target produced performance superior to free recall, but free recall was superior to cued recall when the cues were not congruous with the original encoding. The results were interpreted within the encoding specificity framework, in particular as confirming the debated proposit ion that a cue strongly associated to a target i tem may be ineffective in promoting retrieval of the target if the relation was not encoded when the target was studied. Previous failures to confirm this prediction may have been due to inadequate manipulation of the encoding context in order to provide distinctive encodings.