Assessing the separate contributions of recollection and familiarity for the mirror effect: A process dissociation approach

In four separate experiments, we employed Jacoby\u27s (1991) process dissociation procedure (PDP) to evaluate the extent to which the mirror effect in recognition memory can be accounted for by recollection and In the first experiment, participants listened to two study lists of words, one in a female voice and the other in a male voice. Their memory for one set of words was gauged via an inclusion test, while their recognition of the other set was gauged via an exclusion test As predicted, a significant mirror effect was observed for both the inclusion and exclusion conditions, and the PDP recollection estimate was higher for low than for high frequency words. Contrary to the dual component hypothesis, however, was the finding that the PDP estimate of familiarity was also greater for low than for high frequency words. In Experiment 2, we manipulated the difficulty of the source discrimination test by having participants and between an auditorily and a visually presented list. A recognition advantage for low frequency words was obtained for new items only, and the PDP recollection estimate and independence estimate of familiarity were greater on the auditory/visual discrimination test Source discriminability was also manipulated in Experiment 3 using a nouns versus nonnouns stimulus comparison. Superior recognition of nouns was manifested in new items, as well as list 2 items, and the independence estimate of familiarity was greater in the visual/visual condition Nonetheless, the recollection estimate did not differ between the visual/visual and auditory/visual treatments. Word frequency was again manipulated in Experiment 4, with test instructions run as a between-subjects variable. A mirror effect favouring low frequency words emerged under inclusion instructions, as well as for new and list 2 items under exclusion instructions. Although the estimates of familiarity calculated with both redundancy and independence were greater for low than for high frequency targets, recollection did not differ as a function of word frequency. These results are discussed in terms of their implications for dual component and source monitoring theories of recognition memory, and some concerns regarding the use of the use PD procedure to investigate the mirror effect are raised