Recognition Induced Forgetting: The Role of Visual and Semantic Similarity
Loading...
Authors
Crocco, Lisa N.
Date
Type
thesis
Language
eng
Keyword
Recognition Memory , Forgetting , Similarity
Alternative Title
Abstract
Recognition-induced forgetting (RecIF) is a phenomenon that occurs when recognizing one visual item impairs memory for another related item. This study aims to address the limited exploration of RecIF mechanisms, often tied to semantic categorization. The first objective was to investigate visual similarity as a trigger for RecIF, assessing the need for semantic categorization. The second objective was to elucidate whether the inhibition model or the feature suppression model best explains RecIF. The inhibition model involves competition among similar memories, where an inhibitory mechanism suppresses the memory of items similar to retrieved items. The feature suppression model emphasizes that during the practice phase shared features with a practiced item are enhanced, while differing features are suppressed. In Experiment 1, visual similarity was isolated to study RecIF independently of semantic categorization. Participants engaged in a modified RecIF paradigm where they studied 60 sets of three items simultaneously: a target item (palm tree), an item similar to the target (broccoli) (i.e., in terms of shape, colour, and/or orientation), and an item dissimilar to the target (shrimp). In the subsequent Practice Phase, only half of the target items (excluding the similar or dissimilar items) were presented from the Study Phase, alongside 30 novel items. In the Test Phase, participants identified studied items paired with novel items, reporting confidence in their memory responses. Results revealed a significant RecIF effect across all levels of similarity (low, moderate, and high). High confidence responses triggered RecIF across the low, moderate, and high similarity conditions, providing evidence with the inhibition model. In Experiment 2, I investigated the impact of a semantic similarity manipulation on RecIF. The procedure resembled Experiment 1, the difference being that items were categorized based on superordinate (i.e., potato and rocking horse), basic (i.e., basketball and ping pong paddle), or subordinate (i.e., water gun and shotgun) levels for the similarity assessment. Results showed a significant RecIF effect at the superordinate similarity level, supporting the feature suppression model. Together, these results suggest that both feature suppression and inhibition models contribute to RecIF with their roles dependent on the type of stimuli information provided.