Poster Session D, Monday, March 25, 8:00 – 10:00 am, Pacific Concourse
Semantic influences on episodic memory distortions
Alexa Tompary1, Sharon L. Thompson-Schill1; 1University of Pennsylvania
Prior knowledge can facilitate or impair the formation of new episodic memories, depending on the consistency between old and new information. Although this phenomenon is well studied, past work rarely considers how the structure of semantic knowledge leads to variation in how it biases newly acquired information. We aim to quantify distortions in episodic memories due to prior knowledge by examining how differences in category membership bias new encoding. In experiments conducted on Amazon Mechanical Turk, participants (n=25) encoded and retrieved 70 image-location associations by dragging each image to its associated location. We used this continuous retrieval measure to disentangle biases driven by semantic knowledge from errors due to forgetting. Critically, the locations associated with each image were determined by similarity ratings generated by a separate cohort, such that members of the same category (e.g. birds) were located near each other. These ratings were also used to identify typical and atypical members of each category, which were assigned to random locations that were inconsistent with the similarity ratings. First, memory was more precise for image-location associations that were consistent with similarity ratings, relative to those that were inconsistent. Second, retrieval of typical category members was more biased towards the locations of their semantic neighbors, relative to retrieval of atypical members. Both effects replicated in a separate cohort (n=35) and were disrupted in control experiments where images’ locations were not related to semantic ratings (n=43). Taken together, this suggests the structure of semantic knowledge can bias encoding of new memories.
Topic Area: LONG-TERM MEMORY: Episodic