Semantic clustering and semantic path length: Two measures of free recall organization with different functional properties

Poster Session B - Sunday, April 14, 2024, 8:00 – 10:00 am EDT, Sheraton Hall ABC

Felicia Chaisson¹ (fchais1@lsu.edu), Heather D. Lucas¹, Amber Alford¹, Christopher R. Cox¹; ¹Louisiana State University

In free recall studies, the order in which words are recalled provides important insight into how memories are organized. Semantic organization is often measured via a clustering score in which the semantic distance of each recall transition is ranked against all possible distances given the entire set of not-yet-recalled words. One potential limitation is that this measure assumes that all items presented during the study phase were equally available for encoding, which may not be the case for very long study blocks or other situations in which attention is likely to fluctuate. Here, we introduce an alternate “path length” measure that involves computing the overall distance “traveled” during a recall attempt and standardizing it against a null distribution created from random permutations of only the set of words that were actually recalled, while disregarding the non-recalled words. To compare the functional properties of these measures, we applied them to free recall data obtained in two experiments in which acute anxiety was manipulated during study. In Exp. 1, participants studied two blocks of 72 unrelated words, while Exp. 2 included two blocks of 48 words from four categories. Although the path length and clustering measures were highly correlated across participants, path length showed a consistently stronger relationship to recall accuracy. In addition, acute anxiety during study significantly impacted path length but not semantic clustering. These results suggest that semantic path length and semantic clustering scores provide distinct and complementary information about the role of semantic processing in episodic memory organization.

Topic Area: LONG-TERM MEMORY: Episodic