Generalization of associative item-memory EEG features to associative recognition
Yvonne Y Chen1,2, Jeremy B Caplan1; 1University of Alberta, 2Baylor College of Medicine
Item memory is known to be enhanced by associative processes. Accordingly, some electroencephalographic (EEG) features that are implicated in successful item memory have been thought to reflect associative processes. We test this interpretation by asking if these features explain memory success in the associative recognition task, for which item-memory alone cannot help with the memory judgment. Participants were shown lists of word-pairs (e.g., A-B, C-D,...) and later asked to distinguished intact (e.g., A-B) from recombined (e.g., C-B) probe pairs. At study, we found a significant difference between later-remembered and later-forgotten pairs (subsequent memory effect) at posterior Slow Wave (t(58)=2.34), thought to reflect elaborative processes. However, the difference observed in the posterior Slow Wave did not correlate with performance across individuals. At test, we found no significant difference between remembered and forgotten pairs (retrieval success effect) at the left parietal positivity, thought to reflect retrieval of associative information that supports recollection; neither did the difference measure correlate with performance across participants. In contrast, theta oscillations (4-8 Hz rhythmic activity) showed a significant subsequent memory effect (t(58)=2.23) and retrieval success effect (t(58)=5.64). Furthermore, the differences in theta activity were correlated with participants’ memory performance at both study (r(58)=0.35) and test (r(58)=0.34). Our results suggest that if ERPs reflect associative processes that influence item memory, they are quite different from ERPs that support item-item association memory. However, theta oscillations apparently reflect a cognitive state that drives memory success in both item and association memory tasks.
Topic Area: LONG-TERM MEMORY: Episodic