The Selective Consolidation of Declarative Memories
MetadataShow full metadata
During sleep, recently learned memories are stabilized through a process known as consolidation. This process does not uniformly preserve all recent experiences; some memories fade, whereas others become resistant to disruption and appear to last indefinitely. One possibility consistent with recent results is that memory processing during sleep may reflect an adaptive mechanism for evaluating recent experiences and relating them to future goals. The present study was designed to further test this possibility, by determining whether event-related potentials (ERPs) associated with goal- consistent memories change more across a night of sleep than ERPs associated with goal- irrelevant memories. Participants completed a memory task involving two sessions separated by a 24-hour delay. In the initial session, participants studied a series of images appearing in one of eight locations on a screen and their memory was immediately tested for the image locations. Half of the images were money-related (goal-relevant) while the other half were office supply (goal-irrelevant) images. After two study-test cycles were complete, each participant was informed that when they came back the next day they would complete one final memory test, and that if their memory for the locations associated with money images improved the following day, they would receive an extra five dollars. The instruction was meant to bias them by making the money-related images in line with their goal of increasing financial gains. ERP’s were recorded for three time components (FN400, LPC, LFE) during the final test on day 1 (baseline) and the test on day 2 (delayed). Results revealed no difference in memory change from the baseline to the delayed test between the money-related and office supply images. However, ERP results revealed that there was a greater increase in the late positive component (500 – 800 ms) over the posterior regions from the baseline to the delayed test. This result suggests that memory for locations associated with money-related images may have been preferentially consolidated compared with office supply images. Additionally, the FN400 showed increased negativity at the delayed test compared to the baseline test over frontal regions, suggesting that familiarity increased over time for both image types. Finally, the LFE showed an increased positivity in the central ROI consistent with behavioral results suggesting that money-related images required more effortful processing to accurately retrieve.