No Evidence for a Wakeful Rest Benefit on Associative Memory: A Within-Participant EEG Study

Abstract

Wakeful rest after learning has been proposed to facilitate memory consolidation compared to engaging in a distraction task, with prior EEG studies linking slow oscillation power during rest to better memory performance. However, replication attempts have yielded mixed results. We investigated whether 10 minutes of wakeful rest would enhance associative memory performance relative to 10 minutes of a hippocampus-dependent auditory short-term memory distraction task, using a within-participant design with continuous EEG recording. We employed both a replication-inspired analytical approach, closely modeled on prior work, and a data-specific approach adapted to our dataset. Contrary to our hypotheses, we found no advantage of rest over distraction on associative memory performance. We did, however, observe an order effect: performance was better for the second learning than the first, and this improvement was more pronounced when rest was performed second compared to first. At the neurophysiological level, neither slow oscillation nor alpha power during the post-learning period correlated with memory performance, regardless of analytical pipeline, although cross-over analyses revealed that the choice of EEG reference influenced the direction of some correlations. At the phenomenological level, self-reported mental activity during rest and distraction, as well as trait daydreaming frequency, were not related to memory outcomes, despite the two conditions inducing distinct subjective cognitive states. Together, these findings do not support a robust benefit of post-learning wakeful rest over a hippocampus-dependent distraction task for associative memory, nor do they replicate prior EEG correlates of consolidation. We discuss methodological factors, including task-learning effects in within-participant designs, the coarseness of averaged spectral power measures, and variability in EEG preprocessing pipelines, that may contribute to inconsistencies across the literature, and we call for greater standardization and transparency in future studies.