The impact of sleep on novel concept learning.

Prior research demonstrates that sleep benefits memory consolidation. But beyond its role in memory retention, sleep may also facilitate the reorganization and flexible use of new information. In the present study, we investigated the effect of sleep on conceptual knowledge. Participants classified abstract dot patterns into novel categories, and were later tested on both previously seen dot patterns as well as on new patterns. A Wake group (n=17) trained at 9AM, continued with their daily activities, and then tested at 9PM that evening. A Sleep group (n=20) trained at 9PM, went home to sleep, and was tested the following morning at 9AM. Two Immediate Test control groups completed testing immediately following training in either the morning (n=18) or evening (n=18). Post-training sleep led to superior classification of all stimulus types, including the specific exemplars learned during training, novel patterns that had not previously been seen, and "prototype" patterns from which the exemplars were derived. However, performance did not improve significantly above baseline after a night of sleep. Instead, sleep appeared to maintain performance, relative to a performance decline across a day of wakefulness. There was additionally evidence of a time of day effect on performance. Together with prior observations, these data support the notion that sleep may be involved in an important process whereby we extract commonalities from our experiences to construct useful mental models of the world around us.