Fig. 2.

Schapiro et al. (29) category learning paradigm and results. (A) Satellite exemplars studied from three categories. (B) Performance change for unique and shared features from presleep to postsleep tests. *P < 0.05. (Results are collapsed across the frequency manipulation included in that study.) (C) Experimental protocol: Participants studied the satellites in the evening, had an immediate test, and were tested again 12 h later, after a night of sleep.