Figure 3.

A, B, Effects of retrieval expectancy on retention of declarative (card locations) (A) and procedural (finger sequence tapping skills) (B) memories during sleep in experiment 2. Mean (± SEM) retention performance is indicated for subjects who expected (Exp, black bars) and did not expect retrieval testing (Unexp, empty bars), as well as for subjects who suspected the retrieval although they were never explicitly informed about it, i.e., these subjects were originally assigned to the unexpected group (Susp, gray bars); *p < 0.05, for differences between groups. Retention of card locations is indicated by the percentage of card locations recalled at retrieval testing, with performance on the criterion trial during learning set to 100%. Overnight gain in finger tapping skill is indicated by the average percentage of the number of correct sequences per 30 s trial on the three trials at retrieval testing, with the average performance at the end of learning (trial 10–12) set to 100%. C, Spindle count during SWS in the expected and unexpected groups. D, Correlation between spindle count during SWS and overnight retention of card locations in subjects who expected (filled circles) and who did not expect (empty circles) retrieval testing. Correlation coefficients are indicated; *p < 0.05, **p < 0.01, for differences between expected and unexpected groups.