Figure 2.

Relationship between retrieval expectancy, slow oscillation power, and retention of declarative memory. A, C, Mean (± SEM) slow oscillation power within the first 120 min of non-REM sleep for six successive 20 min intervals (0–20, 20–40 min, etc.) for subjects who expected (solid lines) and who did not expect (dotted lines) retrieval testing in experiment 1 (A) and experiment 2 (C). B, D, Correlation between slow oscillation power (during the first 20 min of non-REM sleep) and retention of word pairs (experiment 1) (B) and card locations (experiment 2) (D) separately for the subjects who expected (filled circles) and the subjects who did not expect retrieval testing (empty circles); *p < 0.05, **p < 0.01, ***p < 0.001, for differences between groups for single 20 min intervals; p < 0.05, for comparisons across the whole 120 min interval. Correlation coefficients are indicated. Retention performance is indicated by the percentage of recalled items (word pairs, card locations) at retrieval, with performance on the criterion trial during learning set to 100%.