Figure 1. Experimental procedure.

(a,b) After studying 120 Dutch–German word pairs in the evening, participants slept for 3 h. During the retention interval 80 Dutch words (40 cued, 40 cued+feedback) were repeatedly presented in sleep groups I and II. The type of feedback for the ‘cued+feedback' words differed between the two experimental groups. While the correct-feedback group (N=14) received the correct German translation after the Dutch word during sleep, the false-feedback group (N=13) received an incorrect German translation. Participants of the control group (N=16) received 40 Dutch cues, which were followed by late correct feedback (1,500 ms instead of 200 ms interstimulus interval) and 40 Dutch cues that were immediately followed by a pure sinus tone. In the cueing categories, the proportion of remembered and forgotten word pairs of the last learning trial was maintained. All words were exclusively replayed during NREM sleep. A cued recall procedure was applied after sleep testing the participant's memory for the German translations.