Figure 3. Results for oscillations in the theta and spindle range in both sleep groups (N=20).

Oscillatory activity during vocabulary replay was specifically analysed with regards to words for which cueing during sleep resulted in a behavioural change in memory performance. Words, which were not remembered before sleep but correctly retrieved after sleep, were labelled as ‘Gains'. In contrast, ‘Losses' refers to words, which were successfully retrieved before sleep but not after the sleep interval. Words were indexed as ‘hithit' in case they were remembered before and after the sleep interval, while words neither retrieved before nor after sleep entered the category ‘missmiss'. Successful cueing of Dutch words was associated with enhanced power in the theta (a) and spindle (c) band. (b) Representative electrode F3. Verbal cues were presented at time 0 ms. The rectangle illustrates the time window used for the repeated measures ANOVA and post hoc comparisons in the bar chart. Top and bottom panels indicate significant differences in post hoc t-tests (in black) between ‘gains' and ‘losses' in spindle and theta power, respectively. (d–f) The differences in theta (d) and spindle band (f) vanished when cues were immediately followed by auditory feedback during sleep (that is, the false or correct German translation). Values are mean±s.e.m. ^**P≤0.01.