Fig. 2.

Study design. 120 participants were randomly assigned to one of 8 conditions (groups). All sleep groups performed the encoding, training and pre-delay retrieval in the evening and the post-delay retrieval 12 h later in the morning. The wake groups performed the encoding, training and pre-delay retrieval in the morning and the post-delay retrieval 12 h later in the evening. Besides the between-subjects factor Delay (sleep vs wake), pre-sleep memory strength was manipulated via the training threshold (A and C for weaker memories and B and D for stronger memories). Additionally, half of the participants were given an interference task before the post-delay retrieval to increase retrieval difficulty (between-subjects factor Retrieval Difficulty, A and B for no interference, C and D for interference). E = encoding, T = training, pvt = Psychomotor Vigilance Task, R1 = pre-delay retrieval, R2 = post-delay retrieval, i = interference.