Figure 1

Reward task. Cues representing probabilities (25, 50, 75, and 100 percent) and course of the delayed incentive task as used during training and scanning: Subjects first saw the cue and were instructed to expect a certain amount of money (magnitude) at the announced probability during the delay. To preserve themselves the chance to win, they had to react correctly to the target with a button press. Reward was displayed dependent on the previously announced probabilities. For Model 1, regressors were defined over the whole length of the trial, for Model 2, Cue (expectation phase) and Reward (outcome phase) were modeled separately.