Figure 1: Instrumental learning task design.

Subjects were required to pick one of four options on each trial. 3 seconds after selection, the option yielded an outcome that could comprise reward (token for €1) and/or punishment (token for -€1). The figure shows an example of combined reward and punishment outcome. The probabilities of reward and punishment were fully independent from each other, independent between options, and varied slowly over trials as indicated in the exemplar graph for one option. Following the outcome there was an inter-trial interval of 3 secs.