Figure 1. Experimental design and behaviour results.

(a) The participants engaged in a dictator game involving a money-allocation task inside a magnetic resonance imaging (MRI) scanner. The participants played the role of the dictator. Three others played the role of receivers who each had a different moral trait, good, neutral, or bad. In each trial, the face of one of the three others was presented for 3 seconds during the face phase, after a 3–7 seconds task-cue phase. In the choice phase, a pair of choices was randomly presented. The participants allocated money to themselves and others by making a binary choice within 5 seconds. (b) In the face-choice task as the control task, two photographs were presented in the choice phase. The participants chose the photograph that was presented in the face phase within 5 seconds. (c) Each dot indicates an option for money allocation that was used in the choice phase of the game. The horizontal axis indicates the money obtained by the participants, and the vertical axis indicates the money that was allocated to others. Nine options of money allocation were used. (d) The group averages and standard errors of the money allocated to others by the participants. The money allocated to others was evaluated using a trial average of the money allocated to others. When the participants chose only options that allocated more money to others, the money to others was “max”. Conversely, when the participants chose only options that allocated less money to others, the money to others was “min”. G, N, and B indicate good, neutral, and bad others, respectively.