Figure 1. Task design. A, Design of the trust game. First, player A chooses In or Out, which reveals a belief probability of the likeliness that player B will choose Cooperate. If player A chooses Out (i.e., does not trust player B), player A and B receive zA
and zB
, respectively. If player A chooses In (i.e., trusts player B), then with the knowledge of player A’s belief probability, player B decides whether to Cooperate or Defect. If player B chooses Defect, players A and B receive yA
and yB
, respectively; if Cooperate, players A and B receive xA
and xB
, respectively. The actual assignment of
x,
y,
z and τA
for the 45 trials is shown in Extended Data Figure 1-1. B, An outline and example of experimental trials. After the green fixation period (2–5 s; cue phase), a task condition is presented for 5 s (choice phase), and participants are asked to press the Cooperate or Defect button (blue and red, respectively). Then, a yellow fixation cross is shown for 6–15 s (rest phase). C, An illustration of the complete experimental paradigm. For both the fMRI and online studies, in the first experiment, participants (as player A) chose In or Out and reveal their belief probability that player B would choose Cooperate. In the second experiment, participants (as player B) chose to Cooperate or Defect. Participants make their decisions while being scanned in the fMRI experiment. Instructions for the first and second experiments are shown in Extended Data Figure 1-2.