Fig. 5: Different option positions elicit different gaze sequences that in turn relate to different cooperation rates.

a Cooperation rate for the four most frequent gaze sequences in each position. The two most frequent first fixation and subsequent second fixation are represented (See Supplementary Fig. 7 for a more detailed description of the gaze sequences), and the cooperation rate for each participant when following these sequences is computed. The bars represent the average cooperation across participants and error bars represent the s.e.m. The percentage displayed on top of each bar is the share of all trials in the specific option position that follows the represented sequence. The letters A and B represent special gaze paths illustrated in (b). b Example of frequent gaze sequences in the different positions, where the first fixation (1) is identical, but the second gaze is different (2A and 2B), resulting in strong differences in cooperation rates (displayed in (a)). c Frequency of the second fixation being within-participant (First and second gaze to the same participant: lateral, vertical, or diagonal) or between participants (within-cell or diagonal). Each dot represents the frequency for one participant, the error bars represent mean and s.e.m. across participants. d Most frequent gaze patterns: on 26% of the trials, participants start looking up left at their own payoff, on 25%, they start looking up right at their own payoff. The next most frequent saccades are within-participant, lateral or towards the bottom row (see Supplementary Fig. 6 for more details). All panels: N = 88 participants.