Figure 1.

Task design, receiving social feedback from peers after a real-life interaction. A, Participants came to the laboratory in groups of five on 2 consecutive days. On the first day they got to know each other by playing the board game Monopoly for 1 h 15 min. Afterward, each person rated three of the other players on 40 positive and 40 negative trait adjectives on a Likert scale from 1 (this trait does not apply to the person at all) to 8 (this trait applies to the person very much). On the first day participants did not rate themselves (yellow) and did not rate one of the other players (green). See Table 1 for a list of the trait adjectives. B, On the second day participants performed the following task in the fMRI scanner. They first saw a cue indicating whether the following trial was about themselves or about the other person they had not rated on the previous day. They then saw one of 40 positive or 40 negative trait adjectives and had to imagine how much the trait applied to themselves or to the other person. They first gave their own rating and then saw the feedback in the form of the mean rating they believed three other participants had given on the previous day. The absolute difference between participants' own ratings and the feedback ratings they received was conceptualized as feedback discrepancies and manipulated during the experiment. Outside the scanner participants rated themselves and the other player a second time so that we could assess how much they updated their ratings after receiving feedback. C, For the main behavioral analyses we used a 2 by 2 design with the factors feedback target (self/other) and feedback desirability (desirable/undesirable). Desirable feedback was defined as feedback ratings that were higher than participants' own first ratings (e.g., own first rating for polite was 6 and feedback rating was 8.0). Conversely, undesirable feedback was defined as feedback ratings lower than participants' first ratings (e.g., own first rating for polite was 6 and feedback rating was 4.0). All ratings for negative trait adjectives were reverse coded. Thus, feedback desirability was independent of the valence of the trait adjective.