Abstract
A total of 96 subjects, half women and half men, opposed a simulated player who was 50% cooperative in the first block of 10 trials and 100% cooperative in the third (last) block. In the second block the latter either was 50% cooperative (control condition) or, in the GRIT condition, communicated and carried out a conciliatory strategy that included some retaliation. The subjects made their choices on each trial either before, after, or simultaneous with the simulated player. It was hypothesized that, beginning in the second block, there would be greater cooperation in the GRIT than in the control condition in all three sequences, including the condition in which the subject chose second, making the simulated player's cooperation very vulnerable to exploitation. The results supported the hypothesis; and the difference persisted in the third block, where there was no longer any retaliation for exploitation. Subjects choosing second were less cooperative in the control condition than those in other sequences. It was concluded that the GRIT approach is effective even when the sequence of responding makes the initiator's acts especially vulnerable-the condition ordinarily holding true in real-life conflict.