Forgiveness in a Repeated Prisoner’s Dilemma for Dhildren

• In Event: Expanding the Study of Early Forgiveness: Intentions, Individual Differences, and Culture

Abstract

Forgiveness researchers have noted the potential for using the Repeated Prisoner’s Dilemma (RPD) task as a tool for studying the process of forgiveness over time (Exline et al., 2004; McCullough, Kurzban & Tabak, 2013). The RPD is a two-player game in which simultaneous decisions can result in cooperation or transgressions over multiple rounds. It has typically been used to study cooperation and test game theoretic predictions in adults, but has only rarely been used with children due to the difficulty of presenting the task in a simple interface. We designed a child-friendly RPD with programmed partners in order to study forgiveness in real-time behavior. The partners were designed to transgress (defected) at specific points in the game. We examined children’s responses to those transgressions (cooperate or defect) over subsequent rounds, measuring forgiveness as a return to cooperation.

Two samples of children were tested: 9-11-year-olds (N=153) and 7-13-year-olds (N=108 twins, ongoing). Parent-report measures of children’s aggression were used as our main predictor: lower aggression should predict more forgiveness. For the first sample, results showed a clear effect of reactive aggression on children’s decisions after the transgression. Children overall began the game with about a 50% probability of cooperating. They continued to cooperate until the partner defected and then most children defected in response. However, children with low reactive aggression scores quickly returned to cooperation whereas children with higher scores continued to defect (coeff=-0.711, p=0.004). A similar effect occurred when the partner defected a second time. By contrast, children with higher reactive aggression scores did not cooperate even after multiple cooperative moves by the partner.

For the second sample, data collection is ongoing, but a preliminary examination of the data for one randomly selected twin from each pair confirms the general pattern of defection in response to partner defection, with a return to cooperation for about half of the subset. Both phenotypic and twin analyses of the link between aggression and forgiveness will be presented.

These results suggest that pre-adolescent children have a typical, defensive response after a partner injures the relationship, but that forgiveness quickly follows and cooperation resumes. A more atypical response occurs for children with more impulsive, aggressive traits. Reactive aggression may lead to “holding a grudge” for longer which can damage potentially healthy relationships. This behavior resembles a game theoretic strategy called GRIM, but the psychological processes which produce this failure to forgive remain unclear. Future work can manipulate the partner’s responses in the RPD to determine which cues are most effect for encouraging forgiveness or what cognitive interventions can prevent a shift to grudges.

Authors

• Meia Chita-Tegmark, Tufts University
  Presenting Author

• Megan Flom, Boston University
  Non-Presenting Author

• Kimberly J Saudino, Boston University
  Non-Presenting Author

• Peter R. Blake, Boston University
  Non-Presenting Author