Predicting Trust Dynamics and Transfer of Learning in Games of Strategic Interaction as a Function of a Player's Strategy and Level of Trustworthiness

摘要

Individuals playing a sequence of different games have shown to learn about the other player's behavior during their initial interaction and apply this knowledge when playing another game with the same individual in the future. Here we use a published computational cognitive model to generate predictions for an upcoming human study. The model plays both Prisoner's Dilemma and Chicken Game with a confederate agent who uses one of two predetermined strategies and whose level of trustworthiness is manipulated. We go beyond the standard postdictive practice and adopt the increasingly popular practice of using the model to make a priori predictions before the human data will be collected in an upcoming study.