Cooperation and Cognition

This project tests a model of cooperation in which breaking an explicit cooperative promise (“defection”) carries a cognitive cost, and individuals anticipate and manage that cost when deciding how to act. Participants complete an online experiment (implemented in Gorilla and run on Prolific) that combines memory tasks with a one-shot trust game. The key manipulation is the size of a bonus for a second memory test offered later in the study ($1 vs $20). The central prediction is that a larger downstream bonus will reduce the rate of promise-breaking in the trust game. We will test our primary hypothesis with a preregistered logistic regression predicting Defection from Condition (low vs high bonus), supplemented by a two-proportion test and pre-specified robustness analyses that re-code Defection among promisors only and, descriptively, separate non-promisers from promise-keepers and defectors.