How ex-ante information design affects cognitive conflict and cooperation depending on agents’ tendency to cooperate: a mouse tracking study

This study investigates how the design of ex-ante information (given before decisions are taken) affects the cooperative decisions made by team members, depending on their inclination to cooperate. Also analyzed is the effect of this information on cognitive conflict (when an agent internally experiences contrary demands or opposing forces). Moreover, the relationship between cognitive conflict and cooperation is explored. We design an experiment in which participants play 15 rounds, in pairs, of three social dilemmas. The ex-ante information is manipulated in three different ways: displaying only private and individual earnings (the “I” frame); displaying the joint profits (the “We” frame); and displaying both types of information (the “I&We” frame). Mouse movements are tracked using a specific software. Individual inclinations to cooperate are measured using the Honesty-Humility (HH) dimension of the HEXACO personality model. The agents are classified as HH_highs (high tendency to cooperate) and HH_lows (low tendency to cooperate). We measure the cooperation level as the percentage of cooperative decisions and the cognitive conflict level based on the curvature of mouse movements. Ex-ante information design is not found to affect cooperation levels in the case of HH_highs but does affect cognitive conflict levels. The opposite is observed for HH_lows. The main result is therefore that the cooperation of non-cooperative agents can be increased through framing (“I&We” being the best framing). No effect on cognitive conflict is found for HH_lows. Finally, a relationship between cognitive conflict and cooperative decisions for HH_highs is only observed in the case of the “I&We” frame.