Collective incentives reduce over-exploitation of social information in unconstrained human groups

This study investigates how individuals weigh personal and social information in collective decision-making, using a naturalistic immersive-reality experiment. Groups searched for rewards in environments with varying resource distributions and incentive structures. Results showed individual gains but group losses from high social information use and proximity in concentrated environments. Group-level incentives mitigated the negative consequences of excessive scrounging. A Social Hidden Markov Decision model revealed increased sensitivity to social information in concentrated environments, with group incentives reducing this responsiveness. Time-lagged regressions showed a collective exploration-exploitation trade-off.

Dominik Deffner, David Mezey, Benjamin Kahl, Alexander Schakowski, Pawel Romanczuk, Charley M. Wu, Ralf H. J. M. Kurvers