Learning to Make Collective Decisions: The Impact of Confidence Escalation

Little is known about how people learn to take into account others’ opinions in joint decisions. To address this question, we combined computational and empirical approaches. Human dyads made individual and joint visual perceptual decision and rated their confidence in those decisions (data previously published). We trained a reinforcement (temporal difference) learning agent to get the participants’ confidence level and learn to arrive at a dyadic decision by finding the policy that either maximized the accuracy of the model decisions or maximally conformed to the empirical dyadic decisions. When confidences were shared visually without verbal interaction, RL agents successfully captured social learning. When participants exchanged confidences visually and interacted verbally, no collective benefit was achieved and the model failed to predict the dyadic behaviour. Behaviourally, dyad members’ confidence increased progressively and verbal interaction accelerated this escalation. The success of the model in drawing collective benefit from dyad members was inversely related to confidence escalation rate. The findings show an automated learning agent can, in principle, combine individual opinions and achieve collective benefit but the same agent cannot discount the escalation suggesting that one cognitive component of collective decision making in human may involve discounting of overconfidence arising from interactions.

[1]  C. Frith,et al.  Coming to Terms , 2012, Psychological science.

[2]  J. Banks,et al.  Information Aggregation, Rationality, and the Condorcet Jury Theorem , 1996, American Political Science Review.

[3]  Richard S. Sutton,et al.  Reinforcement Learning: An Introduction , 1998, IEEE Trans. Neural Networks.

[4]  H. Gintis,et al.  A Cooperative Species: Human Reciprocity and Its Evolution , 2011 .

[5]  R. Zajonc SOCIAL FACILITATION. , 1965, Science.

[6]  Csaba Szepesvári,et al.  Algorithms for Reinforcement Learning , 2010, Synthesis Lectures on Artificial Intelligence and Machine Learning.

[7]  S. Asch Effects of Group Pressure Upon the Modification and Distortion of Judgments , 1951 .

[8]  Scott L. Feld,et al.  Collective Decision Making: An Economic Outlook , 1986, American Political Science Review.

[9]  D. Wolpert,et al.  Two Eyes for an Eye: The Neuroscience of Force Escalation , 2003, Science.

[10]  K. A. Ericsson,et al.  Verbal reports as data. , 1980 .

[11]  P. Neudecker,et al.  When Are Two Heads Better than One and Why? , 2022 .

[12]  Majid Nili Ahmadabadi,et al.  Learning Active Fusion of Multiple Experts' Decisions: An Attention-Based Approach , 2011, Neural Computation.

[13]  Stanislas Dehaene,et al.  Limits on Introspection Distorted Subjective Time During the Dual-Task Bottleneck , 2008 .

[14]  P. Latham,et al.  Optimally Interacting Minds , 2022 .

[15]  Reid Hastie,et al.  The robust beauty of majority rules in group decisions. , 2005, Psychological review.

[16]  C. Frith,et al.  What failure in collective decision-making tells us about metacognition , 2012, Philosophical Transactions of the Royal Society B: Biological Sciences.

[17]  Mark W Woolrich,et al.  Associative learning of social value , 2008, Nature.

[18]  R D Sorkin,et al.  Signal-detection analysis of group decision making. , 2001, Psychological review.

[19]  Richard S. Sutton,et al.  Reinforcement Learning , 1992, Handbook of Machine Learning.

[20]  Ilan Yaniv,et al.  The Benefit of Additional Opinions , 2004 .

[21]  Martin H. Levinson Not by Genes Alone: How Culture Transformed Human Evolution , 2006 .

[22]  Geraint Rees,et al.  Collective Enumeration , 2012, Journal of experimental psychology. Human perception and performance.

[23]  Hauke R. Heekeren,et al.  The Neural Basis of Following Advice , 2011, PLoS biology.

[24]  Peter Bossaerts,et al.  Neural correlates of mentalizing-related computations during strategic interactions in humans , 2008, Proceedings of the National Academy of Sciences.

[25]  Richard Gonzalez,et al.  Interaction with Others Increases Decision Confidence but Not Decision Quality: Evidence against Information Collection Views of Interactive Decision Making , 1995 .