Adversarial Cooperation in Group Decision Making

Brown University, Providence RI

Investigators

Abstract

The national interest depends on high-quality decision-making in both the public and private sector. Most important decisions are the result of group deliberation. Therefore, a theory of effective discourse is proposed called adversarial cooperation. The proposed process divides the cognitive labor of a group into two distinct phases: conjecture and refutation. During the conjecture phase, the group leverages its members’ unique knowledge and skills to generate hypothesized solutions to a problem. In the refutation phase, the group stress-tests each solution, with group members acting as either supporters of their own proposed solution or as antagonists of other solutions. The group achieves consensus for the best decision on the assumption that reasoners are better at evaluating hypotheses that they disagree with than their own. So as long as the conjecture phase generates enough distinct potential solutions, the refutation phase is likely to converge on the best solution. The proposed research program has four strands of behavioral experiments. In all cases, small groups of partisans are asked to collectively reason about politicized problems. The prediction is that the groups whose members are more adversarial that decide by deliberation will outperform 1) more ideologically heterogeneous groups that deliberate and 2) groups that decide via some other collective action mechanism, such as judgment aggregation. The first experimental strand examines reasoning that requires creative idea generation (i.e. conjecturing). The second and the third strands examine cases that require rigorous stress-testing of existing ideas (i.e. refutations). The second strand focuses on evidence-based refutations and the third strand on argument-based refutations. The final strand applies the proposal to judgment at a larger scale. Are larger groups that include people with adversarial perspectives able to make more accurate judgments and predictions than group lacking adversarial perspectives? This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.