The Neurocognitive Process of Group Decision in a Naturalistic Context

Poster Session D - Monday, April 15, 2024, 8:00 – 10:00 am EDT, Sheraton Hall ABC

Jiahui Dai¹ (202221061047@mail.bnu.edu.cn), Zhaonan Meng¹, Xiangyu He¹, Chunming Lu¹; ¹Beijing Normal University

Group decision-making is crucial in human life, but the underlying cognitive process and neural bases are unclear. Here, we addressed this issue by inventing a large-group naturalistic decision task. In experiment (Exp) 1, groups of 16-20 individuals were asked to freely discuss a difficult problem which required them to choose 5 out of 10 options and report their personal selections every 4 minutes. The results indicated a significant increase of selection convergence over time. Employing the Hidden Markov models (HMM), we found that a two-state model best fit their behaviors. One state was more associated with personal preference and the other one with social influence. Moreover, the latter predicted final selection better than the former at early phases. In Exp 2, we tested which architecture of the group organization was optimal for group decision. The results showed that groups with one leader performed better than other organizations (e.g., no leader, four equal leaders, four leaders with different status). In Exp 3, the hemodynamic concentration changes of each individual were simultaneously collected using functional Near-Infrared Spectroscopy (fNIRS) hyperscanning and subject-to-group neural synchronization (GNS) were calculated. Through representational similarity analysis, we observed a close association between GNS in the temporoparietal junction (TPJ) and the social influence state. No GNS was found for personal preference. Together, these results suggested that the TPJ-associated social influence played a pivotal role in group decision, and groups with a single leader outperformed other types of group organization during decision.

Topic Area: THINKING: Decision making