Decision-Making in a Social Multi-Armed Bandit Task: Behavior, Electrophysiology and Pupillometry

TL;DR

This study introduces a social multi-armed bandit task to explore how humans incorporate observed actions and outcomes of others into their decision-making, revealing that social context influences learning strategies and neural responses.

Contribution

The paper presents a novel experimental paradigm for studying social influence in decision-making and provides insights into neural and behavioral differences between solo and dyadic conditions.

Findings

Participants integrate observed and personal information similarly, leading to suboptimal decisions.

Event-related potentials show attenuated amplitudes in solo versus dyadic conditions, indicating differences in arousal and attention.

Social context affects neural responses to rewards, influencing decision strategies.

Abstract

Understanding, predicting, and learning from other people's actions are fundamental human social-cognitive skills. Little is known about how and when we consider other's actions and outcomes when making our own decisions. We developed a novel task to study social influence in decision-making: the social multi-armed bandit task. This task assesses how people learn policies for optimal choices based on their own outcomes and another player's (observed) outcomes. The majority of participants integrated information gained through observation of their partner similarly as information gained through their own actions. This lead to a suboptimal decision-making strategy. Interestingly, event-related potentials time-locked to stimulus onset qualitatively similar but the amplitudes are attenuated in the solo compared to the dyadic version. This might indicate that arousal and attention after receiving a reward are sustained when a second agent is present but not when playing alone.