"How do demographic factors and GPT familiarity affect contribution decisions in public goods games?"

Abstract

The public goods game has been used extensively to investigate cooperation in situations of social dilemmas. In a public goods game, individual members of a group determine how much of their individual endowment to contribute to a group account. Contributions are multiplied by a constant m (>1) and then distributed evenly among all group members, regardless of their individual contributions (Thielmann et al., 2020). In this paper, we test the influence of various demographic variables, such as gender, age, investment background, statistical background, and level of GPT familiarity, on individuals’ contribution decisions in the public goods game, building on the study by Bao and Pei (2023), where they identified cognitive uncertainty to be a significant factor in influencing contribution decisions. We design 3 linear regression models to investigate these effects and conduct OLS regression. Notably, we observe a nuanced relationship between GPT exposure and cognitive uncertainty. Individuals encountering GPT for the first time exhibit heightened cognitive uncertainty. Females usually exhibit higher cognitive uncertainty and greater contribution deviation compared to male participants. Individuals with a background in investments and statistics tend to exhibit lower cognitive uncertainty and tend to contribute more. Generally, older age tends to be associated with lower cognitive uncertainty and lower contribution deviation. Collectively, these insights confirm the hypothesised effects of GPT familiarity, demographic attributes, and game dynamics on economic decision-making.