6/9(火) 15:00-16:00　Dr.Andrea Gradassiのトークセミナー「The effects of social network structure and individual differences on cooperation and social norms in Collective Risk Social Dilemmas」を開催します

2026.06.01

Place

Wako CBS East 1F Seminar Room / 脳科学東研究棟1 階 セミナー室

Tittle

The effects of social network structure and individual differences on cooperation and social norms in Collective Risk Social Dilemmas

Abstract

Understanding what makes humans cooperate in situations of risk is essential for addressing the most pressing global challenges, such as the increased risk of extreme climate events. These scenarios have been modelled using Collective Risk Social Dilemmas (CRSDs), where groups must collectively contribute to meet a threshold that prevents the loss of resources when a catastrophe hits. While factors such as catastrophe probability and threshold uncertainty influence group success, the mechanisms through which individuals become cooperators remain poorly understood. One limitation of previous studies is that participants played in homogeneous groups with full access to others' decisions. In real-life social networks, people occupy a range of positions, affecting the information they can access and the influence they can have on others. This structural variability can interact with individual differences in preferences for cooperation: a central position may be occupied by a cooperator or a free-rider. This study systematically manipulates network structure and participants' positions in a social network to investigate how they jointly drive cooperation in a CRSD. We employ a two-phase design. First, participants complete a battery of questionnaires measuring their Social Value Orientation (SVO), risk preferences, and strategic behaviour in a CRSD. Based on these data, we classify players' characteristics and assign them to groups in either a hierarchical or circular structure. Within these groups, participants play a longitudinal CRSD. In each round, participants observe the contributions of their neighbours and decide how much to contribute. Failure to meet the threshold by the final round results in losing most of their endowment. We vary both network topology (centralized vs. decentralized) and the position of cooperators (central vs. peripheral). We test two competing hypotheses: (a) centrally positioned cooperators may promote cooperation by signalling to others the existence of a cooperative social norm, or (b) they may incentivise others to free-ride as they reduce the need for everyone to contribute their fair share. Furthermore, we hypothesize that decentralized networks should prove more resilient to free-riders due to their having a diluted influence, but also slower to reach the threshold over the rounds of the game. By identifying which combinations of individual profiles and network positions facilitate collective action, our study contributes to a more mechanistic understanding of how collective action succeeds and cooperative norms emerge under collective risk, and provides insights for designing interventions to promote collective action in real-world crises.

Speaker

Dr Andrea Gradassi
Laboratory for Agent-Based Social Simulation, ISTC-CNR, Rome.
https://marwakavelaars.weebly.com/