Publication
Learned Futility: How Social Learning Can Lead to the Diffusion of Ineffective Strategies
Veronika Kurchyna; Lilian Kojan; Jan Schneider; Bernd Wurpts; Anastasia Golovin; André Calero Valdez; Jan Ole Berndt; Ingo Timm
In: Adela Coman; Simona Vasilache (Hrsg.). Social Computing and Social Media: 16th International Conference, SCSM 2024, Held as Part of the 26th HCI International Conference, HCII 2024, Washington, DC, USA, June 29 - July 4, 2024, Proceedings, Part III. International Conference on Social Computing and Social Media (SCSM-2024), June 29 - July 4, Washington DC, District of Columbia, USA, Pages 120-139, Lecture Notes in Computer Science (LNCS), Vol. 14705, ISBN 978-3-031-61311-1, Springer, 2024.
Abstract
When facing a crisis, such as a novel type of disease, individuals need to learn about effective health measures and practices to prevent the spread of illness. They do so both through reflection about their own actions as well as the communal experience of their peers. Here, we present an agent-based model to examine the resulting dynamics in the diffusion of health behaviours and practices. In the model, we employ reinforcement learning and bounded confidence opinion dynamics to model varying degrees of external, e.g. social, and internal knowledge gains in the context of protective measures against a novel disease as use case. Our study shows that social influence is critical for the adoption of potentially effective low-cost strategies, while individual learning modes limit the spread of potentially harmful high-cost strategies. On the downside, social learning also facilitates the spread of ineffective or even harmful health measures and practices. Our findings suggest that cultural variation emerges in times of crisis among learning individuals.