FAIR2 Data Chats: Advancing Public-Interest Data Science Through Participatory Research

This study introduces FAIR2 Data Chats as a community-participatory research tool to advance data science for the public interest. The FAIR2 framework expands on the established FAIR data principles — Findable, Accessible, Interoperable, and Reusable — by introducing four additional principles tailored to social data: Frame, Articulate, Identify, and Report. These principles emphasize the importance of embedding historical and community knowledge throughout the data analytic process to enhance the understanding of social data and address discrimination bias. Drawing from foundations in causal inference, bias analysis, health disparities, and algorithmic fairness, FAIR2 Data Chats bring together community and research collaborators to enrich metadata and inform data analytics. This study provides principles and guidelines for the implementation of FAIR2 Data Chats and offers an example that highlights their relevance for research. As collaborators in FAIR2 Data Chats, community members with experiences represented in the data provided critical qualitative context to the data, enhancing its value, interpretation accuracy, and ethical use. Findings demonstrate the value of participatory methods in strengthening data science to advance the public interest.