EPISTEMOLOGIES OF MISSING DATA: COVID DATA BUILDERS AND THE PRODUCTION AND MAINTENANCE OF MARGINALIZED COVID DATASETS

Abstract

During COVID-19, countless dashboards have served as central media where people learn critical information about the pandemic. Varied actors, including news organizations, government agencies, universities, and NGOs created and maintained these dashboards, conducting the onerous labor of collecting, categorizing, and taking care of COVID data. This study uncovers different forms of data practices and labor behind the building of these dashboards, based on in-depth interviews with volunteers and practitioners across India and the United States who have participated in COVID dashboard projects. Specifically, we are interested in projects that have focused on underrepresented or missing COVID data such as COVID cases in prisons and long-term care facilities, racial/ethnic breakdown of cases, as well as deaths due to COVID enforcement. These data builders employed sometimes creative, sometimes mundane and laborious data practices to not simply collect, but to produce these data that are often invisible in the official COVID dataset. In this process of data production, dashboard builders grappled with the questions of how certain data is collected, who/what is missing from the dataset, and how these data voids shape and manipulate our understanding of the pandemic. Interviewing 74 data builders who participated in COVID dashboard projects, this paper demonstrates the range of underrepresented and messy COVID data that these data builders have identified, fixed, and maintained to render them useful: disappearing data, lumped data, and absent data. Such critical engagement with messy COVID data reveals different data injustices that have tremendous potential to affect future pandemic preparation and management.