Studying data communities: Analytical dimensions from and for empirical research

How can data communities be studied? In this paper, we combine existing literature with our own research spanning more than a decade to propose five dimensions and a series of guiding questions which can be used to analyze data communities. These dimensions – disciplinary domains, social orderings, material orderings, data uses, and data expertise – illustrate the variability of data communities and offer starting points for future empirical research, infrastructure design, and policy activities. Analyzing data communities through the multidimensional perspectives which we propose encourages critical interrogation of these heterogeneous groups, while highlighting the complex ways in which data practices constitute particular communities and how those communities simultaneously shape data.