Building an Apparatus: Refractive, Reflective, and Diffractive Readings of Trace Data

We propose a set of methodological principles and strategies for the use of trace data, i.e., data capturing performances carried out on or via information systems, often at a fine level of detail. Trace data comes with a number of methodological and theoretical challenges associated with the inseparable nature of the social and material. Drawing on Haraway and Barad’s distinctions among refraction, reflection, and diffraction, we compare three approaches to trace data analysis. We argue that a diffractive methodology allows us to explore how trace data are not given but created through the construction of a research apparatus to study trace data. By focusing on the diffractive ways in which traces ripple through an apparatus, it is possible to explore some of the taken-for-granted, invisible dynamics of sociomateriality. Equally important, this approach allows us to describe what distinctions emerge and when, within entwined phenomena in the research process. Empirically, we illustrate the guiding methodological principles and strategies by analyzing trace data from Gravity Spy, a crowdsourced citizen science project on Zooniverse.org. We conclude by suggesting that a diffractive methodology helps us draw together quantitative and qualitative research practices in new and productive ways that allow us to study and design for the entwined and dynamic sociomaterial practices found in contemporary organizations.