Situated between methodological traditions that range from thick ethnography and in-depth qualitative fieldwork to big data analysis and computational social science, we have committed ourselves to staying with the trouble rather than choosing sides. Without a deep, situated appreciation of the media and platform logics that premise digital data, larger scale data-driven approaches loose richness and potential. Without a good sense of the organizational circumstances of a data project, the possible avenues of intervention are limited and misguided.

In our methods development efforts we draw as much inspiration from the experienced anthropologist as we do from the computer programmer. We think of the world of digital traces as a field to be explored rather than a data repository to be modelled. We think of algorithms as pragmatic tools that can open for qualitative insights into meaning production rather than a-theoretical tools that detect patterns in a more neutral way that a human being. As a consequence, we are also sceptical of making hard distinctions between quantitative and qualitative analysis. Our methods development seek to overcome this chasm by enabling new modes of connecting qualitative and quantitative data, and new forms of data-scape navigation that enable the analyst to zoom between aggregated overviews of data and in-depth analysis of the underlying data.

We take as a point of departure that workable knowledge always brings together an assemblage of stakeholders, contextual conditions, overall research question, specific tools and data opportunities. Our interest in specific types of data or tools is therefore inspired and driven by reflections on the kinds of knowledge compositions into which they may be incorporated.