Participatory Data Design, or PDD, is an approach to digital methods that commits to working with the problems and challenges of external partners. Much of our work in the lab engages with knowledge-production outside the walls of the University. We are interested in how, when and why digital methods can support new ways of thinking and making decisions. We develop this interest in close collaboration with partners in the public and private sector, or with various kinds of civil society groups and organizations. Fourty years of research in Science and Technology Studies have shown that the collection and analysis of data matters politically to people with a stake in a problem. PDD is about enabling upstream engagement in the data process.
“The core strategy (...) is to collaborate with local participants about the datafication of specific problems and organizational challenges” (Jensen et al, 2016)
PDD requires digital methods researchers to engage external collaborators in choices about data collection, analysis, visualization and interpretation. The situation and perspective of these collaborators must therefore be accomodated in research questions and designs. It also requires those who participate to become collectively attentive to the conditions and limitations of a digital methods project. To facilitate this process of mutual learning we organise so-called data sprints, where we road test our knowledge practices with the native experts of the fields in which we intervene.
"Data sprints (...) are extended research collectives that assemble over several days to collaboratively explore and visualize a set of pertinent questions. They comprise the necessary competencies to a) pose these questions; b) consider their relevance and implications for the controversy; c) operationalize them into feasible digital methods projects; d) procure and prepare the necessary datasets; e) write and adapt the necessary code; f) design and make sense of the relevant data visualizations; and g) elicit feedback and commentary through consecutive versions of these visualizations." (Munk et al. 2018)