Data sprints and coding retreats

by Anders Kristian Munk

One of the ways in which we work with participatory data design is in so called data sprints (Munk et al. 2019a). A data sprint is a one-week occasion to prototype a digital methods project together with those who have a stake in it, a.k.a. the issue experts. This could be actors in a controversy that the project is attempting to map, the intended users of the results, or groups for whom the results could have adverse consequences. The week begins with the issue experts presenting their matters of concern followed by a Q&A session. In the afternoon of that first day, initial ideas are transformed into protocols for digital methods projects that could be feasibly carried out in a week. There is always this intensity that results have to materialize by the end of the week and so participants tend to engage. Within the first few days we begin to build tentative and exploratory visualizations capable of eliciting feedback from the issue experts. The process makes it clear what can and what cannot be done, of course, but more importantly, it tends to prompt reflection and what we actually came to do. Like in an ethnographic interview, in a sprint, participants tend to discover questions rather than answers (Munk et al. 2019b). Spurred by these reflections, projects are recalibrated or completely redesigned during the course of the week. 

Preliminary visualization used to elicit feedback from issue experts during a data sprint with the Royal Danish Theater and their analytics department.

Another way in which we have engaged with critical technical practice is by inviting tool developers to join us for a coding retreat at the lab. These are also workshop-style events that unfold over several days, but there are no issue experts. Rather, the object in focus is a digital methods tool in need of development. Our visitors would have to take on this task anyway but typically appreciate the opportunity to have some days where they can focus on it without distractions. By providing that occasion we get an opportunity to sit in on their discussions and, if relevant, have a say on key choices that will affect our use of the tool. In several cases we have organized these retreats around specific issues where our concerns as digital methods researchers converged with those of the tool developers, for example on how to visualize ambiguity or define a web entity in a crawl. Taken together, coding retreats and data sprints provide some of the occasions for spending time in critical proximity with the technical, that is necessary for the lab to be lab-like and for us to get our fingers properly into the digital minced meat.

Sketches, coding and mock-up visualization at a coding retreat on tools for visualizing ambiguity with Density Design from Milan.




Munk, A. K., Meunier, A., & Venturini, T. (2019a). Data sprints: A collaborative format in digital controversy mapping. digital-STS: A Field Guide for Science & Technology Studies, 472.

Munk, A. K., Madsen, A. K., & Jacomy, M. (2019b). Thinking through the Databody. Designs for experimentation and inquiry: Approaching learning and knowing in digital transformation, 110.