Probabilistic Storytelling and Temporal Exigencies in Predictive Data Journalism

Abstract

The future of data-driven journalism has attracted widespread attention, but what about the future in data journalism? In other words: How do future predictions shape the formulation of knowledge claims in newsmaking that relies on the analysis of large troves of digital data? Based on interviews with professionals working on such projects, we study how they exploited predictive analytics to make evidentiary propositions and we interrogate the epistemological conceptions that underpin this future-oriented data journalism. Despite growing ambitions to generate more precise prognoses in a shorter amount of time, the practitioners downplayed the journalistic relevance of such projections. Instead, they stressed their dependence on past numeric information and the time-consuming effort needed to produce forward-looking stories that connect with the public. We argue that acknowledging the temporal exigencies around anticipatory news allowed those working on data journalistic projects to explore the possibilities of probabilistic storytelling while at the same time maintaining a professional paradigm of fact-based, post hoc reporting.