Year: 2025
Type: preprint
Abstract: Like other fields, data journalism faces the fundamental challenge of conveying uncertainty to the audience. However, unlike scientific communication, data journalism must engage a broad audience with highly variable levels of prior topic-related, statistical, and methodological knowledge. To structure examples and develop better strategies for visualizing uncertainty it is useful to distinguish different types of uncertainty that commonly arise in journalistic contexts. For this purpose we propose a framework that identifies seven key scenarios: Spatial, Sampling, Forecast, Classification, Definite Ranges, Missing Data, and Reconstructed Past. Each scenario highlights a distinct situation where uncertainty plays a crucial role—whether in mapping environmental risks, interpreting election polls, projecting future outcomes, or reconstructing data about historical events. By systematically examining these uncertainty scenarios, this work seeks to better support data journalists in dealing with uncertainty, encourage transparent reporting and promote critical engagement with data-driven narratives.
Authors: Benjamin Possin, Christoph Kinkeldey
Institution: HAW Hamburg
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4413362039