Uncertainty visualization in data journalism : a pilot study on election polls in a DER SPIEGEL online article

Abstract

Uncertainty visualization, such as in political polling, plays a key role in shaping public understanding and interpretation of data. While most research on uncertainty visualization has focused on controlled laboratory settings, we investigate how readers perceive and interpret these visuals in a real-world context. We included a short survey in a DER SPIEGEL pre-election article, gathering 98 voluntary responses. Results indicate that two-thirds of respondents found the visualization useful, but interpretations varied across levels of engagement—from basic recognition to inaccurate interpretations. These findings highlight both the potential and the pitfalls of uncertainty visualizations in journalistic contexts and will inform a follow-up study aimed at reducing misinterpretation and mistrust in probabilistic reporting.