Election predictions in the news: how users perceive and respond to visual election forecasts

ABSTRACT

Political journalism often tries to predict the future, especially the outcomes of elections. This has historically been accomplished through written articles or opinion pieces. A more recent development involves the publication of data-driven predictions in online news media. These news items not only contain an estimate for the election results but also often try to visualize potential uncertainties of the prediction. However, the ways in which users react to these forms of journalism have not yet been studied extensively. In this work, we survey users of a predictive journalism piece on the 2021 German federal elections published by the German newspaper Süddeutsche Zeitung to better understand their reactions. While we found an alignment between the designers' intention to show the inherent uncertainty of election predictions with the audience's reception, we encountered mixed results in users' ability to interpret the uncertainty visualizations presented. Most respondents indicated that the predictions did not influence their thinking about the race, and some remained skeptical toward such predictions published by journalists for various reasons. Based on these findings, we suggest the need for rigorous user testing of visualizations for election prediction and increased awareness and future research on ways to increase the transparency of methods and data to develop appropriate trust toward predictive journalism.

KEYWORDS:

• Journalism
• election predictions
• visualization
• uncertainty

Acknowledgments

The authors would like to thank Süddeutsche Zeitung for supplying the survey data for scientific purposes.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Benedict Witzenberger

Benedict Witzenberger is a PhD student at the Professorship of Computational Social Science and Big Data at the Technical University of Munich, Germany. His research focuses on application of Computational Social Science methods on questions in Communicative Science, and the evolution of data journalism in particular.

Nicholas Diakopoulos

Nicholas Diakopoulos is a professor in Communication Studies and Computer Science (by courtesy) at Northwestern University where he directs the Computational Journalism Lab and is director of Graduate Studies for the Technology and Social Behavior PhD program. His research focuses on computational journalism, including aspects of automation and algorithms in news production, algorithmic accountability and transparency, and social media in news contexts.