Scientific uncertainty and sentiment in COVID-19 reporting: An analysis of headlines and article bodies
Scientific uncertainty and sentiment in COVID-19 reporting: An analysis of headlines and article bodies

Summary/Abstract: The COVID-19 pandemic required media professionals to communicate rapidly evolving and often uncertain science to their audiences. This study investigates how scientific uncertainty and sentiment intersect in journalistic materials published during this global health crisis, focusing on differences between general interest and science-focused media. Our analysis is based on a large corpus of journalistic articles that were published at the onset of the pandemic (March–August 2020) and in two additional timeframes, the same months in 2017 and 2023. We used dictionary-based uncertainty detection and automated sentiment analysis to examine headlines and article body content in the two types of publications. Our findings indicate that articles with high scientific uncertainty were more likely to feature negative headlines across both outlet types, though this trend did not extend to body content. COVID-related articles published in 2020 revealed heightened negativity in both headlines and body content for general interest outlets. For specialized outlets, negativity was primarily present in body content. These results contribute to a more nuanced understanding of the complex dynamics between scientific uncertainty and sentiment in the context of public health crises.

• Details
• Contents