This paper investigates how individuals reason about the authenticity of the news content they consume. While researchers have conducted much work on fake news detection and prevention, we know relatively less about how news readers reason about the content that they read. Using data collected through Amazon Mechanical Turk, we analyzed over 1,000 justifications that news readers provided about why they believe (or fail to believe) given news articles. We included both fake and credible articles in our analyses and examined the novelty of the news topic as a possible contingency factor that differentiated the reasoning provided. Based on our psycholinguistic analyses, we found that news readers employ both cognitive and motivated reasoning and that agreement with the ground truth impacts the reasoning more than a news topic’s novelty. Our insights contribute to the literature on news consumption and reasoning in the context of evaluating fake news. Furthermore, this knowledge contribution has implications in areas such as news veracity intervention and tool design. Lastly, we offer a methodological contribution via using linguistic analysis in a novel way to assess the quality of open-ended survey questions.
Horne, B. D.,
The Reasoning behind Fake News Assessments: A Linguistic Analysis.
AIS Transactions on Human-Computer Interaction, 14(2), 230-253.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.