Seeing through lies: Correlation between physiological and sentiment metrics in the consumption and dissemination of fake news

Authors:
Preprint
Publisher website
Google Scholar
Abstract
This study investigates the distinction between fake and real news articles by analysing both linguistic characteristics and the reader's physiological responses during reading. We employed a combination of eye-tracking and peripheral physiological measurements to capture real-time physiological reactions, alongside self-reported responses, as participants engaged with these articles. Our primary objectives were to assess whether physiological responses and eye-tracking patterns differ between fake and real news and to identify linguistic descriptors that most strongly correlate with these responses. The results reveal that specific physiological metrics can differentiate between fake and real news, and we observed notable correlations between certain linguistic features and these physiological responses.