Investigating Perceived Message Credibility and Detection Accuracy of Fake and Real Information Across Information Types and Modalities.

(Mis-)information thrives on social media, so it has become increasingly important for users to tell real from misleading content because erroneously following misinformation can cause serious consequences. In this study, we investigated users' subjective perceived information credibility and objective detection accuracy of fake and real information across three topics, each of which was delivered via two modalities. We conducted an online within-subject experiment (n = 293) with a three (information topics: health, science, and life) by two (modalities: text and image) by two (veracity: real and fake) design. Overall, our participants were better at identifying fake information from real information. Results also found that information types significantly mediated information modality and veracity. For real information, the text mode helped people perceive credibility in health, life, and science topics. However, for false information, the image mode helped raised the perceived credibility of life information than the text mode.