Truth-Detection in News Stories Presented with Correspondent Images

Fake news was widespread during the 2016 election year, particularly on social media platforms such as Facebook. Generation Z adults are turning to social media sites for news, with Facebook a leading source. Generation Z adults, who are civically engaged but may be politically naïve, are the adults at highest risk of exposure to fake news. Whether information is true or false can be difficult to evaluate, and truth judgement is affected by processing fluency. When reading news stories, or other new information, processing fluency is affected by the presence of non-probative images accompanying text, and by the level of processing fluency attendant to the semantic-relation of those images to the text. Probative images have also been shown to affect belief. This investigation was designed to test how both probative and non-probative images affect truth detection among young adults of Generation Z, in the context of full-length, online news stories. We created a factorial design consisting of a 3 Text Veracity (true vs. false vs. mixed true and false) × 4 Image Content (true-supporting vs. false-supporting vs. non-probative vs. no image) fixed ANOVA. Results indicated that: 1) participants endorsed each story as true significantly more frequently than expected by chance; 2) participants’ truth endorsement was unaffected by the presence of images, or by the level of cognitive fluency promoted by the images; and 3) participants’ confidence in their endorsement was unaffected by the level of cognitive fluency promoted by the images, but instead by whether the participants endorsed the story as true. We conclude that the full-length news stories we used produced sufficient cognitive load to overpower the influences of image presence, and fluency. We suggest that upon endorsing the news story as true, participants experienced a form of confirmation bias, resulting in increased confidence in their endorsement. Future directions are discussed