Pictures Are Not Always Worth a Thousand Words: Nonprobative Pictures Did Not Increase the Effectiveness of Misinformation Corrections

Abstract

Much research has focused on the language used for debunking false beliefs: communications should lead with facts, label misinformation as false, and reinforce true information. Pictures are used in debunking messages, but it remains unclear whether they matter, and what content they should depict. Six experiments explored these issues, focusing on correcting science misconceptions. There were no effects of nonprobative pictures on correcting misinformation beyond the effects of corrective text alone. This was true regardless of whether the image depicted the misconception (Experiments 1, 3–6) or the correct referent (Experiments 2, 3, 5, 6), whether a symbol reinforcing truth or falseness (red X or green checkmark) accompanied the image (Experiments 1–6), whether belief was assessed immediately or after a delay (Experiments 4–6), and in an exploratory last experiment, whether belief in real-world health issues was targeted. Simple messages yielded robust correction, but nonprobative pictorial aids did not increase their potency.