As a Matter of Fact, Facts Don’t Always Matter: Psychological Mechanisms Behind Informational and Persuasive Effects of Corrected Misinformation

Misinformation can have severe consequences, such as eroding trust in democratic institutions and fueling polarization and vaccine hesitancy. It is particularly worrisome because it can steer individuals toward beliefs and actions they might not otherwise adopt. While scholars have extensively studied informational effects, such as perceived accuracy, far less is known about persuasive effects, such as the attitudes and intentions formed when people rely on misleading claims. This distinction is vital because improved accuracy judgments do not automatically translate into shifts in evaluative impressions, meaning that misinformation can continue to shape attitudes, behaviors, and intentions even after people recognize it as false. This dissertation addresses this gap, identifying the psychological mechanisms that explain how corrected misinformation leaves lasting marks on cognition and behavior. Across four chapters, this work investigates the persistence of these persuasive effects. Chapter 2 revealed a clear valence asymmetry in COVID-19-related misinformation: negative claims remained persuasive even after correction, whereas the effects of positive misinformation reversed, leading to more negative evaluations than before exposure. Chapter 3 built on this by comparing intervention timing in the political domain. It found that post-exposure debunking was significantly more effective than pre-exposure forewarning, though negative misinformation again proved most resistant to correction. Chapter 4 directly compared informational and persuasive outcomes. The results showed that corrections, particularly strong, explanatory ones, were more effective at reducing perceived veracity than at mitigating negative attitudes. This suggests that reducing belief in misinformation does not necessarily eliminate its persuasive impact. Finally, Chapter 5 identified a key reason for this persistence: higher-order inferences. While factual beliefs are relatively easy to correct, the broader inferences individuals spontaneously draw (e.g., personality traits attributed to a subject) are not. Because corrections rarely address these secondary inferences, the persuasive influence remains anchored in the individual's mind. The findings of this dissertation offer critical theoretical implications by demonstrating that the continued influence effect extends beyond factual recall to broader evaluative judgments. This suggests that prior research focusing solely on belief accuracy has likely overestimated the effectiveness of corrections. Persuasive effects are more resistant to change, partly because misinformation functions as a powerful tool for initial opinion formation even without prior knowledge. This persistence is not merely a product of motivated reasoning, but a fundamental feature of how individuals integrate new information into their (existing) worldviews. Central to these insights is the role of negativity bias and cognitive format. The resilience of negative misinformation underscores the emotional weight that drives cognitive persistence. Furthermore, by distinguishing between direct factual beliefs and higher-order inferences, this work explains why misinformation remains persuasive after correction. Explanatory corrections emerge as the most viable solution, as they allow individuals to fundamentally update the mental representations that simple corrections leave untouched. Practically, these insights suggest that the true impact of misinformation is likely underestimated. In digital environments where emotionally charged content spreads rapidly, the fact that even corrected information still shapes attitudes and behavior is a major concern. These findings recommend that platforms and policymakers move beyond "truth-testing" toward strategies that target broader competencies, such as media literacy and source evaluation. While pre-exposure warnings may have a signaling function, they must be paired with robust, explanatory corrections to be effective. In conclusion, this dissertation demonstrates that addressing misinformation requires looking beyond informational outcomes. While corrections can successfully debunk false claims, their impact on attitudes and intentions is often limited. By clarifying when and why misinformation leaves a lasting mark, this work contributes to a more comprehensive framework for mitigating its influence in an increasingly complex information landscape.