People frequently encounter false or misleading information online, and initial exposure can continue to influence beliefs even after corrections are issued. This study examined how retraction affects belief in research findings.
Thirty-six psychology research summaries were drawn from the Retraction Watch Database and randomly assigned to retracted or published conditions. The only difference between conditions was the final sentence, which stated either that the study had been retracted or that it had been made publicly available. To counterbalance the retraction effect, two versions of the study materials were created so that each research summary had an equal chance of appearing in either the retracted or the published condition. Participants rated their agreement with each study's main conclusion before and one day after reading the summaries and also reported their level of comprehension.
Endorsement increased after reading, consistent with a familiarity-as-truth effect. However, among participants with high comprehension, belief change was smaller for retracted studies than for published studies. No such difference appeared among participants with low comprehension.These findings indicate that retraction can reduce the persuasive impact of false research, but only when readers have a strong understanding of the material. Improving public comprehension of scientific content may therefore be critical for the corrective power of retractions to take hold.