Effect of Social Media on Continued Use of Retracted Articles : Evidence from Retraction Watch

Continued use of retracted articles amplifies the adverse consequences of propagating questionable knowledge. Because traditional approaches have not been adequate in alleviating continued use of retracted articles, scholars have called for the use of social media to aid in disseminating retraction information. However, it is not known whether (and under what circumstances) social media reporting can reduce continued use of retracted articles. It is also not known if social media reporting can help researchers to make use of retracted articles in a more rational way. To answer these questions, this study conducted a natural experiment (based on Retraction Watch) to quantify the effect of social media reporting on continued use of retracted articles. The DID estimates reveal that social media reporting significantly reduced post-retraction citations of non-swiftly retracted articles in the first two years after retraction. Further estimates show that additional article-related information attenuated while additional author-related information enhanced the effect of social media reporting on post-retraction citations. In addition, results of subsample analysis demonstrate that the effect of social media reporting on post-retraction citations was more salient for articles published in higher impact journals, articles not self-retracted by authors, and articles retracted due to research misconduct, plagiarism, or duplicate publication. For articles retracted due to research error, social media reporting increased valid citations of retracted articles. This study makes important theoretical contributions to the scientific community literature and offers useful practical implications about how to use social media reporting to reduce continued use of retracted articles