Abstract

For unverified papers deposited as preprints, users must independently judge the reliability and accuracy of the research content. Therefore, this study aims to determine whether the influence of the officially published version of the paper can be predicted in advance using various usage indicators of preprints. In addition, we analyzed whether, despite being an unverified, unpublished preprint, it was being spread socially through media reports and cited in academic papers. Total of 460 papers were collected and analyzed, including 288 papers initially archived as preprints in the medRxiv health informatics field and 172 post-publication versions of the papers. As a result, first, it was confirmed that the initial interest in research results published as a preprint was related to the number of citations in the post-publication version. It was analyzed that using the usage indicators accumulated over the first six months could predict the influence that may appear in future post-publication versions at around 40%(β = 0.409 p < 0.01), helping users to select research results. Second, it was analyzed that unpublished preprints were also spreading socially and being cited in academic papers. In addition, it was analyzed that it is being actively cited not only in other preprints but also in excellent academic journals with a high impact factor.