Author preprint behaviour and non‐compliance with journal preprint policies: One biomedical journal's experience

We investigated author use of preprints including compliance with journal preprint policy, timing of posting, likelihood of preprint review impacting manuscript content, and consistency in licensing. Submissions to one biomedical journal were reviewed over 2.5 years. About 24% of authors were non‐compliant with the journal's preprint policy. Only three journal articles out of 76 posted as preprints received publicly available comments on the preprints. Additionally, most preprints and articles were submitted so closely together temporally (<10 days apart) that even if comments had been posted online in response to the preprint, they were unlikely to have led to changes in manuscript content submitted to the journal. iThenticate plagiarism detection software scored preprints and articles 83% similar (median), providing additional evidence that preprint posting does not notably affect manuscript content. Lastly, preprint versus journal publication licence types were compared. Most authors chose different licence types for the preprint and subsequent journal article, setting up an inherent conflict about reuse of very similar work and incompatible dual licensing of essentially a single product. These insights into author use of preprints and compliance with policy furthers understanding of author utilization of preprints, informs future author education efforts, and supports attempts to harmonize licensing.