News

In this talk, I will report on recent computational studies on open peer review in the context of academic journals. In one study, we compared the content of 258,000 published and unpublished peer review reports from 233 medical journals from Elsevier and Springer Nature, all submitted during a comparable time window (2016-2021). We found that open peer review reports were more informative than confidential reports, although there were differences between reports submitted by reviewers from Western and non-Western institutions, and by reviewer gender. In a study comparing open peer review reports submitted to F1000 with reports submitted to other journals under single/double blind peer review, we found similar patterns in terms of content, but did not find a correlation between report content and manuscript change. In a study of open post-peer review at e-Life, we found a decrease in the informative content of reports after the transition to a no-rejection editorial model of publication. These results allow discussion of the pros and cons of open peer review and demonstrate the benefits of computational research in evaluating peer review reforms.