Barely a week goes by without a new proposal to improve peer review: how to make it faster, better at spotting errors, more transparent, less prone to bias, less burdensome. But it’s difficult to track this ferment — and to glean lessons from it. So Nature welcomes the launch of a registry of platforms and experiments around peer review. Called ReimagineReview, the online hub invites researchers to add projects and to raise awareness of peer-review trials.
Nature will use the platform to document its own peer-review experimentation. More than a decade ago, this journal trialled allowing public comments on manuscripts while the papers were being evaluated (Nature 444, 971–972; 2006). Since 2015, Nature-branded journals have offered authors double-blind peer review, in which reviewers and authors do not know each others’ names. An analysis of more than 128,000 manuscripts in this trial found that authors from less-prestigious institutions were more likely to choose double-blind review (B. McGillivray and E. De Ranieri Res. Integr. Peer Rev. 3, 5; 2018). We are currently investigating publishing anonymous referee reports on Nature Research journals, as already offered by Nature Communications (unless authors opt out).
Many publishers are trialling ideas. BMC (part of Springer Nature) and the British Medical Journal were the first to offer open peer review; eLife and F1000Research, among others, have experimented with open formats, such as allowing authors and reviewers to interact directly, or publishing manuscripts before full review. The term ‘open review’ covers many practices, and it is not easy to measure their pros and cons. Last week, researchers published guidelines for editors wanting to move to open review; these sprang from a workshop at Nature’s offices (T. Ross-Hellauer and E. Görögh Res. Integr, Peer Rev. 4, 4; 2019). Understanding which data on peer-review trials can be captured and how to report them is an effort we’re committed to, to help authors and reviewers. It is in the scientific community’s best interests for everyone to share what they learn.
Nature 567, 5 (2019)