With over 4,700 open access repositories in operation, the use of preprints (research papers made publicly available before publication in a peer reviewed journal) is growing. However, the lack of peer review means that it may be harder to identify high-quality papers and to interpret and cite the findings published in these manuscripts. In an article published on the LSE Impact Blog, Wang LingFeng discusses how peer review for preprints could be implemented.
A survey conducted by arXiv in 2016, found that 58% of its users supported a peer review system for preprints. A Community Peer Review model was suggested, in which registered readers comment on and rate manuscripts. LingFeng highlights 3 potential pitfalls of this system: the potential for unethical behaviour (soliciting favourable comments), papers receiving uneven review attention regardless of quality, and the lack of control over whether, and how rapidly, papers are reviewed.
LingFeng offers an alternative concept – self-organising peer review (SOPR) – an automated method which would match reviewers to articles, by ranking both. Under SOPR:
- Authors of deposited preprints would be registered as reviewers and ranked based on their publication record.
- Submitted preprints would be given an estimated quality level, based on the ratings of recent submissions by the corresponding author.
- Anonymised preprints would be posted and matched to reviewers with a similar ranking (also considering their field and associations with the authors, and excluding recent reviewers).
- Each preprint would be reviewed by 3 people, prior to publication of the author list.
- Reviewers failing to accept an assignment or to complete it on time would be penalised (submissions would not be reviewed and database access suspended).
LingFeng proposes that this rules-based approach should foil any attempts to game the system and should ensure that every preprint receives a timely review, helping to improve the recognition and acceptance of preprints in the academic community. LingFeng also suggests that this approach could enable publishing to flourish under open science initiatives such as Plan S.