The process of peer review is in a horrible mess. There, we said it. Many others have done so and countless more think it but don’t speak out. Not a day appears to go by without the emergence of bold-faced cheating – 170 articles have now been retracted for fake peer review, or some new dubious practice – Editor quits journal over pay-for-expedited peer-review offer.
Peer review itself however remains a central tenet of academic discourse but the integrity of science is being compromised and it is at risk of being forever tarnished by scandals with the result that public trust will decline further.
That this is not a desirable outcome goes without saying. The question then becomes “what are we prepared to do about it and will researchers ever embrace a different process?”
At ScienceOpen, the research + Open Access publishing network, we’ve spent a great deal of time and effort rethinking scientific publishing and developing a better way to do peer review. In an effort to “lead by example” we facilitate non-anonymous, open, expert (only those with 5 or more publications per ORCID can review) Post-Publication Peer Review.
Other publishers also focus on reforming peer review, e.g. F1000 Research and The Winnower. Our observation is that despite vocally demanding reform, the scientific community is very resistant to change, though. Some commentators believe this is due to simple inertia and that probably plays a part – after all, scientific publishing remained unchanged for hundreds of years prior to these more turbulent times and people frequently acquiesce to a bad system because they are “used to it”.
More importantly, the system of promotion and tenure compels scientists to avoid “rocking the boat” since their published output remains a prime measure of their competence. Among the digital cognoscenti, the Impact Factor of the journal they choose to publish in is showing some signs of declining power but it still continues its vice-like grip in the minds of the majority.
The question that ScienceOpen is currently addressing is “how do we build more peer review choice and innovation into our publishing model without participating in (the increasingly problematic) anonymous pre-publication peer review as is practiced by the vast majority of publishers”?
Enter Jan Velterop, stage left (to audience applause!). For most of you, Jan needs no introduction. Originally a marine geophysicist, he became a science publisher and has worked at Elsevier, Academic Press, Nature and BioMedCentral. He participated in the Budapest Open Access Initiative. In 2005 he joined Springer, based in the UK as Director of Open Access. In 2008 he left to help further develop semantic approaches to accelerate scientific discovery.
Today we are delighted to announce that Jan is joining our Advisory Board. He will help us launch “Peer Review by Endorsement” which occurs, just as usual peer review, before publication, but is entirely open and transparent. Authors will be able to choose the “Peer Review by Endorsement” option. Articles published this way will also be available for Post-Publication Peer Review, as are all 1.5 million OA articles aggregated on our site. This option will go live on our site during the summer of 2015.
So what is Peer Review by Endorsement? Rather than publisher-mediated peer review before publication, the scientific community takes this role and the publisher verifies the results. As Jan puts it:
It is more efficient and cost effective to hand peer review entirely back to the scientific community, where it rightly belongs, than for publishers to find the right, appropriate, available, reliable, expert reviewers.
Authors would be expected to arrange (or ask their Scholarly Society to arrange) for at least two peers to check the scientific soundness of their work and, if they are satisfied, to openly endorse its publication by declaring that in their view the work is suitable for being published as part of the scientific discourse. The work’s ‘significance’ is not an issue here (as that can often only be established after some time in the open anyway, and it has the considerable drawback of preventing some articles, e.g. null-results, from being published). The rules are that peer-reviewers/endorsers must be active researchers, and not be, or for at least five years have been, at the same institution as, or a co-author of, any of the authors. Once two signed and open peer reviews/endorsements are available, the article will be immediately published and, as usual for all articles published on ScienceOpen, available for further Post-Publication Peer Review.
We hope by introducing a two stage peer review process (Peer Review by Endorsement and Post-Publication Peer Review) to improve this mechanism for all. In the unlikely event of manipulation (present on a near daily basis in the traditional system), it will be transparent for all to see, which is bound to be a powerful antidote. As ScienceOpen is integrated with ORCID and reviews/endorsements are signed and non-anonymous, there is very little danger of sub-standard articles being published, as endorsers/reviewers would not want to put their reputations at risk.
Improvements to the original manuscripts, we believe, should be among the aims of peer review. Author-arranged Peer Review by Endorsement is conducive to an iterative process between authors and reviewers/endorsers, delivering those improvements.
Since arranging traditional pre-publication peer-review can be difficult for publishers, and can be slowed down by the necessary research to find appropriate reviewers, it can be quite costly. Especially since the cost of reviewing all submissions is usually carried only by those articles that are accepted for publication (this applies to the open access as well as pay-walled publishing models). The Peer Review by Endorsement option avoids that and authors choosing that option will therefore have their APC’s reduced. The regular Article Processing Fee (APC) for publishing in ScienceOpen is $800 and over the coming months prior to launch we will be seeking community feedback on the most appropriate discount level.
According to Jan:
The Peer Review by Endorsement approach leaves peer review to the community (with safeguards in place) and lets the publisher focus strongly on the technical integrity of the article presentation, preservation, machine-readability and the like, which often leaves a lot to be desired in the current system. The cost to authors (and their funders) of open access publishing will be materially be lower as a result.
Jan will be speaking about Peer Review by Endorsement at The Future of Scholarly Scientific Communication Part 1 event at The Royal Society (London, 20-21 April 2015). These meetings are being held in recognition of the 350th anniversary of Philosophical Transactions, the world’s first science journal.
On Monday afternoon, he is taking part in a session entitled “Peering at Review”, where he will join a conversation about “Future developments, evolution and alternatives”, together with Dr Richard Sever (Cold Spring Harbor Press) and Elizabeth Marincola, (CEO, PLOS) who recently said “At PLOS, in the not-too-distant future, we want to improve many aspects of peer review”.
Building a better Peer Review mechanism is certainly not easy but here at ScienceOpen we are committed to demonstrating a creative vision for a healthier ecosystem.