Peer review would be improved by discussion across the lab
The current individual-based review system has its limitations. Modern research is both multidisciplinary and technical, and it is often difficult to find reviewers that have significant expertise across subfields together with technical know-how in a specific discipline. They are also given only a limited time to complete their review. So reviewers, even those within the same field of research, often differ in their evaluation of a paper. As an alternative, I suggest we use a system which I call peer group review.
In peer group review, an editor would have two options when sending a manuscript out for review. A single reviewer could evaluate the whole work on their own and submit a recommendation to the editor on the basis of its scientific and technical merits; this is identical to current practice at many journals. Or the reviewer could formally discuss the manuscript with his or her research group, gathering input from graduate students, postdocs, technicians and other researchers, and then use this discussion to make a recommendation to the journal. This second option is analogous to the journal clubs already in place at many labs, albeit more formal, and is the main idea of the peer group review system.
Regular journal clubs discuss papers only after publication. It is not uncommon, in these discussions, to hear students or colleagues say: “The authors should have done this control or that experiment” or “Why don’t I get such picture-perfect data?” With a formal peer group review system, these comments would be more likely to come out before publication, and authors could address them in the final manuscript.
An important difference between informal journal clubs and peer group review would be the removal of non-scientific and non-technical information from the manuscript to maintain confidentiality of the source. Hiding authors’ names would prevent publication bias towards ‘star’ authors (see ‘Fraud: anonymous ‘stars’ would not dazzle reviewers’. The main reviewer would be responsible for ensuring this confidentiality.
The short story
I propose that authors provide two versions of their manuscript at submission: a full-length paper and a shorter version containing key text and important figures only. Such a dual submission is not without precedent: FASEB Journal regularly solicits both full-length and three-page abbreviated versions for their FASEB Express format. But, unlike FASEB Journal, the short version proposed would have no references, grant support or other investigator-specific information.
Reviewers would be given both versions of a manuscript for use either in the current peer review process as usual, or in the group review system. I believe that technical errors which could be overlooked by a grant-savvy principal investigator, would be caught by a tech-savvy graduate student or experienced technician present in the peer group review.
Centres of excellence
Journals could identify several group review centres on the basis of their expertise and technical skills, which could act as peer-review centres across the world. I further suggest that, as peer reviewers are currently not compensated and not always appreciated, these centres be acknowledged annually
Of course, not all reviewers would be able or willing to discuss paper content in a journal club context. But by maintaining the option of reviewing a paper in the usual way, these reviewers would still be able to participate in the peer-review process.
This proposed system should enhance the consistency and quality of the review process itself and do more to ensure the publication of high-quality papers. In my view, critical discussion and review of a manuscript before publication rather than after it, might have prevented some of the misconduct scandals of recent months (for example, see ‘Verdict: Hwang’s human stem cells were all fakes’
1. Kawasaki, H. & Taira, K. Nature 426, 100 (2003).
Debomoy Lahiri is a professor of neurobiology in psychiatry and of medical and molecular genetics at Indiana University School of Medicine, Indianapolis. He is particularly interested in understanding ageing mechanisms and gene regulation in Alzheimer’s disease, and is editor-in-chief of Current Alzheimer Research.