Keeping with a theme of disagreeing with letters to Science, let’s discuss the concept of sequential reviewing. In a recent correspondence, Dr. Simcha Lev-Yadun points out an obvious problem: the placement of several reviewers on a flawed manuscript wastes that most valuable of publishing resources, the reviewers’ time. Instead of having multiple experts trash a bad manuscript, Dr. Lev-Yadun suggests that editors should send out manuscripts to a single reviewer, allow the authors to make revisions, and then bring in an additional (or more) reviewers to assess the revised and (hopefully) improved manuscript. This saves time for the additional reviewers (but not the first), allowing the late-comers to focus on more important aspects of the manuscript and not on trivial problems.

I disagree. Not only will this slow down the entire publication process, but it will drive authors nuts. Running the gauntlet of peer review is best done all at once.

Cartoon credit: Nick Kim

Let’s consider this for a moment…the first reviewer has several concerns, the author puts time into fixing those, and then potentially receives a new set of concerns from another reviewer requiring another revision and re-writing?? What if the editor then decides to add a third reviewer???

To be fair, this proposal may work for a very specialized society journal in which the scope is very narrow. In this case, refs will have overlapping expertise and perhaps those trivial problems that concern Dr. Lev-Yadun would indeed be cleared up prior to taking up the time of a second reviewer.

But for a paper such as this the strategy would be a complete failure. For the paper I referenced, an editor would need to assign reviewers with expertise in epigenetics, learning & memory and possibly neuro-degenerative disorder models. It is ridiculous to apply a sequential strategy when so much expertise is needed. The overlap will be minimal and thus, conducting the review at the same time is essential to reduce the total time prior to publication. Even just casually looking at the latest issue of a more specialized journal, this manuscript likely had experts independently assessing apoptosis/cell signaling or neuronal receptor trafficking, with perhaps even a separate ref focusing on spinal cord injury models.

Let’s put it another way. If I were an author of a paper that even barely crossed sub-fields (and these days, most studies do), I would be pulling my hair out after being told that I first have to address the concerns regarding Figs. 1-3 and then in the next round, perhaps have to also tack on additional revisions regarding Figs. 4-6, with some minor tweaking again to Figs. 1-3 because the new reviewer found a few other minor problems.

The review system is not perfect, but being on the “inside”, I have to say how impressed I am with its efficiency. Remember, my friends (said with a nod to Senator McCain), this is all pro bono work on the part of the refs. And a vast majority of reviewers take these obligations very seriously. Here is what I think is a general view: scientists like an intellectual challenge, and nothing is more stimulating than breaking down a manuscript and engaging in advanced debate within your favorite specialty.

Looking at this from another angle, sequential reviewing would actually deprive younger scientists of the irreplaceable learning experience of comparing one’s review to those of other experts. Now that my graduate school and post-doc colleagues are getting to the point where they are starting their own labs, they are constantly asking me to send them papers to review so that they can learn. After a decision is made on a manuscript, I send all of the anonymous reviews to the reviewers, not for their records, but so they can see how their views calibrate to those of their peers. By comparing views, the reviewer gets a sense of where others in the field stand, and can perhaps bring a different perspective to their next review (whether to be “softer” or “tougher”), improving the entire review process on the fly!!!

As for the argument that scientists are reluctant to review because of the vast number of requests they receive, this is of course true to a certain point. But sequential reviewing will not fix this problem nearly as well as a review transfer process, as is conducted by the Neuroscience Peer Review Consortium. To tell you the truth, though, I don’t have much of a problem finding reviewers for papers. I just checked my “stats” and my acceptance rate for review requests across all of my subject areas over the past 3 months is 65%. This number is pretty impressive even before considering that the way this number is calculated, it does not exclude requests I made that were positive responses after I already felt I had enough expertise on a paper (e.g., say I make 4 requests on a paper, and all four eventually say yes. I place the first three positive responses on the paper and tell the 4th “thanks, but perhaps I can use you on another one in the near future”).

There is another way to combat Dr. Lev-Yadun’s concern of wasting valuable reviewer time by sending him/her an under-developed, or poorly-written, or illogical manuscript. DON’T SEND OUT THE MANUSCRIPT FOR REVIEW. So I would like to make a counter-proposal, if indeed this problem of wasting reviewers’ time is very real – EDITORS, DO YOUR JOB.