The "peer reviewers get worse" item in this morning's Links Dump drew an immediate comment elsewhere to the effect of "of course they do, because they start pawning reviews off on their students. This one was a surprise to me, so here's a quick poll to see if my subfield of physics is really that much more collegial and ethical than the rest of science:
Refereeing takes place via classical communications channels, so you may only choose one option.
My experience is that while I have written referee reports that my boss didn't want to write, it was done above-board-- that is, my boss contacted the journal editor, said he was too busy to review the paper, and told him to use me as the replacement referee. I have a hard time imagining any of the people I worked for or with having a student write a referee report then handing it in as if they'd done it themselves.
If this really is common in other fields, then, well, one more item for the "Why you really want to be an AMO physicist rather than one of those assholes" file when advising students...
Refereeing is an important part of the business, and one to which students who expect to continue in the field should be exposed. But when it's done with actual papers/proposals, yes, the editor/program officer should be in the loop. That's especially true in my field, where some of the journals actually do publish referee names (provided the referee approves). In that case, of course it would be unethical for Prof. Bigshot to claim the review is his own, as he would be claiming credit for the work of his underling.
As one of the commenters on the Retraction Watch piece noted, sometimes farming out referee reports to your trainees can result in higher quality. Younger scientists know the latest methods and results, and they don't have fond memories of paradigm X which was later disproven. Also, younger scientists are more likely to have the time to do a thorough job; senior scientists have more demands on their time. When Prof. Bigshot actually does do the review, it is all too likely to be along the lines of, "Cite the paper by Youngblood and Bigshot (J. Subfield Phys., 200x) on this topic; no other changes needed."
It would be nice if there were an option between completely unethical and fine. At least in my group, I think there would be no issue with the grad students or the post docs writing at least the first draft of a report, as we're the ones who are more familiar with the current literature on our particular projects than the PI. Still, I wouldn't want a report that was entirely written by such . . .
Most journal review requests I've gotten specifically say, in marginally more formal language, feel free to hand this off to your grad students.
Seems completely fine to me. If you have the editor assign it such, it might be a very good thing.
If you don't, it has to be a *supervised* process; i.e. having the student write the thing and then the faculty editing it appropriately.
A student review submitted without the in name referee even reading it? That's the only thing that trips my ethical sense as wrong.
For the record, if you are given a paper to review and it includes the exact work you have assigned to your research assistant graduate student, it is most appropriate to tell the editor they should get another reviewer and tell your grad student they need to revise their project. Rejecting the paper that is mostly valid even if you think it is imperfect, or sitting on the review to stall for time, and NOT telling the student but just trying to kick them in the ass to work faster is NOT ethical. /yes this happened to me
sitting on the review to stall for time ... is NOT ethical
I agree that somebody who does this for the reason becca gives for this having been done unto her is being particularly unethical, but in general slow refereeing is a pet peeve of mine. If you don't have the time to do the review, don't agree to do it in the first place. The editor/program officer will thank you for not making him hound you for the review, and for not giving the author a reason to hound him for referee reports on the paper that has been under review for months because you didn't take the time to read it. It's usually easy to find another reviewer. For many of the journals that I deal with, the editor already has another reviewer in mind if you decline.
BTW, the reason I suggested working with the editor/program officer if you plan to involve your trainee in the review is because for many journals the "Suggest Alternate Reviewer" path isn't guaranteed to steer it to your protege. I have received referee requests where, on clicking the link to accept, I have gotten a message saying that the editor has already found enough referees for the paper.
My grad students are way too busy washing my car and doing my laundry to write reviews for me, so I use a website where Indian and Chinese students write the review. It's actually quite reasonably priced...
I have more of a problem with grad students than post docs, but there is also a difference between "writes the report" and "writes a draft" or "comments on a draft". The latter is part of the learning/training process.
I wouldn't do this for grant proposal reviews, but have done for journal articles (with permission from the editor). Obviously you want to look at the review before it is sent off, and discuss it with the student.
How else are students supposed to learn how to do reviews?
As an editor at Nature Materials, my personal opinion on the issue in short is that it is perfectly fine. But a few points from my side:
- I am aware that some scientists routinely "delegate" refereeing to someone in their group, some even discuss papers with a number of students. This seems ok to me as I assume that the guy doing the referee actually knows much better than the PI how the experiments are done. However, if the report is delivered in their name I would expect the PI that I approached initially to at least go over the report and do some sort of quality control (and perhaps discuss it with the student to train them if you like).
- It should also be good practice to name the person who contributed to the report (in confidential remarks to the editor). As always in science, I think credit to whom credit is due, so this should be good practice. Also, if the report is really great, next time it saves me the time to go via the PI. I've learned about young scientists that are great referees this way...
- Of course the option of declining a referee request and suggesting a member of your team instead is perfectly fine also, and makes the involvement of your postdoc for example more official.
This was common practice where I made my PhD. (Don't know if it still is.) Typically you'd find the to-be-reviewed paper on your desk when the review was several weeks overdue already. And of course you were asked to do it because you had to learn it, not because some prof wasn't in the mood.
The most recent request I got to review a paper seems to be ahead of the curve on this. Instead of waiting for us to decide what's ethical and how it should work, this journal (the top journal in its field, BTW) is actively recruiting student/postdoc involvement, with what seems to me to be the right perspective. Here's what they said in the request to review:
"If you are short of time, consider having a highly qualified graduate student or post-doc assist you. Students are particularly good at evaluating technical details and presentation. However, few students have the perspective needed to determine whether a paper contributes important new data and ideas -- that is why we would want you to be the responsible person. In any case, involving a student is an excellent way to refresh the [name of journal] community (if you mention his/her name when you return the review I will pass that information on to the Associate Editor, who can thus become acquainted with him/her), builds the student's critical faculties, and is a very effective way to improve his/her own writing skills."