Wednesday, June 10, 2009

Co-authorial Rights

Something caught my eye in a recent NYT article about a supposedly falsified biomed study in which the first author forged the signatures of co-authors whom he selected without informing them about the study or their co-authorship. When problems with the (published) paper were revealed and the co-authors realized they were co-authors on a paper they had never seen, involving research they had not done, at least one of them tried to gain access to the reviews of the manuscript and other editorial correspondence with the first author. He was denied access.

This is from the article, with additional info added by me in brackets:

Dr. Andersen [one of the co-authors who didn't know he was a co-author until the paper was published], curious about what Dr. Kuklo [the first author] had actually submitted, asked Dr. Heckman [a journal editor] for copies of those reviews. But the editor turned him down, even though Dr. Andersen was supposedly one of the study’s authors. In a recent interview, Dr. Heckman said that his journal, like many others, considered such reviews confidential and shared them only with a study’s lead author.

“It is all confidential information,” Dr. Heckman said, when asked by a reporter for the reviews. “It is protected by the peer-review process.”

I can see why a random person couldn't write to a journal editor and request to see so-and-so's reviews, but why is this information confidential with respect to co-authors?

This general issue reminded me of a conversation I had at a conference last spring with a colleague. He hates the fact that he almost never gets to see the reviews and editor comments for manuscripts to which he has contributed as a co-author. Some of his first-author colleagues won't send him the reviews even when he asks. He had never tried asking an editor if he could see the reviews as well, but perhaps it wouldn't have mattered.

Is Dr. Heckman right? I am not sure he is, but if he is, why can't co-authors see review materials? If co-authors are responsible for the content of papers, shouldn't they have the right to see the reviews?

In a few instances, I (in my role as first author) have not wanted to share review and editorial correspondence with co-authors, for reasons I will outline below, but if any of these co-authors had asked me directly if they could see the reviews, I would have complied with their request. And if a journal had a policy of giving co-authors access to reviews, I would not object.

The exceptions I can think of at the moment occurred when:

(1) the co-author was a somewhat junior student and the reviews were 'not constructive' (= hostile and unprofessional, possibly including insulting comments). Eventually students should see reviews in their raw form and learn how to deal with negative comments displaying various magnitudes of rudeness; I expect senior grad students to participate fully in reading and responding to reviews. However, I have seen the crippling effects of harsh first-reviews on students, and would prefer to ease them into the experience of being attacked for no obvious good reason.

(2) I hated my co-author. I can think of one case in which I ended up not having much choice but to co-author a paper with someone who was not only hated by me but by most of the rest of the world. My other option was not submitting the manuscript, but I had a lot of time and effort (and $$) invested in this project and was unwilling to drop it without at least attempting to publish one paper. Communicating with my odious co-author about anything, however benign a topic, tended to unleash paranoid rantings about all the people he hated (they were wrong about everything, he was right), and I didn't want to know what his response would be to reviewer comments, even though they were mostly mild and constructive. I took care of all the revisions myself and presented the finished manuscript to him as a fait accompli.

Despite my aberrant and hypocritical behavior in these cases, my general opinion is that co-authors have a right to see reviews and be fully informed of the review process and editor decisions.

I realize that some manuscripts have 57 authors and it might not be practical to involve everyone, but perhaps in these cases the corresponding author could indicate the 5 or 10 co-authors who should have access, if they so choose, to information related to the review process.

Or am I missing something? Is there a downside to allowing some or all co-authors to have access to reviews?

29 comments:

Anonymous said...

Some journals in biological sciences now demand the e-mail addresses of all co-authors (not just the corresponding author) and automatically send the reviews to one and all. Only the corresponding author is empowered to respond by submitted a revised manuscript, but all co-authors are included in the review process. This would work against your desire to project jr students from the nasties of the world, but is a simple and straightforward way of emphasizing that all authors are accountable.

female Science Professor said...

That system makes a lot of sense to me.

Matt said...

That is outrageous! In my experience, even on papers with up to 20 authors, reviews are circulated to all co-authors (by the corresponding author) and the co-authors are consulted before a response is sent back to the journal. I would hesitate to collaborate again with any author who resisted this transparency.

Sara said...

Even though the review process is meant to ensure the quality of articles, isn't it a way to learn through feedback too? If the co-authors is denied to see the reviews they also get deprived of the possibility to learn from the feedback. In the end that is detrimental for the whole research community, is it not?

Kevin said...

I thought it was standard to send referee comments to all authors. Certainly any claim that the peer-review comments are confidential from the authors is completely against the basic ethics of peer review, no matter what the editor said.

Some conferences and journals have a practice of sending the referee comments to all the referees. I have found this to be a very useful practice---it raises the awareness of the sloppy referees to what they should have noticed, and it helps the referees feel more confident of their judgments when they see that all three referees have independently identified the same flaws in the work. It also cuts down somewhat on the vitriolic reviews, except in the case of really awful papers.

Maxine said...

This is an interesting debate. At Nature (where I am an editor) we would send a coauthor the reviews upon request - coauthors are registered at submission of the manuscript. However, it is a nightmare for a journal in terms of version control, enormous time/staff costs, if all the coauthors contact the journal wanting changes, etc, so as you and the above commenter point out, we have a policy of dealing only with the corresponding author so far as the manuscript itself is concerned. We have never had an issue, so far as I am aware, where we've had dissent between authors during the peer-review process - if it happens, it happens somewhere else!

We have a set of authorship "ethics" policies which the corresponding author has to check that they've read and agreed to as part of the submission process - these include maintaining good communication with coauthors. (The average number of authors on a Nature paper is 6, consortia excepted.) Obviously, this is not perfect in that some will check without having read, but it does provide some guidance to corresponding authors (who may not have gone through this process before) as to their responsibilities to their coauthors, and signals to them that the journal takes these matters seriously.

To me, it seems amazing that a corresponding author would not share reviews with coauthors as part of the standard revision process (all submissions to Nature go through at least 2 rounds of peer-review before they are published.) Surely this is the whole point of scientific publishing?

When it comes to proofs, we explicity ask the corresponding author to circulate the proof among coauthors and coordinate their changes. Most, but not all, do. It's quite surprising to me how many details the author misses on proof, and it is not unknown for a coauthor to contact us after publication to point out an error (eg a name spelt wrong) - becuase they were not sent a proof. Part of Nature's standard procedure is to send the corresponding author the edited file before the final PDF version is sent to them, so the author does have two opportunities to make edits/corrections.

Meena said...

I think every author has the right to see every review, and to submit revisions. The corresponding author should come into the picture only after all revisions, when the article is being handled by non-academic publishing staff -- who will send (and resend, and resend in different formats ...) the source files, who will check proofs ... . it should be absolutely clear that at this stage, no changes are being made to content.

I do not see a single reason why things should be otherwise.

Anonymous said...

Just asked around the office about this one--our guess is maybe the editor is trying to cover himself. While it is typically the corresponding author's job to distribute the referee reports (if they choose to do so), a co-author writing in and asking would not be denied here.

-an editor for a scientific publishing company

Anonymous said...

I would be very interested to know the policies of Science and Nature in this regard, especially since Nature requires a statement detailing each author's contribution to the paper.

Anybody know?

qaz said...

I think it is important for junior students to see the nastiness of the world. The reviews just need to be discussed with them rather than just sent to them.

I have to say I have been really surprised at this story (which has now been discussed on a couple of blogs). I always assumed that all co-authors legally had a right (and a responsibility) to see the reviews, good or bad.

Just as I would never send a paper in without that final "OK" email from the co-authors, I would never hide the reviews. I might attach a comment like "these are stupid, you can ignore them", but I'd never not send them.

Anonymous said...

I like anonymous 12:52's biological sciences journal practices. I think it is absolutely ridiculous to keep reviews from co-authors because of "confidentiality" issues.

Regarding sharing reviews with others, I don't agree with keeping them from junior graduate students. It's good to help them develop coping skills early, since they will no doubt face many ridiculous, negative reviews in their academic careers. A friend of mine happened to be protected from negative reviews (a combination of luck for his own papers and a protective advisor for others) until he landed his professor gig, and the first two rejections almost did him in.

Sam said...

I had no idea that some journals only gave reviews to the lead author. I guess I never really thought about it before.

I am blessed with a wonderful advisor that does the following:
1. Has all grad students in the lab (5 of us) read, copy-edit, and make comments on manuscripts about to be submitted for publication (even if not all of us are on the author list).
2. Shares the review comments from the papers with all the grad students in the lab, even if they weren't on the author list (though I haven't seen any personal attacks, so maybe she edits those out before sending to us...)
3. Believes it is part of a grad student's education to participate in paper writing, editing, choosing the journal, responding to reviews, etc. Aside from the research, this is a process we need to learn to navigate professionally if we are to survive in academia.

I appreciate my advisor a great deal - she is an FSP as well, in neuroscience - and I thank my lucky stars that I have her as a mentor.

Oak said...

The only instance I can see whereby
review commments might be withheld from the co-authors is the one with the JS you mention. If the case is that I wish to withhold harshness from JS co-authors, I generally withhold the review comments until I have a chance to answer them, and show the lot to the co-authors then. Apart from that, no further issues should arise from any perspective - co-authors also assume co-responsability, so why should they be kept off in other instances, whatever they may be? As Anon mentioned as the case for some journals in the biological sciences, some journals in the Environmental/Earth sciences do the same right now.

a physicist said...

I agree, all co-authors should have access to the reviews -- at the very least, at the journal level! I think it's bizarre the editor wouldn't share the reviews with a co-author.

Jenn, PhD said...

Another biological scientist here - although with limited submission/review experience. But if I'm to be held responsible for the entire content of a manuscript that carries my name, then I want to have access to the entire manuscript, throughout the review process, including the reviews. I wouldn't expect to be automatically supplied with all the details (sometimes a collaboration is a very small part of the big manuscript), but if I specifically asked to see them, I would have a problem if that access was refused.

Aisling said...

Anonymous: I was going to make the exact same comment. In fact, I think it is good practice in general to collect all authors' email addresses. This system is used by a journal where a manuscript I contributed to was recently submitted, and upon submission by the corresponding author, an e-mail was sent to all co-authors to inform us that a manuscript we co-authored had been submitted on our behalf by *corresponding author*. This system would avoid manuscripts being submitted without co-authors' knowledge (unless fake e-mails were given as well...). As a corresponding author, I always forward e-mails from the journal to co-authors to keep them informed of the progress and status of a submission, so having the journal do it automatically would save me time :-) However, I have not been involved in any manuscript with more than a handful of co-authors. I have also been lucky enough to collaborate with colleagues who were at worst unresponsive when receiving reviews, so that no hateful discussion sparked and there were no obstacles to a resubmission where some co-authors had very minimal involvement. So basically, I'm unaware of any problem that could arise from sharing reviews with co-authors.

Rosie Redfield said...

Maybe the editor has gotten legal advice about this particular paper, and been told not to give out any information to anyone. The story about confidentiality may be just a misguided attempt to cover his (her?) ass.

EscapedWestOfTheBigMuddy said...

Particle physics here. Some papers have hundreds or thousands of co-authors, so it is impractical for all authors to be included in all correspondence. The practice is for the corresponding author to provide a summary ("...reviews were mostly positive, but some changes were requested to figure 3...") to the collaboration as a whole and details only to the committee designated to write this paper and anyone who requests them.

Certainly a single point of contact is in order for usual communication (submission, reviews, changes, etc) between the journal and the authors.

I don't, however, see where the journal comes off thinking they should decide which authors are more equal than the rest. In the case in hand, denying access to an "author" with concerns that they have been wronged would be particularly egregious.

Anne M. Archibald said...

I just had an article published in Science. Correspondence was generally just with me, though I did provide email addresses for all my (seventeen!) coauthors. Since I am pretty much the most junior one of the bunch, there was certainly no question of dealing with everything myself, and in any case I would have made sure to forward them the referee's report. The only time I didn't communicate everything to everyone was when I was trying to converge on a final version under a tight deadline - sending a draft out to everyone encourages everyone to find some minor comment, often changing something back to the way it used to be. So many revisions went out only to a short list of the most active coauthors. But of course the final version went to everyone, asking for an okay.

It seems to me that if the journal gets a request directly from one of the coauthors, once they've verified their identity, it's a sign something is seriously wrong: if the first author isn't communicating with the others, there's a serious question about whether the coauthors are actually willing to put their names to the paper. It seems like that should raise all sorts of red flags in the editor's mind...

maxine said...

One of the anonymouses above asks whether anyone knows of Nature's policies in this area. I am a (non-anonymous) editor at Nature, and have described our polices a few comments up. Perhaps these were still in moderation when anonymous wrote his/her question. But if anyone is interested, please scroll up or for more details send me an email at m.clarke@nature.com.

dave said...

On one recent (accepted) paper I (accidentally) swapped the order of two of the figures in the originally submitted version, so I almost edited out that bit of the referees' comments to hide my own incompetence.........

Anonymous said...

The arrangement Anonymous described @6/10/2009 12:52:00 AM makes sense to me.

Related: on a collaborative paper, the PI (a faculty peer) decided how to respond to reviews and did so before running his plan past me as a faculty level co-author. The same PI tried to keep me (and I assume others) from knowing which editor handled our submission. IMHO these steps made the PI look narrow.

Years have gone by and the paper still isn't accepted.

Doctor Pion said...

I cannot believe that any journal editor would refuse to send correspondence to a person who shares legal liability for the content of work published under his or her name.

I have not worked on the editorial side of the Physical Review, just the author side, but I cannot imagine that they (meaning one editor I know personally) would not send a copy of editorial correspondence to a co-author upon request.

In my one experience with something like this situation, where a coauthor submitted a paper with my name on it without my permission (containing nonsensical descriptions of the theory work I had helped do for his experiment), I had no difficulty getting the editor to send it back even though it made the guilty prof angry because he looked bad. But not as bad as I (and he) would have looked if that paper gotten out.

Ms.PhD said...

I've been left out of the review and revision process on more than one occasion, with different collaborators who never sent me the version they submitted. In one case the data that was supposed to be part of a main figure got moved to supplemental, but they never told me, and I didn't even know the paper came out until I saw it in Pubmed. The other paper has been revised and split into two manuscripts, and I still haven't seen the version that was originally rejected or the version that will be resubmitted. There is only so much I can do as a junior person and middle-author. If the journal requires them to provide everyone's email address, they must not have enforced it or done anything with the information.

I think the ethics of science publishing was always weakly outlined and is only getting worse. Something has to change!

maxine said...

I had a few rotten things done to me when I was a researcher, re authorship and papers, too. However, I think what some people are criticising here is the ethics of some scientists, rather than the journals. Journals aren't a police force, they are a publication vehicle, focused on assessing the research submitted to them. Authors are adults and if they all behave accordingly, there wouldn't be these problems! Of course, journals have authorship policies and should be prepared to enforce them, but they can't be held responsible for inter-coauthor squabbles and unethical behaviour.

Kevin said...

maxine wrote
Of course, journals have authorship policies and should be prepared to enforce them, but they can't be held responsible for inter-coauthor squabbles and unethical behaviour.

While publishers can't be responsible for inter-author behavior, they can and should adopt procedures that avoid fraud. In particular, they should inform all co-authors of any activity on the manuscript (submission, referee comments, re-submission, availability of proofs, publication). This gives the co-authors a chance to have their name removed from a paper that was submitted without their approval, at least. Since e-mail to 100 co-authors is no harder than e-mail to one, this form of automatic notification should be routine (and is for some journals).

Anonymous said...

I think that in the interest of saving time and resources on the part of the journal editors, that it is not realistic for the journal to send all co-authors copies of the reviews since some papers have many co-authors and if they all contacted the journal individually it would be a nightmare.

Thus I'm in favor of the "corresponding author" being the one that the journal sends reviews to. HOWEVER, I think that the corresponding author does have a moral and professional obligation to share the reviews with the co-authors. In my experience, not all corresponding authors will do this so it boils down to what the individual's personality, working style and agenda is.


I do, however, feel that if the corresponding author flat out refused to let the co-authors see the reviews, then the co-authors should have the right to get these from the journal. If your name is being attached to a piece of work, you should know what is going on before it is finalized forever. who knows, maybe a reviewer raised very legitimate concerns about YOUR part of the work...

If I was a co-author and the lead author refused to let me see the reviews, this is a red flag. I would e-mail the journal editor and request copies of the reviews, giving the explanation, and that if I cannot see the reviews I might consider retracting my name (and thus my contributions) from the manuscript.

Thomas said...

On what planet is this ok?

If I had unhelpful coauthors I might send them an email saying that I would send the reviews if they wanted, and hope that they didn't get around to replying. Actually withholding the reviews is outrageous.

I have encountered the scheme your first anonymous commenter described, and I like it.

Kevin said...

I just got a notification on a paper for which I am a co-author not corresponding author (it is not really a collaboration---the four groups are competitors, who were asked to summarize the state of the art):


Your revised manuscript ... has been successfully uploaded to the ... web site for editorial processing. The manuscript number is ... Please make note of this number. Your co-authors are blind-copied on this message.
You may keep track of the processing status for your manuscript by logging on periodically to your User Account at http:... where the status will be displayed in your Author Center.


This sort of automatic notification costs the journal nothing, and is very important in making sure that all co-authors agreed to have their names on the paper.