Trainee vs. Advisor: A tale of two publications

AKA Part III of “Who owns research data and the rights to publish?”

[Let's say it again...Nothing contained in this post should be construed as legal advice.]

This series of posts started out with a question: Is it legal for a student to publish their dissertation research without their advisor’s permission? In the last post we saw that the answer depends on many factors, such as the conditions under which the data were collected and whether the student holds copyright on the dissertation. But there are more than just legal considerations in situations like this. It’s about time we discussed the non-legal aspects
of deciding to publish. No more copyright law or legal jargon. Now to ask the trickier question: Should trainees (including students, postdocs, etc.) publish their research without the permission of their advisor? Here we examine two cases in which trainees did just that with very different outcomes.

Shmuel & Leopold vs. Logothetis

In May of 2008, Amir Shmuel and David A. Leopold published “Neuronal Correlates of Spontaneous Fluctuations…” in the journal Human Brain Mapping. Shmuel and Leopold were former trainees of Nikos Logothetis, a researcher at the Max Planck Institute for Biological Cybernetics and a big name in the field of neuroscience. Although the article clearly acknowledged that the data were collected in Logothetis’ laboratory and that funding came in part from grants awarded to him, Logothetis was not an author on the paper. Two months later, Logothetis went public in the form of a Nature News article to denounce the article’s publication, saying Shmuel and Leopold had published data belonging to his laboratory without his consent:

Logothetis is furious about the publication of data, which he believes will mislead the field, and about the fact that the authors of the paper allege that he tried to stop them publishing the data for personal reasons.

At first glance, this looked like a losing battle for Shmuel and Leopold. Here were two relatively unknown researchers going up against a giant in the field of neuroscience who had the backing of the Max Planck Institute and the ear of editors at Nature. But the tables quickly turned. Although some came out in favor of Logothetis, calling Shmuel’s and Leopold’s behavior “unethical”, comments posted on the Nature piece demonstrated broader disapproval of Logothetis’ actions:

It is sad to see that such prominent scientist refers to administrative means of settling scientific dispute.
-Vladimir Goncharov

I think it was wrong for Shmuel and Leopold…to publish the data without permission of the lab head where the data was collected. But this was handled very poorly by Logothetis….Try to squelch publication of a peer-reviewed paper and soil the careers of his own students is inappropriate and not good for anyone, including him.
-John Rico

…Logothetis could have played a better role to prevent all this nonsense. Scientific controversies should be tackled scientifically”
-Anand Prakash

If THE director is so furious, he should just write a better paper exposing the claims by his students.
-Ashish Asgekar

…Nikos Logothetis did not handle this well. I agree with others  that his best option would have been to publish his own article opinion, not to publically attack his own students.
-Walt Kurtz

And my personal favorite in the spirit of open science:

While it may not be politically savvy to publish data without the consent of a former supervisor, to say that data belongs to the PI is wrong-headed. Data belongs to the scientific community
-John Galt

The even more surprising support came when Human Brain Mapping published an editorial piece explaining why they decided to publish the article, and criticizing Logothetis’ behavior. They pulled no punches:

…Logothetis contacted the editors seeking to block publication. Concurrently, Logothetis pressured Shmuel and Leopold through other channels to withdraw the manuscript. When these efforts to block the manuscript failed, Logothetis made public statements discrediting the authors and the editors. In our opinion, the accusations against both the authors and the editors are not supported by the facts of the case; rather, Dr Logothetis’ conduct in this episode clearly fell short of widely accepted ethical standards.

The editors explain how Shmuel and Leopold followed appropriate procedures to obtain permission to use the data. The editors also argue that peer review was properly conducted and write:

Once the editors found no scientific grounds for withholding the manuscript…we were ethically bound to publish the manuscript.

The response from Max Planck following the publication of the editorial was swift and harsh. They sued. Again, Nature News covered the story:

The Max Planck Society (MPS) in Germany has begun legal proceedings against publishers Wiley International in a dispute over an editorial in the February issue of Human Brain Mapping.  The society alleges that the editorial grievously misrepresents it and harms the reputation of one of its scientists. It wants the journal to publish a letter from the society addressing these concerns without delay.

 The letter was published two months later and included a link to a longer article entitled “The Other Side of the Coin” [pdf] posted on Logothetis’ laboratory website rebutting the editorial.

The presentation of the unpleasant episode in the above-mentioned editorial contains serious factual flaws and omissions as well as pure speculations about the intentions of the different parties which discredit the people and institutions involved.

Alongside the letter, was published an editorial reply:

We note with regret that your reply did not address issues of ethical principles but focused only on perceived “factual flaws and omissions” (Jackle, 2009a). However, we are heartened that we agree completely on “the only point that is indisputable: Schmuel [sic] and Leopold had the right to publish” (Jackle, 2009b, pg 11).

Meanwhile, Logothetis published a rebuttal article in May of 2009 in the journal Neuroimage entitled “How not to study spontaneous activity”. To date, the article by Shmuel and Leopold has been cited 144 times according to Google Scholar. The rebuttal article by Logothetis has been cited 29 times. (As a side note, I’d love to hear from any neuroscientists who have read both papers. Which interpretation of the data do you buy?)

In sum, two trainees went up against their advisor, exercised their right to publish, and won.  Although it’s hard to judge the influence of a single paper on an investigator’s career, it’s likely that paper was a stepping stone for both of them.  Amir Shmuel and David A. Leopold are now professors directing their own laboratories, the former at McGill University and the latter at the National Institutes of Health. So, all’s well that ends well, right?

Except it doesn’t always end that well…

Liu & Xue vs. Noll**

In January of 2012, Wei Liu and Lei Xue published the article “Functional conservation of the Drosophila gooseberry gene…” in PLoS ONE. Both authors were former research trainees of Markus Noll, Professor of the Institute of Molecular Biology at the University of Zürich.
Although Noll was listed in the acknowledgements for his advisement, he was not listed as an author. Just six months after the article’s publication, PLoS issued this retraction notice. Unlike the Shmuel and Leopold case, this one flew largely under the radar; there were no press releases, no editorials, no lawsuits – just a quick retraction and done. From the details provided in the retraction notice, it seems that Noll contacted the editors after the article was
published to say that Liu and Xue had submitted the manuscript without his knowledge or permission. He also questioned the validity of some of the reported results. The editors then asked the authors of the article to produce the original data to substantiate the results. The retraction reads:

The authors have indicated that they do not have access to the raw data on which the results are based, as a result, the authors retract the article given that it is currently not possible to verify the validity of the results.

Written between the lines here is that Noll presumably did have access to the raw data, but refused to provide it for the purposes of validating the published results. Again, as far as I can find, this case received little reaction from the scientific community, save a comment posted by Mike Taylor (@MikeTaylor) in response to the retraction notice. Mike effectively dismantled the alleged reasons for the retraction one-by-one and I highly encourage others to read his entire response. I’ll just quote the one line which I think best sums things up:

All in all, this does not look pretty. It smells more like a personal conflict than a genuine matter of scientific misconduct.

To date, no response to Mike’s comment has been posted and the retractions stands. Determining the effects of the retraction on the authors’ careers is difficult, due in part to the recency of the events. As far as I can tell, the authors have not published anything in the 5 months since the retraction. Wei Liu is now a professor at Northwest A&F University.  However, I was not able to determine if, or where, Lei Xue holds a position. One thing is clear, though: Whenever someone searches for these authors’ publications on Pubmed or Google Scholar, that retraction notice rears its big, ugly head. I would imagine that has constituted, if not a road block, a significant hurdle for their careers.

Analysis and comparison of the two cases

In my opinion, the Shmuel and Leopold case is most interesting because of the response of the journal editors. They could easily have catered to a well-known researcher’s demands, pulled the paper, and likely avoided the publicity and lawsuit. Instead, they chose to take a
stand, defend their decision to publish the article, and call out what they viewed as unethical behavior. Their actions made all the difference in this case, and they should be commended, not necessarily for their support of Shmuel and Leopold, but for the benefit to the scientific community as a whole. Had the editors caved and pulled the article prior to publication, the only interpretation of the data on record would now be that of Logothetis. By refusing to pull the article – one they determined had been properly peer-reviewed – they made it possible for scientists to decide for themselves which interpretation is valid.

As for Logothetis, in my opinion, he was out of line. The majority of his criticisms were based on the interpretation of the data; a scientific issue that, as many pointed out, should have been addressed in a scientific arena, not taken to the level of institutional intervention or splashed all over Nature News. If he really believed Shmuel and Leopold were incorrect in their interpretations, let them fall on the sword. He could have gracefully declined their offer to be an author, thereby absolving himself of any responsibility for the ideas they presented, and published his own article explaining the scientific reasons for his concerns regarding their interpretations.  Of course, he eventually did this, but only as a last resort, preferring first to exhaust all attempts to strong-arm the students, and later the editors, into not publishing. I really do not understand this position. If he was confident in his analysis and interpretations of the data, it should not have been a threat that there was another report out there with conflicting results.

The Liu and Xue case involved similar behavior on the part of the Principal Investigator.  Like Logothetis, instead of settling the issue with healthy scientific discourse, Noll chose to try to shut the whole thing down and tarnish the reputations of his trainees in the process. Again, why was the publication of the article such a threat? If the students were really wrong and Noll had solid data to show this, as indicated in the retraction notice, he could have published a rebuttal article. This approach would likely have earned him not only commendation for the research, but respect for his dedication to the scientific process.

Unfortunately, Noll, unlike Logothetis, was successful in getting the article pulled and this seems to have had everything to do with the behavior of the editors. In this case, I find myself in the unusual position of criticizing PLoS. I am a huge fan of PLoS ONE, but I think the editors handled this situation badly. Mike’s comment really says it all, so I won’t rehash each point here. But in a nutshell, only one of the four stated ‘concerns’ could have justified a retraction: if the data really did not support the conclusions. It is not clear from the retraction notice if after the students said they did not have access to the raw data PLoS then asked Noll to provide the data for the purposes of verification. In my opinion, this should have been done. The article went through a process of peer-review and was determined to be scientifically sound.  Subsequent to this, the burden of proof should have been on anyone contesting this notion (i.e.  Noll) to show why the data or conclusions were not sound. The only way to do this would have been for Noll, being the only person with access to the raw data, to provide both data sets for comparison. If Noll was so confident in the second set of results, he should have had no issue complying with such a request for a proper investigation.  Instead, it looks as if the editors merely decided to take Noll’s word that the second data set was more reliable. I’ll be the first to apologize to PLoS if I’m wrong in my assumptions. But if there were additional actions taken by the editors, or more solid reasons for believing that Noll’s second data set was more reliable than the first, then that information should have been provided in the retraction notice so that others could make an informed judgement of the retraction’s legitimacy. Unless and until there is more information forthcoming on this case, I’m going to have to agree with Mike that this really does look like a personal conflict.

Having said all that, I think Liu and Xue also did not handle the situation as well as they could have. If they had been transparent from the beginning, they might have found more allies. Since the work in question was part of one of the author’s dissertations, it is possible the student had copyright and had the legal right to publish even if Noll did not give consent. The best approach would have been to invite Noll to be an author. Then, if he refused, they could have proceeded to submission with all the necessary documentation in hand to show that they attempted to involve Noll. Second, although the trainees were surely prevented from taking the original data with them when they left the lab, they were likely permitted to take copies of the data (especially since it was dissertation work) and should have done so if they intended to publish.  This way, they could have performed the necessary analyses requested by PLoS to verify the results. If they were not permitted to take copies, something is very wrong about that in itself, but the decision to publish was questionable given that they should have realized at some point someone might ask for additional analyses or substantiation of their results.

What have we learned? And what does all this have to do with open science?

There are no right answers here, but these cases teach us a few valuable lessons. First, if you’re going to publish without your advisor’s permission, make sure you cross all your t’s and dot your i’s: (1) Establish that you have the legal right to publish, (2) Provide a draft of the manuscript to your advisor and offer he/she the chance to make their own edits and be an author on the paper, (3) If they refuse, notify them of your intent to publish alone, (4) Save all correspondence, and (5) Be forthcoming to the editors handling your submission. Second, be aware that deciding to publish without your advisor’s consent can be a risky game. It’s a gamble. Sometimes you get lucky and fall into the hands of editors like those at Human Brain Mapping, who are willing to support you. Sometimes you don’t. It’s a decision you have to make with your eyes open.

But here’s my take. If you feel you have something with scientific value, you should do everything in your power to get it out there. Fight. Because this is not about the potential career benefits to you; those, if they happen, are secondary. This is about the fact that you presumably got into science to learn more and to share that knowledge with others. It’s about being open. That beautiful piece of science you have there does no one any good sitting on your hard drive.  And the sooner everyone in the scientific community, including faculty advisors, recognize that, the better off science as a whole will be.

Update 04/12/2013: In December of 2012, PLoS editors posted a reply to my comment on the retraction, saying they “would like to clarify several aspects in relation to the decision to retract this publication”. The editors say that concerns about the article prompted an institutional investigation, during which the authors themselves requested that the article be retracted. The university later made the same request after concluding their investigation. The editors write, “In keeping with PLOS’ commitment to transparency, these details should have been included in the retraction notice”.

References

Shmuel, A. and Leopold D.A. (2008). Neuronal correlates of spontaneous fluctuations in fMRI signals in monkey visual cortex: Implications for functional connectivity at rest. Human Brain Mapping, 29: 751-761.

Liu W, Xue L (2012). Functional conservation of the Drosophila gooseberry gene and its evolutionary alleles. PLoS ONE: e30980. doi:10.1371/journal.pone.0030980.

*Thanks to @MnkyMnd for bringing the Shmuel & Leopold case to my attention.
**Thanks to @CaseyBergman for bringing the Liu & Xue case to my attention.

About these ads

8 thoughts on “Trainee vs. Advisor: A tale of two publications”

  1. I agree 100% with your conclusions.

    Only one point of conflict here: “The best approach would have been to invite Noll to be an author”. On what grounds? Is there any evidence that he made a scientific contribution worthy of authorship? Merely being someone’s supervisor does not constitute such a contribution.

    1. Merely being someone’s supervisor does not constitute such a contribution.

      I couldn’t agree more. I object to the all-too-common practice of tacking the advisor’s name on the end of every paper. There should be a significant intellectual and sustained contribution to warrant authorship. I guess I was giving Noll the benefit of the doubt in this case that he made such a contribution. But even if he didn’t, I think it would have been the politically smart move in this case for the trainees to offer him authorship. Unfortunately, the reality is that many in the life sciences expect to see the advisor’s name there and I think the editors recoiled in part because authorship wasn’t even offered. It made it look as if the trainees were intentionally trying to hide the existence of the publication from the PI. Again, I don’t think it’s right, but it seems to me a simple cost-benefit analysis. Offering Noll authorship would have been a small price to pay to ensure that the science was published (and not later retracted). Ultimately, I hope this convention regarding advisor authorship changes and trainees do not have to make the politically smart decision, but only the scientifically correct decision, in the future.

      Aside from the debate over authorship, I think the key lesson here is that transparency matters. If the trainees really felt strongly that Noll’s contribution did not warrant authorship, they should still have notified him of their intent to publish. I think the editors would have been more inclined to support them had they done this.

  2. This is ridiculous. The data, lab notebooks, etc belong to the lab and technically to the University, not to the investigators/trainees. Especially if said data resulted from a federally funded project. Working in a lab it is truly hard to say who contributed intellectually, since the PI generates an environment in which is conducive to new ideas and provides the resources to put this ideas to the test. If the project relied upon these resources awarded to the PI’s lab, then the PI should be an author in the paper. Period. Ultimately the PI is accountable for research that emerges from their lab, and thus should be included in all publications. Any trainee that believes their contributions to their lab while being a trainee are truly independent and belong to them, is truly deluding themselves.

    In the above cases it seems like the case stems from instances where the PI and the trainees disagreed on the data, and the PI refused to publish. In this case, there are other recourses available to the trainees to file grievances, but they still do not own the data. If their data so strongly disagrees with that of the PI’s, then they can always replicate it in their labs as independent investigators.

    1. The PI generates an environment in which is conducive to new ideas and provides the resources to put this ideas to the test. If the project relied upon these resources awarded to the PI’s lab, then the PI should be an author in the paper. Period.

      This is completely alien to me. In palaeontology you don’t get an authorship for creating an atmosphere, you get it for writing part of the paper. (I published seven papers while doing my Ph.D. My advisor is not a co-author on any one of them, and never so much as hinted that he should be.)

      1. Maybe it’s field dependent, but in life sciences it is basically impossible for a trainee to work in an advisor’s lab without drawing upon resources and funding provided by the advisor. Many of these resources and techniques were developed over the years in the advisor’s lab and are an integral part of the lab’s line of research. I wrote 5 papers from my PhD dissertation and my advisor was in all of them, as well other people in the lab who contributed, as they well should have been.

      2. OK. Evidently the moral here is that expectations of advisor authorship vary between fields, and academic editors who work in one field should be careful about making unwarranted assumptions regarding the field of the paper they’re handling.

    2. The data, lab notebooks, etc belong to the lab and technically to the University, not to the investigators/trainees. Especially if said data resulted from a federally funded project.

      In most cases, this is correct. But this is a separate issue. Funding and ownership of the data does not translate directly into authorship on a scientific paper. If it did, we would list the University (or even the funding agency) as an author, which I’m sure we all agree really would be ridiculous. Second, the trainees were not claiming ownership of the data itself. They did not release whole data sets to the public or remove any data from the lab. This particular case involved publishing part of the author’s dissertation, to which the student likely held copyright. Thus, the student was really exercising ownership rights over the contents of the dissertation.

      …the PI generates an environment in which is conducive to new ideas

      This is not always true. In fact, I would say that since the PIs in these cases tried to block trainees from publishing data that disagreed with their interpretations, it is indicative that the lab environment was not at all conducive to new ideas.

      Any trainee that believes their contributions to their lab while being a trainee are truly independent and belong to them, is truly deluding themselves.

      I agree that it’s rarely the case that contributions of a trainee are completely independent; otherwise they wouldn’t be trainees. However, I think there are cases in which the trainee is the major driving force behind a project and the PI plays a very minor role. See Hoffman v. Lemaitre for an example of a case in which the trainee claimed just that (and has fairly good documentation on his website to back it up, I might add).

      If their data so strongly disagrees with that of the PI’s, then they can always replicate it in their labs as independent investigators.

      Here’s the problem. If the trainee is blocked from publishing their work by the PI, they may have difficulty getting a faculty position with their own lab, let alone the funding to support replicating the experiments. How are they supposed to demonstrate to others that they can perform the work if the publications aren’t there and they can’t even use the data as supporting evidence in a grant application? The actions of the PI can effectively cut things off at the pass.

      1. A note on Lemaitre’s website linked in my last reply: The site previously included documentation about the activities and contributions in the lab during the years the Nobel-prize winning work was conducted. The site appears to be down, but I will leave the link up in the hope that the site will be restored.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s