About these ads

Retraction Watch

Tracking retractions as a window into the scientific process

Who deserves to be an author on a scientific paper?

with 29 comments

labtimes 2-2013Although authorship issues are not the most common reason we see for retractions, they’re one of the most vexing. We’ve seen multiple cases in which papers are retracted because colleagues say authors didn’t have a right to publish data, for example. In other cases, authors who didn’t know about a paper are surprised when it comes out.

So for our most recent column in LabTimes, we decided to look at these situations and try to answer some questions:

What’s the best way to acknowledge the work of colleagues who might have helped provide intellectual background for a particular study but who did not participate in the collection of the data or the preparation of the manuscript? And what of the lab head, whose ability to bring in money keeps the lights on and the rats fed?

We discuss inappropriate authors, excluded authors, and even ghost authors, drawing on guidelines from publishers, COPE, and the ICMJE. Have a look, and let us know what you think.

About these ads

Written by Ivan Oransky

March 27, 2013 at 8:30 am

29 Responses

Subscribe to comments with RSS.

  1. I always follow the authorship rule passed on to me by a great postdoc in the lab where I did my Ph.D: If you’ve done any of the actual experiments, written any of the paper, or intellectully contributed to coming up with the experimental design, then you’re on my paper. Otherwise, you’ll fall into the acknowledgements or gift basket category.

    mmmbitesizescience

    March 27, 2013 at 8:42 am

    • That works well for experimental science.
      What about this common situation from epidemiology? A group of scientists spends years collecting thousands of people for observation in a cohort. A large amount of data is collected — questionnaire based and also biological specimens.
      Later those specimens and data are analyzed by other investigators who were in no way involved in the original cohort assembly or original data collection. Analyses could include methods that did not exist when the original cohort was assembled. (Such data and specimen sharing is in fact essentially mandatory.)

      Does designing the original cohort count as “designing the experiment”?
      Or is it merely “data collection” that does not justify authorship (ICJME)? (Come to think of it, what do they mean by ” data collection…alone” anyway…?). Seems a bit harsh considering the years and years of effort.

      And it gets worse when such a study is conducted over multiple sites (e.g., clinics) around the world. Potentially over decades. Must each paper track down the original PI or head of each site for authorship on each paper? And etc.
      In the end I would say, No. It puts a big burden on junior investigators (to find everyone and get endless rounds of revision) and is a disproportionately large authorship gift to the original investigators.
      But my answer is by no means standardized across epidemiology or clinical science.

      S

      March 27, 2013 at 1:17 pm

      • “If you’ve done any of the actual experiments, written any of the paper, or intellectully contributed to coming up with the experimental design, then you’re on my paper.” – these three conditions should be met concurrently, so the “or” should read “and”. Besides, the “any” bit would make it possible to abuse the system as even the most miniscule contribution (e.g. writing up a sentence, a comment made during a coffee break, turning on an instrument) could be construed as sufficient. This, in fact, is what happens most often when PIs rob (for the lack of a better word) the deserving scientists by giving the co-authorship to random people. This is a zero-sum game – rewarding the co-authorship to the undeserving, diminishes the credit the others are due.

        chirality

        March 28, 2013 at 3:38 am

        • “This is a zero-sum game – rewarding the co-authorship to the undeserving, diminishes the credit the others are due.”
          The quality of authorship is not strain’d,
          It droppeth as the gentle rain from heaven
          Upon the place beneath: it is twice blest;
          It blesseth him that gives and him that takes:

          littlegreyrabbit

          March 28, 2013 at 4:52 am

        • I have to disagree with you, chirality: if we followed your ‘and’ rule, most PI’s wouldn’t be on their own paper. In my experience, they very rarely do the experiments or write the paper, but they do contribute hugely to the experimental design. I firmly believe that when it comes to authorship, a peppering of common sense and the ability to pipe up when you see someone unfairly included/excluded is key.

          mmmbitesizescience

          March 28, 2013 at 10:28 am

          • My problem is who should be the FIRST AUTHOR in a scientific paper wherein you are the one who has done the entire writing? Even if you are in a group, the one that has contributed a lot in the paper should be given the most credit: the one who had conceptualized and who’s so eager to run and test the experiment, the one who had interpreted the results, the one who had read a lot of scientific papers to come up with a good paper and the one that finally written almost 90% of the paper. Just common sense can deal with such a problem and no further discussions. But is it really a big deal to be the first author when paper is presented internationally?

            Edna Mae Cruz

            August 12, 2013 at 9:16 pm

      • I am actually going through this issue currently, I have been working on data collections – I had collected almost 30,000 records for public health research, cleaned it, processed and made it all pretty for the analysis stage. Yet when it came to analysis another bunch of investigators took over. Now, after analysis and at the stage of publishing I have been told i’m just going to be in the acknowledgements (I personally think this is unfair), and yet all those who did the analysis and even just PROOF READ are being put as authors.
        Do i have a right to be an author on the paper? I have put in so many hours, days, and even years into the collection (i was even told that 80% of the work in the data collection), isn’t it unfair to just be put in the acknowledgment? But i’m am very confused about whether i have the rights to ask to put as an author….please help!!!!

        Malariageek

        July 18, 2013 at 3:15 am

  2. Authorship disputes are one of the most common problems editors see. Last May the International Workshop on Contributorship and Scholarly Attribution identified some of the issues with current authorship conventions and attribution in a world of increasing diversity of research outputs. The meeting report summarises them and some of the possible solutions that were discussed (declaration: I wrote the report). It also includes measures that might help research groups and those involved in collaborative projects avoid or minimize authorship misunderstandings and disputes (page 6):

    (i) Have a clear authorship/contributorship policy.
    (ii)Discuss and document individual contributor roles and provisional authorship early on,
    ideally at the start of the project before work begins.
    (iii)Review contributions as the work progresses, revise roles and authorship accordingly until journal submission.
    (iv)Keep a descriptive authorship contribution list.
    (v)Document the reasons for author/contributor additions and deletions, and get agreement for changes from all individuals.
    (vi)Make sure all authors/contributors see and approve the final manuscript

    Report available at http://projects.iq.harvard.edu/attribution_workshop and http://figshare.com/articles/Report_on_the_International_Workshop_on_Contributorship_and_Scholarly_Attribution,_May_16,_2012/96831

    Re why authors make up co-authors – I’ve heard it’s sometimes to have someone to blame when things go wrong.

    Irene Hames

    March 27, 2013 at 9:32 am

    • “Review contributions as the work progresses, revise roles and authorship accordingly until journal submission”
      This is one of the sticky ones. On my last paper as corresponding author, I had to remove two people from the author list when their data was written out of the manuscript. I am hoping that the data will ultimately be published later, but it felt like I was kicking them in the gut. It was also no fun getting approvals on revisions and re-revisions from multiple co-authors at different locations, but was absolutely necessary.

      stpnrazr

      March 27, 2013 at 9:46 am

  3. In addition to the Harvard workshop report that Irene beat me to posting, this short one from Research Trends: http://www.researchtrends.com/issue-31-november-2012/fixing-authorship-towards-a-practical-model-of-contributorship/

    Dorothea Salo

    March 27, 2013 at 10:01 am

  4. My view is that the problem with who should be listed as author has been successfully solved. And, it’s not that the correct solution was unknown 30 years ago. Simply, in these 30 years, science and scientists underwent the “change”, and what was known became unknown, and the rules became needed.

    So, the rules are now written, polished, published, discussed at the retreats and republished. A much more important problem, however, remains: the very people who wrote the rules and currently are in charge of applying the rules, are able and willing to pervert the rules, although they do not do this publicly…

    Now, I will speak about the Committee on Publication Ethics (the COPE, quoted here and everywhere else as a role model of integrity). I more than once tried to draw attention here, at the RW, to my exchange of emails with COPE: http://www.universitytorontofraud.com/committee.htm I supposed that not only the COPE rules should be studied, but also their actual practice that violently contravenes their own rules and all other rules.

    Yet, I never noticed any interest in studying this email exchange on the part of the learned public. My question now: Does scientific community has any interest whatsoever in the actual impact of COPE on science, scientific publications, retractions, etc.?

    pyshnov

    March 27, 2013 at 12:59 pm

  5. I don’t worry about a definitive rule, because I think there are numerous and reasonable variations in institutional practice. My standard: If you’re a co-author on my first-authored paper, I will ask you for advice and comments all the way through the process. If you have more expertise on some aspect of the paper than I do, I’ll ask even more of you. And, if you make me a co-author on your paper, I’ll ask up front how much help you need from me, and will always make time to do what’s needed to get the paper done. (Short of writing it for you, of course!)

    drjuliebug

    March 27, 2013 at 1:17 pm

  6. Then there is in-joke authors:

    http://dx.doi.org/10.1016/0550-3213(79)90017-8

    To understand this one, you should know that the physics institute in Munich has office labels on everybodies office door. However, a cable funnel door (Kabelschacht in German) has exactly the same label. This particular author has already five papers…

    More seriously, there is the “alpha-beta-gamma” paper:

    http://en.wikipedia.org/wiki/Alpher%E2%80%93Bethe%E2%80%93Gamow_paper

    Rutger Boels

    March 28, 2013 at 4:08 am

  7. First authorship, order of authors, guest authorship, and what effort needs to be invested in a project are all very contentious and highly heterogeneous between groups. I’ve worked with many teams in the past, and I’ve had my share to heated discussions on these topics. Some individuals feel that if they were asked to be part of a research team, but didn’t actually contribute anything, that they still deserve authorship. Others feel that the only ‘real authors’ on a paper are the first, second and last authors and therefore you can ‘Christmas tree’ the author list with as many additional authors as you want (especially guest authors) and that this practice is both ‘in accordance with current norms and expected’. Some have even warned of retaliation by people who are not showered with authorship on papers they actually know nothing about, or just read the final manuscript before submitting it to a journal.

    In the end, it’s a messed up tradition, but I don’t know of any better models out there.

    Ahmed

    P.S. Unfortunately, when I try to use logic and evidence from documents like ICJME and COPE, people laugh at me and act like I am crazy. They say that no one actually cares what they documents say and that if you don’t ‘grease the scientific wheels’ then you will never grow in a University setting. I pray that they are wrong, but as the days go by, as I see the fruits of my hard labor is being served as gifts to others who don’t deserve it, and as my scientific credit is continuously violated by others who have power over me, I wonder whether it’s worth a University setting is the best place for me.

    Ahmed M. Abou-Setta, M.D.

    March 28, 2013 at 4:11 am

  8. Even sticking to the comical “To be an author on a paper, you should have at least read it” would probably prune pubmed sigificantly.

    Charlie Bond

    March 28, 2013 at 4:33 am

  9. The problem is now very serious in developing countries (including China, Taiwan, etc.). The salary and promotion of professors are based on number of papers and IF (impact factor). Some professors push his/her graduate students to publish papers and ask to be the first author or the corresponding author. But actually they contribute nothings in science. They offer only scholarship, space, equipment, etc.

    I have caught quite a few cases of plagiarism and fabrication. (Fabrication is very difficult to be sued, because authors they refused to show their experimental records. Other professors want always play as “good guy”.)

    I believe all scientific journals should have strict and rigid rule on “corresponding author”. The corresponding author should be always the one who submit the manuscript and communicate with the editor. (But this does work sometimes. As I know that some professors let their graduate students to use professor’s e-mail to submit manuscripts. The bad guys can always find bypasses.)

    It seems we can never find a way for a 100% prevention of unethical and immoral behavior. But at least we should do our best to prevent people doing that.

    Prof. Dr. Hsin Chi

    March 28, 2013 at 5:22 am

    • Completely agree with Dr Chi. This is a serious issue also in Brazil.

      Eduardo G P Fox

      March 28, 2013 at 10:58 am

    • Yes, I totally agree with you. I was shocked when I saw an immoral behavior at a famous institute in Japan: A “famous” boss pushed the graduate students and postdocs to include his name as the corresponding author even if he is just a host profesor while the main scientific contribution to the paper or communacation with the editors are done by the other “co-author”.

      W. Yang

      March 28, 2013 at 12:02 pm

      • This is a real problem in both national and international labs right now, and it’s something I really do not like seeing. I know exactly who did what on all of my peer reviewed publications on which I am the correspondent author, and I am honest about my student’s work in my cover letters to the journals.

        We are Working On It. Scientific publishing ought to be about clear communication of discoveries and, when it’s a university lab, a testament to the work the student did. (My PhD Boss, Peter Tonge, made me re-write the introduction several times in my 1st author JACS before he was satisfied with my language!) But we have a lot on our plates right now so give us (and, more importantly, our Editors) a bit of time.

        Dr. Allison L. Stelling

        March 28, 2013 at 12:34 pm

  10. Also, the biology and medical folks need to remember that math and physics (well, theoretical physics) are not team sports. If you cannot do the math & picture rendering in your head, you’re probably not an Author.

    Dr. Allison L. Stelling

    March 28, 2013 at 1:32 pm

  11. This article and these discussions still fail to address the very important issue of data ownership versus paper authorship that was at the very heart of the original dispute. While paying the peons doesn’t merit authorship, per se, it does grant ownership of the data. Therefore, a peon cannot publish data without the consent of the data owner. That doesn’t mean that someone who only owned data can claim authorship; however, it does mean that the data owner could choose to disallow its publishing OR to insist that they contribute to the manuscript in a meaningful way so as to merit authorship. Further, said data owner, if they want to be a bad boss, could choose to write the paper entirely on their own and merit sole authorship (as simply collecting data is not sufficient for authorship under ICMJE standards (however, said peon would have to be listed as an author for Elsiver journals).

    QAQ

    March 30, 2013 at 5:09 pm

  12. I will be pursuing an authorship and plagiarism dispute for a study I was involved in. I identified a problem in health care as an ordinary bedside nurse, spent years trying to get someone to do something, finally brought the problem to an educator/researcher who took me seriously, worked side by side with her developing the auditing took and collecting the data. Then when it came time to publish, she started publishing and presenting our startling results without my knowledge or name on it, complete with millions of dollars in cost savings, I couldn’t be a part of the manuscript writing because she did it without my knowledge. I know find myself in this world of science as a dirty business.

    plagiarized

    April 2, 2013 at 12:45 pm

    • Sorry for the typos above, I got a little passionate : )

      plagiarized

      April 2, 2013 at 12:46 pm

  13. I find myself today a surprised co-author! (I am not the surprised Laura of the post about chemistry, by the way). The work isn’t bad, but there are mistakes and wrong references in ‘my’ bit of the paper. Any advice?

    Surprised of Norway

    April 3, 2013 at 4:29 am

    • I recommend you don’t worry about it. If you had seen the paper earlier you would have spent a lot of time trying to improve it. As it is, since you are not the first author, no-one will blame you for it. Maybe this view is a little cynical, but mostly you can’t fix everything.

      Michael Kovari

      April 6, 2013 at 6:36 pm

    • I recommend you ask the corresponding author to correct this. And if he does not, contact the Editor of the journal.
      In contrast to Michael Kovari’s view, you *are* a co-author on the paper, and if other scientists know you are the expert in that section of the paper in which there are mistakes, they *will* blame you, since they *expect* you to have written that part!

      Marco

      April 7, 2013 at 12:41 am

    • If it’s still a manuscript that has not yet been submitted, you should have the mistakes corrected. As Marco says, if it is published your name will appear in the list of authors so you will be held publicly responsible and accountable for errors. If any of the errors in the references involve publications by any of your peer reviewers, they will not be impressed.

      If the article has already been published, you can ask the other authors to agree that the journal should be asked to publish a correction. This is embarrassing for all coauthors and creates extra work for the editors and publisher, but it’s best to correct the record.

      If the manuscript has been submitted but not yet accepted (or rejected), the right thing to do is to contact the editor to say that mistakes have been noticed and will be corrected in the revised manuscript (assuming you are offered a chance to revise and resubmit). This “honesty is the best policy” decision, however, may be uncomfortable for (or even rejected by) the other coauthors and may bias the editor against the manuscript since he or she may suspect something worse than honest error.

      If the journal’s instructions or guidelines for authors state clearly that submittal of a manuscript will be understood to mean that all authors have seen and approved it, this experience may be a useful lesson for all coauthors. Editors usually react badly to news that the conditions for manuscripts submittal were willingly and knowingly violated.

      Karen Shashok

      April 7, 2013 at 6:35 am

      • Thank you to all of you for your advice – nice to see a balance, too! The article in question is already published. I discussed it with my boss and considered the comments here. In the end I wrote a rather cross email to the lead author, pointing out that they signed something that said that all coauthors had seen it. The mistakes are so very minor that it does not change the paper outcomes in any way – just out-of-date/not very useful references. The lead author apologised, and provided a ‘paper trail’ to show they had tried to contact me. So it’s just down to an over-enthusiastic mail filter at my old workplace. Shows how sometimes, authorship issues aren’t really anyone’s ‘fault’ but still happen.

        PS to ‘plagiarized’ the nurse, I’m not surprised you are cross enough to have made typos! But please believe me that not everyone in science ‘plays dirty’ – most of us do our very best to act in an ethical manner. I hope you get a good resolution to your case,sounds like you have been ripped off appallingly.

        Surprised of Norway

        April 9, 2013 at 11:44 am


We welcome comments. Please read our comments policy at http://retractionwatch.wordpress.com/the-retraction-watch-faq/ and leave your comment below.

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

Follow

Get every new post delivered to your Inbox.

Join 34,865 other followers

%d bloggers like this: