3

For my master thesis in clinical medicine I have collected and analyzed data for a retrospective cohort study. Now one year later, a research assistant of the group is about to submit a manuscript based on my master thesis and I have been offered shared first co-authorship with said research assistant. Upon internal review I have noticed the following issues:

  • Definitions of clinical outcomes of the study have been changed such that they are now not well defined and not even of clinical interest
  • The statistical analysis I had performed for my thesis was completely ignored and omitted in the manuscript. For the manuscript, the research assistant employed statistical models I deem inappropriate as confirmed by fellow researchers external to the group. In addition, reporting of methods and results is poorly written with critical information missing.

I have provided feedback concerning the statistical methods but it was ignored. My advisor refused my request to see the new statistical analysis (in particular, I asked for the statistical software files edited by the research assistant), telling me that even though I have shared co-first authorship for the paper, I am not entitled to any insight in the statistical analysis, since the research assistant is in charge of it. I’m having a hard time accepting publication of a study of which I’m co-first author but haven’t seen the analysis behind it, especially since I highly suspect it is inappropriate. My advisor threatened to remove me from the paper completely if I don’t agree blindly.

Shall I renounce and let it be published without my name? Shall I agree and accept to have a paper published with my name and the knowledge that the quality might be really bad? Shall I contact the editors of the journal they want to submit it to?

Edit: Thank you all so much for your thoughtful answers and comments! I appreciate your help!

  • 2
    When you say "your advisor", do you mean the professor that was your advisor at the time, or is this person still your current advisor now? – Ian Sudbery Jul 7 at 13:17
  • If the reviewers act with due diligence, they will want to see the files as well. And if the methodology is as flawed as you suggest, the paper isn't likely to get published anyway. This might ease the decision not to be included as co-author in the prospective publication. – henning -- reinstate Monica Jul 8 at 11:33
  • 1
    What you describe sounds terrible. I'd go as far as saying that your advisor is engaging in academic misconduct. – user2705196 Jul 8 at 11:49
  • @henning Reviews are often pretty rushed and I have never had an experience of a reviewer asking for analysis files. The recently retracted Lancet hydroxychloroquine paper shows how even the most prestigious journals will accept papers without corroborating any of the primary data or analysis. – D Greenwood Jul 8 at 15:08
  • @DGreenwood unfortunately, reviewers aren't always diligent and the review process can be quite random, that's true. – henning -- reinstate Monica Jul 8 at 15:27
7

If you are to co-author a paper, you need to agree with its methodology and conclusions. You don't.

My advice, in general, is to offer to work with others to improve the paper to meet everyone's standards. You indicate that this might not be possible, which would be a fault of the PI. If you can't convince them to work with you, then I don't see a real way forward without saying that you won't be part of it in any way.

They can then go forward if they like, citing your past work properly. If they don't do that, it would be time to contact editors.

| improve this answer | |
  • A difficulty might arise in that master's theses are not generally published, so would they be not be citable. In theory a student can withhold permission to use unpublished data from their advisor. In practice we all know how hard that is to enforce. – Ian Sudbery Jul 7 at 13:18
  • 6
    @IanSudbery, actually unpublished work is citable and the failure to do so is still potentially plagiarism. "J. Smith. Baskets in Ten Lessons, unpublished MS thesis, Harvard College, 2001". – Buffy Jul 7 at 13:58
  • Incredibly difficult to enforce though. I have to say I have never in my life seen a reference to an unpublished MS thesis in a peer reviewed article. – Ian Sudbery Jul 7 at 16:49
  • 1
    A master thesis should be an official source. It is somehow published by its nature. At least this apply in my country and the others I've worked in. @Ian Sudbery – Alchimista Jul 8 at 11:27
  • Master's thesis' here are sent to the examiner, marked, moderated, viva'd, kept for a couple of years in case of appeal against the mark and then disposed of. – Ian Sudbery Jul 8 at 12:45
7

For me, the most telling phrase is this one:

My advisor refused my request to see the new statistical analysis (in particular, I asked for the statistical software files edited by the research assistant), telling me that even though I have shared co-first authorship for the paper, I am not entitled to any insight in the statistical analysis, since the research assistant is in charge of it.

If you are an author of a paper, you should be allowed access to all aspects of it. Your advisor is completely out of line in telling you that you cannot get insight in the statistical analysis performed. This is a huge red flag for me, and I would suggest that you withdraw your name from the author list: if your co-authors won't allow you to even look at the statistical analysis, I wouldn't want to take responsibility for their actions by putting my name at the top of the paper.

| improve this answer | |
-2

You have to decide between not being an author and keeping yourself away from the paper or being an author with all the benefits and, possibly consequences, that come with it.

Will it be worth the authorship or not - you need to decide, but given the way you express your concerns I feel you are leaning to not being an author.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.