9

Two years ago, I was about three weeks (fortunately I keep notes) into a postdoc research position with a federal government organization, and I was excited to participate in a meeting with a federal researcher who was asking for help with ideas after a paper was rejected by journal reviewers for lacking any kind of validation of the results or approach.

During the meeting, the researcher presented the model, and I pointed out that rather than constructing an empirical distribution for P(y=1|x), the method was estimating P(x|y=1), which none of the internal or journal reviewers had noticed. I suggested two ways out: sort it out with Bayes' theorem or take an entirely different approach by creating a binary dataset with y taking on 0 or 1 instead of just 1, which was possible due to the sampling scheme.

The researcher took my second idea and quickly published a paper using that idea and included a long list of other folks in the author list and acknowledgments, but there is no mention of my contribution or name.

Now I am writing a paper about my own research that implements the same idea, but with substantial enhancements, and as I contemplate the literature review, I am extremely aware that I would be citing my own idea as the foundation for my approach, but with no obvious way to let the world know that it was my idea or that I contributed in any way.

A footnote explaining my role in the cited research seems weak because it would lack corroboration (unless I put the federal researcher's name on the author list [their only role was providing the dataset, which anyone can ask for and receive], a topic that is a whole other can of worms...).

Reading the responses here and here suggests that perhaps this is a more dramatic ethical issue than I expected (e.g., mentions of plagiarism).

I am also hoping to move on from being a postdoc, so I'm trying to accumulate recommendations so I can take the next steps in my career. I am working with the same federal researcher on another project (that they lead) and I get lots of credit for my ideas that have greatly modernized the approach and for teaching the team all about it. I have been expecting to rely on this researcher for glowing and detailed recommendations, which makes this a bit more difficult...

Broadly, this research is in the area of biostatistics and software development.

5
  • 7
    A footnote explaining my role in the cited research seems weak because it would lack corroboration --- Just an idea (without knowing the conventions of your field or the personalities of those involved), but perhaps you could include a footnote saying that the idea occurred to you in conversation with XX about YY problem, which XX was able to successfully use [or to give XX more credit (to lessen potential fallback): "XX found a way to successfully implement"] in ZZ paper, and the object of the present paper is to further develop this idea by applying it to show . . . Nov 17 '21 at 18:35
  • 14
    From what you've written (which may be incomplete), it honestly doesn't sound like you played a big role in terms of the science itself. (By which I mean, it doesn't sound like anybody would have any reason to remember you as, say, "the guy who suggested using Bayes's theorem for this problem".) So I would caution against accusations of plagiarism here unless you know there were others with equal or lesser contributions than you who nevertheless received credit, and you have reason to believe you were omitted deliberately. And even then, it might be too petty of a concern.
    – user541686
    Nov 18 '21 at 7:49
  • 3
    Since when do you get acknowledgement/authorship for pointing out a mistake in a paper ? Made in a one-off meeting on the topic ? If you complain about this, don't be surprised if people will stop talking to you. Back when I still did science, I had a wrong distribution fitted somewhere. Three other academics didn't see it either, the 4rth one pointed it out, everyone thought "that explains a lot" and went their merry way. I doubt the professor was still thinking about this by the time they left the building. You could ask them to see if they even remember and add that to your question. Nov 18 '21 at 13:11
  • Are you worried that that your methods will be viewed as plagiarism of the paper you contributed to without credit?
    – user121330
    Nov 18 '21 at 20:16
  • @Marianne013 An acknowledgement seems fine to me for pointing out a mistake. Not mandatory, but a reasonable gesture (and honestly I'd think less of an author that didn't add an acknowledgement for someone who helped them). Coauthorship seems excessive, unless the mistake completely invalidated the paper. Nov 19 '21 at 9:03
27

Getting input from fellow researchers is a common thing. There is nothing wrong with building on discussed ideas. This is one way how research progresses, as you have just experienced. Acknowledging that in some way is fair, of course.

But sometimes, it happens that you overestimate the input you gave. Authors of papers usually thought a lot about their work, so many ideas and approaches are already in their heads. Additionally, implementing an idea or adapting experiments and data analysis also takes a lot of effort most of the time. From the perspective of your discussion partners, your input then might be very small, too small to result in co-authorship. Butterfly effects like that happen all the time. I advise to exercise this change in perspective to answer your question yourself, as it is hard to tell from the distance.

Usually, when something as I describe above happens to me, I am actually quite happy that I was able to help. Well, occasionally only after an initial phase of disappointment, I admit. It also depends if you have reasons to suspect that your discussion partners have a tendency to treat people in an unfair way. Cooperation should be founded on mutual trust. If you trust these people, then you can assume that they acted with no malicious intent. And you might even be able to ask them about their perspective on your contribution.

17

No, I don't think you were unreasonably left out of the author list, assuming I'm reading the situation correctly from your description.

I have several times made comments of this magnitude on other people's papers regarding mistakes in their statistical analyses. Similarly, I have had reviewers of papers I have submitted that made comments of this magnitude and made corresponding adjustments in the manuscript. Other places that you might see this sort of thing happen would be in discussions at poster presentations or during/after a talk at a conference or speaking visit. These formal and informal opportunities to comment on someone else's work are not bids for authorship. None of these circumstances led to changes in authorship in my experience. If they take your feedback and improve their work, then that's just taking ordinary feedback as part of the scientific process.

In some rare cases, contributions like these might lead to authorship, but not from just the simple suggestion alone. Instead, the way it would happen would be for that one comment to lead to further involvement in the project. For example, if I were to suggest a different statistical analysis, and the current authors invited me to run the analysis myself, write the methods and results portions of the paper that involved that analysis, and to further review the remainder of the manuscript, I would certainly expect to be made an author.

Unless I'm missing something, it seems that all you've done is to identify an error. It may be a simple error or one ultimately resulting in a profound reinterpretation of the results, but still, it sounds like something of the magnitude that a peer reviewer should have found or that would have come up in one of the other settings I've mentioned.

2
  • 3
    Actually, I think you are wrong here. The error had been identified (no validation) and the paper rejected. The OP claims to have made two suggestions, one of which led to success. Providing "head slapping" insight is pretty big in some theoretical fields. I'll guess that our "action advice" to the OP would be about the same, though.
    – Buffy
    Nov 17 '21 at 20:34
  • 5
    @Buffy All I can do is speak from my own experience in situations that sound similar to me to what was described. It's certainly possible the insight was more profound than I am appreciating, and it would have probably have been appropriate to offer an acknowledgement in any case, but I think it's a bit of a leap to think this is so crucial to be worth authorship by itself. And yes, as you point out, not much productive to be done with it now, anyways.
    – Bryan Krause
    Nov 17 '21 at 20:42
9

It certainly seems like the paper has your "fingerprints" on it and that you made an important intellectual contribution. Yes, I'd call it plagiarism if you weren't at least acknowledged and probably an author.

My field is math/cs, however, where a few minutes of input can have major impacts. I don't know how similar things work in biostatistics, but for theoretical work, I'd think they are similar. Some people think that it is "effort" or "time spent" that makes you an author. But really it is who contributed the ideas. Especially the crux ideas. It feels like you qualify.

I can't, however, suggest a solution. Think long term, however. If this turns out to be a single "glitch" in your career you can probably leave it behind you. But future "collaboration" with such people isn't advised.

3
  • 17
    It sounds to me like OP made an improvement using a method that should be basic knowledge for a (bio)statistician. I think that in most cases more is needed to warrant authorship (for example actually doing the analysis for the coworker). If I got a penny (or authorship) every time I pointed out a mistake or made a suggestion to a co-worker in a group meeting I would be rich.
    – Louic
    Nov 17 '21 at 14:42
  • 5
    @Louic, certainly true for some interactions, but I read this one differently. Missed by reviewers, etc. Might have been the crux. It wasn't pointing out a mistake, but pointing to a solution. Quite different.
    – Buffy
    Nov 17 '21 at 14:49
  • 4
    It really doesn't sound like this was some kind of genius idea that the OP gave to call this plagiarism. It really sounds like many, many other researchers would have quite likely come up with the exact same "solutions" if they had a mixup between P(x | y) vs. P(y | x) pointed out to them. As such, claims of plagiarism seem kind of petty (is the next generation of scientist really going to look back and think you had some massive role here?) and might even reflect poorly on the person claiming it.
    – user541686
    Nov 18 '21 at 7:39
9

You could cite the paper you mentioned and add another reference "personal communication" with you as the first author and the federal author as the second one. This links you to idea, it makes clear there is no official record, and it does not claim authorship of the other paper.

Example

We test our hypothesis by creating a binary dataset
with y taking on 0 or 1 instead of just 1 as done in [1,2].

References
[1] Chappy Hickens, Federal Agent: *Personal communication, adding this idea to [2]*
[2] Federal Agent et al: *Different topic, using idea with y taking on 0 or 1*
1
  • 1
    That's probably what the Federal Agent should have done as well. Giving co-authorship to OP might have been a bit much.
    – glglgl
    Nov 20 '21 at 12:45
3

No. It does not appear to me that you were unreasonably excluded.

In addition to the excellent answers, it is also worth mentioning that different fields have very (very, very) different author inclusion expectations.

For example, in my experience, biologists tend to include several --perhaps too many-- authors. It is simply a matter of culture. Compare that with pure maths, which is the opposite.

Perhaps you were coming from a more biology background/culture? Whereas your collaborators were more statistics/comp sci, and never even thought about including you as a minor contributor.

1

It would have been prudent of the authors to have mentioned you in Acknowledgements but take this as a learning experience:

  1. Always consider inviting colleagues as co-authors, if they show interest and come with valuable inputs.
  2. Be aware whenever you are about to hit the "submit" button: Is there anyone I have forgotten to acknowledge?
  3. If you ever feel cheated for a co-authorship, let it pass but do blacklist the offenders; the tit-for-tat strategy from game theory works nicely in this domain.

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.