Statistician as co-author

The first rule of thumb is never do work without clarifying authorship. (This is usually harder said than done, especially if one is a graduate student).

Based on the limited information you have provided, it seems that authorship may be warranted here. If your analysis was used, especially if your "explanations of how to interpret the results" were used verbatim, you need to be given authorship. If you are in a position to do such, I would send a short email to the professor and explain that you feel your contributions deserve authorship. It costs the professor nothing to add you as an author. Since you are outside academia, you have the option of not working with the professor further. Regardless of the outcome, you can still list the project as a project you participated in as a statistician.

This will be somewhat field specific. Such practices could be confirmed by looking at other similar papers and seeing what type of attribution the statistician was given. This will also back your claim if statisticians are usually given authorship in the field.

Also note that a poster is not the same as a journal article. Authorship for posters is usually somewhat loose; authorship for journal articles is much more official.


Let me chime in with my experience. I (mostly) left academia a couple of years ago to start a business, but I'm still doing research. The trouble you describe is not unknown to me. Most of it are well-known conflict lines in inner-academic collaborations as well, but there are a few points that are peculiar to freelancers/consultants/contractors from outside academia:

I had to learn that there are peculiar misconceptions about (scientific) authorship. In particular I've met (a few/some) academics who

  • seem to think that authorship in a publication is bound to holding a position at an academic research institution (seems to correlate with the world-view that valid research is exclusively conducted at academic research institutions), or
  • that being paid on contract basis somehow buys off the authorship*.
  • Or that authorship is a form of payment that can be used in lieu of money*.
  • And I've had people arguing against me being co-author that being industry/business/freelancing I don't need publications any more.

Needless to say, all these points are of no relevance at all for deciding scientific authorship. BTW, I'm in a legislation where authorship is an inalienable right - I could sell the right to economic exploitation (just like academics/their employers do all the time when tranfering rights to the journal publisher; I do that, too: I license my work to the client), but not authorship.

In consequence, whenever I smell such authorship trouble lurking around the corner/whenever the contracted work is of a nature that is likely to end up in a publication, I put into my offer that any publications coming off of the work in question will handle authorship according to the DFG priciples for Safeguarding Good Scientific Practice** This is redundant in the sense that the academic research institutions I'm dealing with are members of the DFG and have agreed to these rules. But by reiterating it in the contract I try to cut short those nonsensical "arguments" above and I emphasize from the beginning that I do care about authorship.

In a way we freelancers/contractors/consultants do have an advantage here over, say, academics in the stats department who do statistics consulting: as the normal way of cooperating with us is a written contract, we can use this contract to clarify our claims to possible authorship***. In practice, it seems to just offset the additional risk of encountering misconceptions about authorship like the ones I've listed above.

  • I agree with the others: the guesstimate from the necessarily limited information about your contribution is that you probably should have been an author.
  • OTOH, in my field (analytical chemistry) only peer-reviewed papers do really count.

    • Therefore, I'd personally not make a big fuss about the poster (as in, not crying over already spilt milk).
    • But I'd express clearly and unambiguously that I'd have thought to have contributed sufficiently to be a co-author. However, I recommend doing starting this discussion in a very polite and professional manner (one-to-one with the prof in question). If the prof does not bring forward valid arguments why I'm mistaken in my view but does not agree to my claim of coauthorship for the paper, it's time firmly make that claim project-publicly.
    • In my experience, there may be some fuss as the client may not want to pay you for working at the paper, and you'll have to decide how to deal with that.
    • Similarly, you'll have to decide how much of a stink you'll raise if they nevertheless submit a paper without your knowledge.
      However, if you are well-known in the relevant scientific community, it may be well to casually let other colleagues know that you are involved in that study and you are the one to do all the statistics and data-management.
  • Importandly, never letting you see the manuscript (so that you don't contribute to the writing/revising of the manuscript - which is a prerequisite for authorship) is not a valid way of excluding you from authorship after you contributed substantially to the published work/study: that would be plagiarism.

  • I've had my authorship to a paper upgraded to co-first-authorship by the editor based on the description of my contribution in the "contributions" section. With that experience, I'd hope that an acknowledgement that acknowledges you doing the data management, statistical analysis and helped with the interpretation of the results of the statistical analysis would trigger reviewers and/or editor to put you as an author. Unfortunately, I think it unlikely that your full contribution will be acknowledged (as opposed to mere help with these points).


* I met academic researchers on full-time contracts who say this without blushing.

** put in here any scientific authorship guidelines that are relevant for your field/location. In my experience, they are pretty much the same in their requirements.

*** For future collaborations: I've discussed concerns with prospective (academic) clients by explaining that though it doesn't have anything to do with them personally, I've had bad experiences with other academic clients in similar circumstances and therefore will work only on particular terms until I know that the collaboration with them (their institution's administration) works smoothly.
My experience with that is in terms of payment delays, which were comparatively easy to discuss with the clients in question (as employees they've probably had delayed reimbursements as well). I'd expect telling the prospective client that you had bad experiences in terms of authorship and in consequence insist on agreeing on the rules for authorship early on should work similarly - again, academics are well aware of all kinds of trouble around legitimate authorship claims, and clarifying authorship at the beginning is recommended for inner-academic collaborations as well.


It is worth having a look at some of the academic literature and ethics documentation on authorship to get an idea of what is considered a valid authorship claim. The Council of Science Editors has published a recent White Paper on Promoting Integrity in Scientific Journal Publications that gives a good summary of research on authorship and attribution in scientific journals (see section 2.2 of the report). Following their 2003 report, they state that ...“there is no universally agreed definition of authorship, although attempts have been made … As a minimum, authors should take responsibility for a particular section of the study" (p. 22).

In the case of a statistician who undertakes the data management and analysis for a project, this seems to me like a "significant contribution" to the paper. Data analysis and interpretation is an expert skill, and a contribution of this kind is a fundamental contribution to a scientific paper. Parker and Berman (1998) give a detailed analysis of authorship by statisticians, and recommend the use of a scoring system which rates the contribution of a statistician to the design, implementation, and analysis of the study. It would be worth your time to read this paper and score your contribution on their scale, to see whether this justifies inclusion as an author on the paper.

Once you have reviewed this material, I would recommend contacting this professor and putting forward your claim to authorship (assuming you still wish to pursue it). You should back up your position by references to documentation of this kind. In the first instance you may be able to negotiate authorship with the professor, but if this is disputed then you should also be able to have recourse to dispute-resolution mechanisms at that professor's university.

Tags:

Authorship