18

The paper in question is an analysis of government statistics to answer a question in sociology. The analysis was applied to a single government survey of a large population.

The sociologist has limited knowledge of statistics. He wrote the introduction, literature review, part of the discussion, and the conclusion. He also listed the variables to be analyzed by a hired statistician.

The statistician wrote the entire results section and a summary for the discussion in his own words, accounting for about 50% of the paper. He designed the analysis and ran all of the calculations.

Is it ethical for the sociologist to list himself as the sole author of the published study? What is customary and ethical in this case?

Glorfindel
  • 613
  • 3
  • 10
  • 20
Eggy
  • 1,445
  • 9
  • 16
  • 5
    Did the statistician run any calculations/models and did the statistician design the statistical analysis? – henning Jul 06 '21 at 13:49
  • 5
    The statistician designed the analysis and ran all of the calculations. The sociologist listed the multiple variables to be analyzed. The analysis was applied to a single government survey of a large population. – Eggy Jul 06 '21 at 13:54
  • 12
    Odd, I didn’t realize there could be such a thing as a sociologist with no knowledge of statistics. Live and learn. – Ed V Jul 06 '21 at 13:54
  • 2
    Some sociologists specialize in qualitative research. But perhaps I should have said "limited knowledge of statistics." – Eggy Jul 06 '21 at 13:55
  • 2
    I am just a passerby from Hot Network Questions, but there is another issue at stake: in some jurisdictions (at least a few European countries to my knowledge), misrepresenting authorship in this way could be illegal under copyright laws. – jaskij Jul 06 '21 at 22:31
  • 3
    A statistician could however give up authorship contractually. – Eggy Jul 06 '21 at 22:59
  • 7
    @Eggy several countries (and academia in general) tend towards the view that the moral rights of the author can't be assigned contractually, only permission to distribute. There's no reason the statistician couldn't publish pseudonymously if he wanted though – origimbo Jul 07 '21 at 00:48
  • 4
    Closely related: When does a statistical consultant become a co-author or collaborator? – Stephan Kolassa Jul 07 '21 at 08:37
  • Can you update the question with the information from comments (as comments may be deleted at any time)? (But *without* "Edit:", "Update:", or similar - the question should appear as if it was written right now).) – Peter Mortensen Jul 07 '21 at 09:23
  • So far we have the description from one side of the dispute. Perhaps hearing from the other side would affect our answer. – GEdgar Jul 08 '21 at 16:18
  • @GEdgar and as if by magic, 1 minute after you said that, this question was posted. – Daniel Hatton Jul 08 '21 at 16:25

6 Answers6

38

It would not be ethical to claim sole authorship in this case and the statistician should be listed as an author. It's always a good idea to get authorship sorted out as early in a project as possible to avoid problems later.

In the fields I'm familiar with (STEM) anyone who made a substantial intellectual contribution to the paper should be an author. Their contractual/payment status shouldn't be part of this decision. The Guidelines for Authorship from the University of Cambridge, and the references cited there, elaborate on this.

From a personal point of view, if questioned on the details of the statistical part, answering that you would need to check with the person who did the work would feel pretty awkward to me if they weren't an author. At least if they are an author they share responsibility for the results, otherwise it's all on you.

atom44
  • 5,680
  • 19
  • 36
  • 5
    Assuming the facts as stated, it would be plagiarism to claim sole authorship. – Buffy Jul 06 '21 at 15:09
  • 1
    Similar to the concern raised by atom44, the author could be questioned about the statistics in a job interview and find himself unable to answer. – Eggy Jul 06 '21 at 15:35
  • "In the fields I'm familiar with" I think you should state what these are, since the OP asks for what is customary (besides ethical). In the law and in econ both my personal experience is that statisticians junior to the main author often prepare their section of the paper and don't get authorship. – Hasse1987 Jul 06 '21 at 23:50
  • 1
    @Hasse1987, I clarified this in an edit. That is actually very interesting about law & econ. It may be worth elaborating on this in an answer. – atom44 Jul 07 '21 at 09:03
  • @Eggy If the sociologist is expected to understand the statistics involved, that could be an issue even if the statistician is listed as a co-author. – chepner Jul 07 '21 at 14:42
  • @Hasse1987 In my limited experience (on the econ side of things), I think it's common that a hired research assistant who conducts a routine analysis under specific directions does not authorship, but I don't think it would be well accepted if an author hires someone to design the whole empirical strategy without giving authorship. – jnanin Jul 09 '21 at 08:41
14

The statistician has a very good case for being included, but not absolutely clear-cut.

The British Sociological Association (BSA) lists a number of criteria for deserving authorship. The BSA criteria are quoted below (emphasis mine). They are similar to the Vancouver Protocol, which relates to medical research but is often referred to beyond medicine.

  1. Everyone who is listed as an author should have made a substantial direct academic contribution (i.e. intellectual responsibility and substantive work) to at least two of the four main components of a typical scientific project or paper:

    a) Conception or design.

    b) Data collection and processing.

    c) Analysis and interpretation of the data.

    d) Writing substantial sections of the paper (e.g. synthesising findings in the literature review or the findings/results section).

  1. Everyone who is listed as an author should have critically reviewed successive drafts of the paper and should approve the final version.

  2. Everyone who is listed as author should be able to defend the paper as a whole (although not necessarily all the technical details).

If a contributor fulfills each main criterion, they must be included as author.

If a contributor does not fulfill each main criterion, they must not be included as author.

If a contributor fulfills some criteria but not others, they should be acknowledged. (This is how I interpret the clause "all those who make a substantial contribution to a paper without fulfilling the criteria for authorship should be acknowledged.")

Paraphrasing from the question and comments:

  • The statistician wrote half of the paper
  • The statistician designed the analysis and ran all of the calculations

The statistician clearly fulfills criterion 1, as they contributed to design, data processing, analysis, and writing.

Since they contributed a crucial component of the paper, they arguably would be able to defend the remainder too, as in criterion 3.

It's not clear whether the statistician approved the final version, as in criterion 2. But it seems likely.

Obviously, the statistician should at least be acknowledged, since they fulfill some criteria. But all things considered, the statistician has a pretty good case for being included as co-author as well, in particular when comparing their contribution to that of the sociologists (which I have not done here).

henning
  • 35,032
  • 10
  • 121
  • 151
  • 6
    While this is a good post, I'm not sure it address the question of the OP (at least not my initial interpretation). While based on what you explain the statistician qualifies for authorship, it doesn't mean he should be listed as an author (it's written "everyone who's an author should have done this", not "everyone who's done this should be an author"). The way I interpreted the question of OP is if it's ok to leave him out of the author list. Ethically speaking, I don't see it as a problem as long as there is a clear agreement between statistician and the author. – cinico Jul 06 '21 at 14:02
  • 7
    @cinico What may be an ethical problem is if the author conceals the fact that none of the statistical analysis/write-up is due to them. It seems to me particularly problematic that the author wouldn't have written the Results section themself. – Kimball Jul 06 '21 at 14:06
  • @cinico fair point. I'll revise this later. – henning Jul 06 '21 at 14:17
  • 2
    @cinico, I think your interpretation is incorrect. Everyone who qualifies as an author needs to be listed. I'm not certain that the statistician has made an "intellectual" contribution to the ideas of the paper, but if they have, they must be listed. Otherwise you are committing plagiarism. Contractual work might or might not rise to the level required, but if it does, list them as author. – Buffy Jul 06 '21 at 15:05
  • The statistician has definitely made an intellectual contribution to the ideas of the paper. – Eggy Jul 06 '21 at 15:26
  • 3
    Henning, I believe you have pinpointed why this situation is difficult: the guidelines I've seen all have several criteria that must be met for authorship, but this statistician's involvement ended when he delivered his report, so the criteria of final review and responsibility aren't met. And yet, common sense tells us there's a fairly severe ethics problem here! – Eggy Jul 06 '21 at 23:03
  • It seems that the statistician satisfies criteria 1.a), b), c) and d) but maybe not 2. and 3. The sociologist on the other hand does not satisfy 1c) but can claim 1.a), b) and d) as well as 2. and 3. So by these criteria the sociologist should be the sole author. – quarague Jul 07 '21 at 08:44
  • @cinico et al. I've revised the question. It should be clearer now. – henning Jul 07 '21 at 09:32
  • Quarague 5, your analysis of who did what is correct, but there's still an ethical problem. Plagiarism is passing another's work off as one's own without acknowledgment. But work that is acknowledged in a note ("Joe Chan reviewed an early draft of this paper) is considered secondary. Here, the statistician's work was primary. The statistician's contribution was greater than the sociologist's. So common criteria for authorship like the Vancouver Protocol aren't infallible, don't cover all possible circumstances. Here, the criteria for sole authorship may be met, but an ethical problem remains. – Eggy Jul 07 '21 at 14:41
  • @Eggy when you (all too) rigorously apply the same criteria to the sociologist and the statistician, the paradoxical outcome will be that neither deserves authorship. To avoid that conclusion, it makes sense to compare both would-be authors' contributions along the criteria above, as you suggest in your comment. That's also what I allude to in my last sentence, and one would have to conclude that if it can't be the case that there is no author, both have to be authors. – henning Jul 07 '21 at 15:01
  • Henning, very good point that when viewed rigorously possibly neither deserves authorship. I noticed that too and I think it's a curious situation. In fact I think it's possible that this paper shouldn't have been undertaken, especially if the sociologist doesn't have the ability to understand the statistician's work. – Eggy Jul 07 '21 at 17:16
  • 2
    @Eggy I wouldn't go that far! Interdisciplinary collaboration is important. I'd rather say that both deserve authorship, since the only other remaining conclusion, when all contributiors are held to the same standards, would be absurd. – henning Jul 07 '21 at 17:23
  • 3
    @quarague and Eggy: The document linked in the answer as "Vancouver protocol" has 'The criteria are not intended for use as a means to disqualify colleagues from authorship who otherwise meet authorship criteria by denying them the opportunity to meet criterion #s 2 or 3. Therefore, all individuals who meet the first criterion should have the opportunity to participate in the review, drafting, and final approval of the manuscript.' I have a feeling that may be the first point where things have gone wrong here. – Daniel Hatton Jul 07 '21 at 22:32
  • 1
    ICJME (http://www.icmje.org/recommendations/browse/roles-and-responsibilities/defining-the-role-of-authors-and-contributors.html) expands their criteria a bit more in details: *"The criteria are not intended for use as a means to disqualify colleagues from authorship who otherwise meet authorship criteria by denying them the opportunity to meet criterion #s 2 or 3 [drafting/revising and final approval]. Therefore, all individuals who meet the first criterion [contribution to study] should have the opportunity to participate in the review, drafting, and final approval of the manuscript. " – cbeleites unhappy with SX Jul 08 '21 at 15:16
11

I do not know what the standards are in sociology, but the contribution of the statistician looks quite substantial to me. Without even looking at standard guidelines as other responses do, from my gut impression that the contribution you describe is well above the threshold for co-authorship.

Leaving them out feels quite wrong, even mentioning it in an acknowledgement them feels like a downgrade. It is perfectly fine for not everyone to be expert in everything and for different people to providing complementary contributions, but leaving them out is just plain wrong.

There is a reason why interdisciplinary research usually has a hard time being justified - one often either drastically over- or underestimates the expertise contribution of the neighbouring fields, leading to a distorted view of the intellectual value of external contributions (in both directions). [yes, I know, statistics is well established in sociology, but the present question precisely reflects the classical interdisciplinary dilemma]

Captain Emacs
  • 47,698
  • 12
  • 112
  • 165
  • In some cases the conflict is that the statistician wants to be given co-authorship and the other author resists. In this case the statistician has not asked for co-authorship to the best of my knowledge. The concern here is the sociologist's ethical obligation as a scholar. – Eggy Jul 06 '21 at 15:31
  • 1
    @Eggy Good point. Authorship is also about accountability, so it's not only a right, but in fact a duty. Of course, if multiple people work together to produce a work under the name of a given company, it's the company's brand at stake, but research is far more individual. Of course, one could end in a dilemma where the statistician does not want to be named, but the sociologist wants to publish. For such an issue where one side wilfully puts the other into an ethical dilemma (assuming the work itself is perfectly scholarly, and not just shoddy rough work), I do not have a good answer. – Captain Emacs Jul 06 '21 at 16:13
  • 3
    I personally think an authorship-entitled person blocking a perfectly acceptable piece of scientific work from being published would itself constitute an ethical breach by that person, but I didn't read that into the question and it would constitute a separate issue. – Captain Emacs Jul 06 '21 at 16:14
7

At this point I will offer what I think is an authoritative answer to the question. I consulted with a senior sociologist who conducts quantitative studies. Here's what he told me: (1) It's common for sociologists to hire statisticians because they don't have and aren't expected to have advanced skills in statistics. (2) Normally the statistician only prepares tables and charts. The statistician does not write a long, publication-ready report. (3) If the latter occurs, the statistician would be a co-author. (4) What is more expectable is that the sociologist has the skill to read and interpret the statistician's work, and then the sociologist writes the narrative results and discussion sections. In that case the sociologist takes sole authorship and might credit the statistician in an acknowledgment. (5) Claiming sole authorship in the present scenario would be plagiarism and it would invite other serious risks, like being unable to discuss the study intelligently in a job interview. (6) In order to claim sole authorship in the present case, the sociologist would need to substantially rewrite the statistician's report, not publish it verbatim.

The senior sociologist questioned why this study was undertaken in the first place, since the junior sociologist has so little knowledge of statistics. There's a case for that, but I also think a junior sociologist could begin working with statisticians to increase skills and learn the ethics of authorship. Not having statistics skills is a handicap for a sociologist, since the answers to many pressing questions in sociology are contained in government statistics.

I'm satisfied with the senior sociologist's answer. It doesn't refer to any formal authorship codes, but it's a good answer from the field. I appreciate everyone's input on this. Thank you!

Eggy
  • 1,445
  • 9
  • 16
  • 1
    I agree that this is pretty common. For instance, my university has a “faculty research support center” with staff that do work like this. The staff will consult on methods and code up the analysis, then hand off the tables to the faculty. They will also review and edit the write up. They may write bullet points as well. So basically the OP example is this model taken to an extreme. – Dawn Jul 07 '21 at 18:42
1

Just to add a search term to the discussion:

Not including someone who did make a substantial intellectual contribution to the study/paper is called ghost authorship and is as such unacceptable.

Academic ghostwriting is not just a matter of the rights of the ghostwriter (who in many jurisdictions can legally agree to their name being excluded from the author list), but

  • it leaves the remaining authors misrepresenting their contributions as including those of the ghost author. This is plagiarism (even though the text is novel and original).
  • it can hide possible conflicts of interest.
cbeleites unhappy with SX
  • 23,007
  • 1
  • 44
  • 91
  • 1
    cbeleites, I stongly agree and appreciate you adding ghost authorship to the discussion. The issue you raise is my main concern. Ghost authorship meets the definition of plagiarism as passing another's work off as one's own. – Eggy Jul 08 '21 at 18:17
0

The BSA criteria for authorship seems like a solid starting point. Indeed if the statistician processed the data, wrote a substantial part of the work and presumably devised the analysis method themselves,it would be a case for acknowledgement or even authorship. But it depends:please take a look at @henning 's criterion no. 2 : the statistician should have seen/ approved of the final form of the paper.It seems trivial but what if someone hires you, a specialist in your field, to do a certain analysis- and you do that , and presumably receive everything, including authorship and some form of remuneration but never get to see the final work? Maybe they used the information you provided in a context you feel it is not correct. Maybe they "load" you with some opinions and interpretations you disagree with, maybe they would misuse your work, or draw the wrong conclusions from a good analysis. And you could potentially be called out for something in "your" paper that you did not effectively sanction -because you never had the chance. So I make a case for "acknowledgement" not "authorship" if the (major) contributor did not have any..editing options, as in rejecting or requiring modifications to be made so as to reflect their own opinion and analysis.

Axel Morisson
  • 1