14

I have seen colleagues present the same data analysis multiple times over several years.

By present, I mean two things. First, present the work to other faculty and PhD students in departmental seminars and informal brown bags. Second, I mean present the research to journals by submitting it for peer review.

Each time I see the work presented, the data analysis is rather similar to the last time I saw it presented. But what the study is "about" changes significantly, sometimes using different theories and claiming to have a new research question.

What does not change is that the paper claims to test hypotheses derived from whatever theory they are using. The hypotheses keep changing, but the analysis doesn't change.

The changes to the theory and research question are explained as having the previous paper rejected at a journal and needing to change the "framing" of the paper or responding to reviewers who "didn't like" the way the analysis was presented.

Is the behavior of my colleagues ethical?

10
  • 8
    How does ethics enter in to this? Maybe it is misguided, but unethical? Maybe they are just getting a deeper understanding.
    – Buffy
    Oct 19, 2018 at 20:35
  • 8
    It would be helpful if you clarify what you mean by "present". If it is just presenting as-in some sort of non-archival presentation or discussion, if they are changing the framing of a study and resubmitting it after a rejection, or if they are publishing year after year the same basic work - these would all call for very different responses!
    – BrianH
    Oct 19, 2018 at 21:06
  • 1
    Should “ethical” be replaced by “lazy”?
    – Solar Mike
    Oct 19, 2018 at 21:07
  • 1
    @BrianH I updated to specify what I mean by "present"
    – birch
    Oct 20, 2018 at 17:38
  • 2
    One thing stands out - you say they are "changing the framing" in response to reviewers comments etc. Does that mean that none of their previous work submitted to journals was accepted after peer-review, or do they keep re-submitting work which has significant overlap (in the analysis, as you say) which was already accepted for publication or published?
    – penelope
    Oct 23, 2018 at 12:44

4 Answers 4

32

What venue are they “presenting” the work in, and does it carry an expectation of originality?

Publishing multiple minor variations of the same piece of work would indeed generally be considered as unethical, or at best borderline-ethical “salami-slicing”. That’s because publishing carries an expectation of originality/novelty: you’re not supposed to publish the same work twice. This is deeply entwined with the ways that publications are used as a metric of productivity. However, it sounds like this isn’t what’s happening here, if your colleagues haven’t succeeded in publishing the work.

Whether presenting multiple minor variations of the work is ethical depends entirely on what venue they’re presenting in (major conferences? informal workshops? the departmental seminar?) and whether it has an expectation of originality like publications would.

This varies between fields and subfields, and between conferences within subfields. In pure maths (at least in the subfields I’m familiar with), conference presentations have no expectation of originality. You can present the same piece of work every year for a decade, and people won’t be very excited or impressed, but it won’t be considered as unethical. In computer science, on the other hand (again, in the subfields I know), presentations at major conferences are like a smaller version of a journal publication, and as such, are explicitly required to be original: re-using material between them would be unethical. But in less formal settings (e.g. local workshops), again there’s no such expectation of originality, and re-use of material (with or without variation) is fine.

2
  • 13
    I upvoted your answer, it is much more comprehensive than mine...
    – damian
    Oct 19, 2018 at 20:15
  • "That’s because publishing carries an expectation of originality/novelty: you’re not supposed to publish the same work twice." It worked fine for George Lucas... Oct 20, 2018 at 23:04
14

It is difficult to say whether it is ethical or not; it also will differ based on what the conferences say (e.g., do you have to explicitly state that the work has not been presented elsewhere?). There is also a gray area here: There is nothing wrong, for instance, with a secondary analysis of previously collected data.

But a colleague who has nothing new to tell for years, has probably a much bigger problem anyway...

1
  • 14
    I upvoted your answer, it is much more succinct than mine…
    – PLL
    Oct 19, 2018 at 20:18
2

Original research may constitute research containing at least one of:

  • original theoretical frame
  • original interpretation of primary material
  • original primary material ("data," "text," "object of analysis") (Source: Australian HERDC standards, if in doubt use the last year that the publications were collected)

One of the responses to the imposition of Fordist-Taylorist labour conditions in "The University" has been for scholars to reduce the scope and originality of research projects.

Whether this is ethical largely depends on your discipline's and institution's ethical and employment frame work, and the tolerance of journals for similar works. For example, permuting the literary texts while keeping the theory and analytical tools constant, will probably result in rejection in a literary journal: the scope for replication is smaller in literary studies.

Outside of ethics I would suggest that this is risky from a labour and employment stand-point; that departmental colleagues, unless rigorously cynical yet highly solidaristic, are likely to throw you under the bus; and, that higher level bosses at school, faculty, college or university level are going to strongly test your belief that tenure still exists.

On the other hand, as long as the volume of papers in preparation and received for publication meet standards, and if the deficiency in originality is made up for in other papers, seeking collegial support for a difficult to bring to publication piece of research is fine. Even if it bores at departmental research seminars. Maybe they need a new coauthor?

2
  • Can you elaborate what you mean by "rigorously cynical yet highly solidaristic"?
    – birch
    Oct 23, 2018 at 13:56
  • Rigorously cynical in the sense that they’ve analysed the effort/reward ratios in “originality” and decided on that basis to accept minimised effort thereby effectively betraying their beliefs about best serving their discipline and advancement of knowledge by serving volume over science. Highly solidaristic in the sense that they have labour or scholarly solidarity for their department colleagues instead of blocking them on the basis of lack of quality / service to knowledge. Minimal originality output strategies allows departmental competitors to attack colleagues on that “quality” basis. Oct 26, 2018 at 1:34
0

In light of the recent comment by the OP, I think a lot of speculation can be removed from the answer. I fully agree with the answer by PLL regarding different possibilities which he covers.

The colleagues in question have been fighting for some time to publish a specific piece of research they've been working on. None of this seems unethical:

  • The work has itself never been published as a peer-reviewed publication.
  • Changing the manuscript in response to the reviewers comments (as OP says) to improve or adapt scope, level of detail and presentation style is in fact expected.
  • Updating the manuscript over time with new references, but also new findings and conclusions is expected, especially if this is an ongoing work.
  • Presenting the work at departmental seminars is an established way to get feedback.

In fact, I have a publication now that I started to write in the first year of my PhD, started to submit in the 2nd, and finally got accepted almost 5 years after starting it initially. It was a survey paper, so I can make limited parallels, but the level of detail, target audience, presentation style, etc. have all drastically changed several times over based on the reviewers comments as well as the target journal.

I'll just echo damian's answer: rather than being unethical, spending years to publish a result probably speaks volumes about your colleague's quality as a researcher. This should also be taken with a grain of salt however, as there might be valid reasons in persisting to publish one specific result (while continuing your other research activities): see my example above, but also MSER descriptors (Google Scholar search link to show the number of citations) got rejected from many top-tier Computer Vision conferences before being accepted and popularized, and the story of early days of deep learning as one example with a really big impact.

1
  • One reason I wonder about the ethics of such long-term paper transformations is that it seems at some point it becomes HARKing--hypothesizing after results are known. I don't see how it is ethical to keep repackaging the same empirical results as different "studies" and each time claiming the original intent of the project was to test the new hypotheses, but I know that the analysis hasn't changed for a long time.
    – birch
    Oct 23, 2018 at 18:06

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .