32

My university has a rule that all students need to publish at-least a couple of papers in SCI rated journals.

A senior told me that it is better not to publish the research contribution in a single paper. The length of paper is not an issue here. The act is meant for increase in paper count and hence becoming eligible to receive the degree.

Assume that the research contribution is on a task T. If I manage to get a couple of methods that perform better than the existing methods. One is accuracy A and another one with accuracy A' where A' > A.

The senior asks to send the algorithm related to A to an SCI journal and submitting method related to A' after enough amount of time on a pre-print site so that it can be sent to another or the same SCI journal for acceptance.

I believe that it does not fall under the category of unfair means. No one can object if I decide to do so. But, I have doubt on whether the act is ethical or unethical to do?

10
  • 29
    The more people game the ridiculous reliance of academia on h-indices to sort people the faster this system will be abandoned for something better. Ergo, it is the height of virtue to maximally inflate your paper count and/or participate in citation rings, etc.
    – Him
    Jun 27 at 15:23
  • 1
    @Him "The more people game the ridiculous ... the faster this system will be abandoned for something better." [citation needed] Since "publish or perish" was imported from the US to Europe around 30 years ago, it went from "nice to have" to "must have". Universities have moved from self- to (science-)externally managed entities which means that KPIs and pseudo-"objective" measures now matter. Yes, it's desirable to abandon them, but it is not going to happen anytime soon. Nonetheless +1 for your comment for the good intention. Jun 27 at 19:51
  • 3
    Does this answer your question? Salami Publication Jun 27 at 20:31
  • 4
    Anyone remember Lance Armstrong's argument? Everyone is doing it so if you don't do it, you might as well not compete. Salami slicing is not the only way. I have seen large research group where everyone's paper puts everyone else's name on it so everyone ends up with a large number of publications. Jun 28 at 11:20
  • 1
    @CaptainEmacs 'Since "publish or perish" was imported from the US to Europe around 30 years ago... Universities have moved from self- to (science-)externally managed' In England and Wales, the legislation that created the "post-1992" universities explicitly listed the professions from which members of the governing bodies of those universities could be drawn. Can you guess what profession was not on the list? Jun 30 at 12:20
87

What you’re describing is basically salami publication. When done with the goal of artificially inflating your publication count and at the cost of reducing the effectiveness of the communication of your ideas and results, I think there is a strong case to be made that it’s (at least mildly) unethical. However, as discussed on the linked Wikipedia page, people who wish to rationalize such behavior have a few reasonably valid arguments they can use to deflect criticism of unethical behavior. So, while it’s certainly not the best practice, it’s also not the worst. Perhaps more than being unethical, it is a shoddy practice used by mediocre people with mediocre ambitions, and will not help a person get a good reputation.

With that being said, if your institution is setting up graduation requirements that strongly incentivize its PhD students to engage in unethical behavior, it really has only itself to blame when they end up engaging in such behavior. In that case, there is also a strong case to be made that it is the institution that is behaving unethically, and that carries a large share of the blame for any unethical practices of its students and faculty.

6
  • 34
    Upvoted, especially for the last paragraph. Jun 26 at 16:02
  • 7
    Is there also a name for the opposite? Cramming together too many loosely related results?
    – Džuris
    Jun 27 at 10:53
  • 4
    @Džuris - There isn't a widely used name for that. But see academia.stackexchange.com/questions/114747/… for some candidate metaphors. Jun 27 at 17:19
  • 1
    It might be worth to also point to publish or perish where you are rewarded for publication numbers (as part of funding/contract extension/promotion).
    – DetlevCM
    Jun 27 at 17:55
  • 3
    Ironically, some journals insist on salami publishing by refusing to publish long papers. Jun 28 at 21:21
17

The use of publication metrics for assessment is unethical. This approach of "salami slicing" publications to game those metrics is an entirely reasonable response to the unethical situation you and your supervisor find yourselves in.

Perhaps, further on in your career, you can fight back against the system you find yourself in and improve it for those that come after you but - for now - you should concentrate on getting your degree and follow your supervisor's advice.

11

This is known as ''salami slicing''. It's not exactly unethical, but it's definitely discouraged.

It's most helpful for other researchers if you collect the relevant and similar results in one place where they can be conveniently accessed. Salami slicing is the opposite of this because you are taking similar, connected results and then dispersing them out and spreading them all out into separate publications in different journals (all of which need to be paid for in theory). This is clearly unhelpful and not good practice for a researcher working in a community of other researchers. Reputation is important in academia, and it will not help your reputation if it is blatant that you are engaging in this practice in order to inflate the quantity of your publications.

In general, this is an important thing to emphasise: quantity is nice, but quality of the publications is the primary thing and this is what your reputation will primarily rest on. If students are being encouraged to reduce the quality of their publications by splitting them into lots of smaller papers, then this is just bad practice.

2
  • 3
    It might be worth to also point to publish or perish where you are rewarded for publication numbers (as part of funding/contract extension/promotion).
    – DetlevCM
    Jun 27 at 17:55
  • 2
    It's officially discouraged, but in practice it is the only meaningful way to consistently achieve the expected publication number. Something like truckdrivers life: A company is paid 100% for their service, only if they bring their goods in 12 hours from Warsaw to Luxembourg. By law, a driver has to rest x hours every 12 hours of driving. The company is discouraging the driver to drive for more than 12 hours, but it is requiring the goods to be in Luxembourg in 12 hours. Warsaw-Luxembourg driving time? 12 hours 40 minutes. Yes, it is discouraged ...
    – EarlGrey
    Jun 29 at 9:42
3

Cynical response here - but do you want to graduate? Do you need these papers to graduate?

Does each piece of work stand on its own? Will it get published in great journals broken apart? If so - they do what you need to do to graduate, and to meet those KPIs that might be stupid but still affect employment/income/graduation.

Sure it might be nice to keep them all together, it probably would make for a stronger paper - but if it means you cannot graduate then you've gained the warm fuzzy feeling of not breaking up your work, and that far worse feeling of failing out of your degree. If it means delaying completion - what will that cost you (tuition fees, funding lost, job opportunities)?

As for ethics - it really is shades of grey. There's absolutely a balance between degrading the quality of publications and meeting KPIs. But if the papers stand on their own, and meet the scientific standards for the field, then do what you need to do. If, however, this means that you have two substandard publications, or that you need to target low-quality journals, then it's not unethical but often counter-productive career wise.

If your second piece of work shows your earlier work to be wrong, and you know this, then yes it would be unethical to publish the earlier piece.

2

I don't think this is merely salami slicing, but something less ethical.

Salami slicing, in my understanding, is when you have two related results which you choose to publish separately rather than in a combined paper. In that case, it is a trade-off between having two weaker papers or one stronger one, and I don't see that there is anything particularly wrong with either option, particularly when it is the natural response to some external pressure.

However, here you appear to only have one result (that you can improve the accuracy to that of A'), and to want to get a weaker version of the same result accepted before you tell anyone what your real result is. Perhaps I've misunderstood, but the fact that you talk about delaying submitting the second paper suggests that the existence of A', if disclosed, would prevent your results on A from getting published. In that case I think you would be misleading the journal by not disclosing this.

6
  • I understood that he has 2 methods, A and A+. A that improves state of the art and A+ that improves it further. Obviously if A+ gets published first, nobody would care for method A, so he needs to publish A first (or together with A+). And if he does publish A first followed by A+, nobody would even notice he intentionally postponed A+ because this "we did X and it is the best", "now we did Y and it is even better" is how most if not all of science works and worked. Jun 28 at 11:34
  • 2
    @ZizyArcher Yes, I think that is accurate. But "we did X and it is the best" is simply not true in this case, since OP has already done better. (And the fact that nobody would know does not make something ethical.) Jun 28 at 11:51
  • 1
    @Zizy: "Nobody would even notice" == "you can get away with". Even your words imply unethical behavior. Just because you won't get caught doesn't mean it's right. Jun 28 at 15:21
  • @EspeciallyLime One can always argue to himself "let's submit A to the peer-reveiw, maybe I missed some important things and it is all wrong" and in the meanwhile double, triple-checking the results from A+ . Yes, double checking means only "let's upload paper A+ to our svn server so I can track down modifications" ...
    – EarlGrey
    Jun 29 at 9:51
  • @CrisLuengo This equality is questionable to put it mildly. I am simply observing that his hypothetical succession of papers is indistinguishable from ordinary progress in field. You either publish just one paper when you abandon the problem/field/..., or you publish many papers that build on previous ones. Jun 29 at 10:30
0

If you send paper A to a journal and A' to a preprint server, there is a clear chance that the reviewers look at the preprint A' while reviewing A. How does it look then?

Irrespective of your institutional guidelines, its always better to put your best foot forward, while submitting a paper to a Journal. Withholding can only backfire.

Edit after comment by Wrzlprmft

Thank you. Well, then my answer makes a "U" turn. If the institute measures the quantum of work by the number of papers, and if you feel that the research is good enough to be split into two and still both be acceptable to a quality journal, then its a good gamble...no question of being unethical. Neither the Journal nor the University has put restrictions on the timing of the research conducted/invented.

8
  • I mentioned after enough amount of time.
    – hanugm
    Jun 28 at 5:21
  • irrespective of the whether they look at preprint or not, reviewers can make out of existence of better method that you have withheld.
    – Rajesh D
    Jun 28 at 5:23
  • Don't try to fool a journal. The editors and reviewers have vast experience.
    – Rajesh D
    Jun 28 at 5:25
  • One of my senior already published atleast 9 papers in such way. So, I am not thinking that editors and reviewers can object.
    – hanugm
    Jun 28 at 5:27
  • Everybody knows, that one you start writing, research is going on. Once the paper is written and sent to a journal, you might have better results. But at some point you have to publish, othewise you could endlessly improve your paper with new results.
    – usr1234567
    Jun 28 at 8:56
-2

In this situation, I believe, one has to worry not so much about ethics but rather about whether one contributes to the overall degradation of the quality of information exchanged through scientific publications. For an analogy, when you try to use your car less so as not to contribute to the air pollution too much, this is not the matter of ethics, it is the matter of survival.

6
  • @Downvoter: sorry if it sounds unjustified but I really believe what I wrote. Jun 29 at 9:22
  • I disagree: The personal price of such behavior is (unfortunately) far exceeded by the personal benefit. That's why people do it. There's a net loss, but this is only in aggregate; selfishness is still rewarded. Mind you, there's a benefit in convincing people otherwise. In game theory, this situation is somewhat of an example of a "commons dilemma."
    – Brian
    Jun 29 at 18:46
  • 1
    @Brian Maybe I expressed myself not clearly enough. I did not mean personal survival but rather collective survival. Specifically in this case, if you feel that it is more and more difficult to find in the literature recent new information relevant for you, you would be probably more and more reluctant to contribute to the increase of this difficulty yourself, no? Jun 29 at 19:02
  • You expressed yourself perfectly clearly, but the expected impact (here, I use expected in the probabilistic sense, not the English sense) of personal behavior on collective survival is negligible. Thus, a selfish actor typically has no reason to care about collective survival. This is very unfortunate and the results are...bad. Going a step further, the impact on personal behavior on collective survival is also typically delayed, exacerbating the problem.
    – Brian
    Jun 29 at 19:12
  • @Brian Maybe you are sort of aggravating? After all, more and more people nowadays become ecologically conscious, sparing water, buying eco-friendly goods, etc. Why not expect the similar tendency with respect to information pollution? Jun 29 at 21:13

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.