Is it time for a new classification system for scientific misconduct?

Toshio Kuroki

Are current classification systems for research misconduct adequate? Toshio Kuroki — special advisor to the Japan Society for the Promotion of Science and professor emeritus at the University of Tokyo and Gifu University — thinks the answer is no. In a new paper in Accountability in Research, Kuroki — who has published on research misconduct before — suggests a new classification system. We asked him a few questions about his proposal. The answers are lightly edited for clarity.

Retraction Watch (RW): Why did you feel that a new classification of misconduct was necessary?

Toshio Kuroki (TK): The STAP affair, starring Haruko Obokata, was my inspiration to become a “misconductologist.” In 2016, I published a book in Japanese on research misconduct for the general public.

The Japanese version — which is now being translated into English for publication by Oxford University Press — included the classical classification of fabrication, falsification and plagiarism (FFP) and questionable research practices (QRP). However, the FFP/QRP classification was not quite right to me. It seems too simple, placing one-dimensionally, or linearly, responsible conduct of Research (RCR)-QRP-FFP on a line moving from white (RCR) to grey (QRP) and finally black (FFP).

In other words, the classical classification emphasizes FFP without distinguishing between betrayal of the truth — fabrication and falsification and trust — plagiarism — while de-emphasizing QRP. QRP covers many types of inappropriate behavior, but not all are necessarily clearly defined. Some of QRPs are truly inadequate actions, more than just “questionable.”

In writing the Oxford University Press version of the book, I conceived of a new classification that takes into account the outcomes of misconduct, while reflecting the core values of truth, trust and risk in a more tangible fashion. Under this classification, various kinds of fraudulent behavior can be understood in two or three dimensions, or non-linearly.

To be honest, I was reluctant to publish the new classification, because it challenges the global standard of classification, which was established after several revisions starting with proposals in the latter half of the 20th century. However, the editor of Accountability in Research kindly accepted and published after revision.

RW: Please explain the three categories, and what goes into them.

TK: The proposed classification includes three classes of misconduct, depending on whether the case reflects betrayal of the truth, of trust, or risk to safety. Each class is further classified under subheadings.

“Class I Misconduct: Betrayal of truth” includes fabrication and falsification, which I consider to be extreme misconduct betraying the truth of nature.

“Class II Misconduct: Betrayal of trust” includes three subcategories, i.e. plagiarism, irreproducibility and inadequate research practice.

  1. Plagiarism is serious misbehavior and common, from student reports to dissertations. It betrays trust in science, rather than the truth of nature.
  2. Irreproducibility is also increases the risk that the public will lose trust in science, but in the classification it is merely included as a QRPs. In the proposed classification, irreproducibility is rated as equally serious to plagiarism, regardless of intent.
  3. The inadequate practice of research prevails in the scientific community. It includes a large variety of fraudulent practice such as ghost authorship, gift authorship, concealing inconvenient data, duplicate publication, and fake peer review. However, some behavior such as harassment is not necessarily directly related to research misconduct, but is obviously violates research integrity.

“Class III Misconduct: Risk to safety” takes social influences into account, although it is not particularly included in the classic classification. Among many undesirable outcomes, risk to the safety of health and industrial products are most serious.

  1. Risk to safety of health is the most obvious and worst outcome resulting from betrayal of truth and trust.
  2. Risk to the safety of industrial products should also be taken seriously. If unqualified products are supplied to markets, they will cause risk in our daily lives. Typical examples include Volkswagen’s fraud involving emissions by diesel engines, and lack of oversight of anti-seismic devices used in many buildings, including hospitals and fire stations, in Japan.

RW: Is the idea that a given act could be two or even three categories at once?

TK: Under some circumstances, acts that betray truth or trust that caused risks to the public should require notification.

Class III is a sort of an optional category of misconduct. If the primary misconduct in terms of betrayal of truth or trust is known, such misconduct can be classified into two or three categories at once. However, there is not much transparency about most misconduct that takes place in industry, and investigation reports are not available. Such cases are solely categorized as Class III misconduct.

Collectively, Class III misconduct is an additional or optional category set up to notify the public of possible risks to their lives.

RW: Should sanctions for some classifications be more severe than for others?

TK: I am concerned that sanctions for misconduct have not been well defined and are ambiguous in terms of assessment of culpability, inequitable judgement, retroactive determination and so on.

I am also aware of the fact that FFP is often the only form of misconduct that is subject to punishment or administrative action, while QRP seems to be underestimated and even sometimes overlooked. The proposed classification would facilitate more tailored analyses and application of appropriate corrective actions in accordance with the level of classification.

Like Retraction Watch? You can make a tax-deductible contribution to support our growth, follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up for an email every time there’s a new post (look for the “follow” button at the lower right part of your screen), or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at

16 thoughts on “Is it time for a new classification system for scientific misconduct?”

  1. Let’s hope this will find it’s way to all the research integrity Commissions and will be common practice.

    A lot of misconduct is now nearly impossible to address properly.

  2. I do totally agree with the proposed classification of misconduct in scientific research and I do hope it will seriously applied by the scientific community. Furthermore, it is also necessary to standardize real sanctions for scientific misconduct in research zccording to the level and seriousness of the misconduct.

  3. I very much appreciate this type classification (i.e., betrayal of truth vs. betrayal of trust). And, yes, it seems to me that some forms of misconduct can fall under all three classifications. Consider plagiarism of others’ data or the covert reuse of one’s own previously disseminated data. Both of these fraudulent acts represent betrayals of trust and of truth. Because of the latter, there is also the possibility of placing the public’s health at risk.

  4. Can someone explain “irreproducibility” to me? I think I must be missing the point. My own work may be thoroughly reproducible in my own lab, but unfortunately turn out to be irreproducible generally (for example because of a unique contamination in the water-supply in my lab, or the fact I happen to use a certain brand of filter-paper; no matter how hard one tries, there is always something that we’ve taken for granted). I can’t ask other labs to reproduce all my work before publishing – they wouldn’t do it! So I do my work as best I can, repeat as best I can, and publish what seems to be reliable. If it turns out irreproducible, it should certainly be marked in the literature as such. But this sort of irreproducibility is an accident of genuine, honest research, not a form of misconduct. It shouldn’t be written alongside plagiarism, because the plagiarist is probably a deliberate criminal, entirely conscious of their intellectual theft, while the person who is hoodwinked by some silly minor detail that actually undermines their entire work is, to a large extent, an innocent victim of the complexity of designing an experiment. We try, we don’t always get it right. We should be castigated for deliberately messing up, cheating, and concealing our errors; we shouldn’t suffer unduly for making errors and admitting them as soon as we’re aware they exist.
    Is it a different sort of irreproducibility that was meant?

    1. Just as much plagiarism on the part of our students is unintentional not checking different filter papers is unintentional

  5. Most of these problems (at least in the USA) will go away the moment those miscreants are treated the way they ought to be. Common criminals with an MD or a PhD (or both) after their names. They waste federal funds (which could have gone to honest scientists to do honest research which will result in outcomes that benefit the US taxpayer), some conduct clinical trials based on the spurious research which will not benefit the participants (and may actually harm them), destroy the future of student trainees, occupy lab space that are used for the nefarious research etc. Get the FBI involved, give some teeth to the ORI at NIH to prosecute these criminals, ask the universities to the return the ill-gotten grant funds, and finally throw some of these criminals in jail. Either these ghouls will disappear from universities, or they will become too scared to fudge research, knowing that sooner or later they will end up in the big-house. If bank managers and stock market crooks can end up behind bars for lesser crimes, why not these scum?

    1. “If bank managers and stock market crooks can end up behind bars for lesser crimes, why not these scum?”
      Lol, how many of those go free? Also, “lesser crimes”? Financial crimes are usually measured in the millions or billions. Research misconduct in the tens or hundreds of thousands. I’m not defending the bad actors, but have some perspective. Plagiarizing an article should get toy fired, but I don’t think it’s worth the FBI’s time. And nobody had their life savings wiped because of it.

      1. Wake’s paper on risks of vaccination, found to be based on fraudulent data, has arguably lead to many deaths in epidemics of preventable diseases. We had an epidemic of mumps on my campus last year due to too many unvaccinated students. I think an accurate perspective on scientific fraud must recognize that it *is* potentially serious, especially but not solely in biomedicine.

        1. Ok, that is a particularly egregious case, one where I think jail time would be justified. But most misconduct (plagiarism, photoshopping western blots to falsify the activity of some gene) are reprehensible, but not deadly. I don’t think the it’s realistic to ask the FBI to get involved.

      2. Hundreds of thousands? No, a typical NIH R01 is in the range of $1.9 million or more. I am not talking about plagiarism-although that is bad enough. I am talking about research fraud on top of plagiarism. It has been going on for too long. NIH has been listing the names of those fraudsters online for years, with a slap on the wrist for the fraudsters. Has that helped? No. PubPeer and Retraction Watch have been doing a thankless job outing them, only to be threatened with lawsuits. The only result has been that the crooks have become more sophisticated. Yes, with millions of dollars and patient safety involved, the FBI needs to be brought in – otherwise the fraud will simply continue. Yes, no one has had their life savings wiped out, but every taxpayer is being bilked by these crooked MDs and PhDs. Think about the clinical trials floated as a result of such spurious research – isn’t that criminal? How many patients may have been affected?

    2. You’re on the right track.

      There’s already a very good mechanism in place to deal with these fraudsters. It’s called Qui Tam, or False Claims Act.

      “The federal government’s massive spending on research and development has been a frequent target for fraudulent and false claims. Many of the most prestigious academic and private research facilities have been accused of research fraud. Some of the common forms of research grant fraud include:

      Falsifying a grant application in order to secure a grant
      Falsifying research data and results
      Over-charging time, costs and other expenses associated with the grant
      Falsifying purchase orders for equipment and materials
      Using grant money for other unrelated research
      Using grant money for personal expenses
      Improper conflicts of interest by the principal investigators
      Falsifying progress reports and other documentation
      Failing to comply with applicable government safety and other regulations”


  6. How to classify this retraction?
    April 27, 2001 The Journal of Biological Chemistry
    276, 14459-14465.
    Insulin Stimulates PKCζ-mediated Phosphorylation of Insulin Receptor Substrate-1 (IRS-1)
    Yan-Fang Liu, Keren Paz, Avia Herschkovitz, Addy Alt‡, Tamar Tennenbaum‡, Sanford R. Sampson‡, Motoi Ohba§, Toshio Kuroki§, Derek LeRoith¶ and Yehiel Zick‖
    – Author Affiliations

    From the Department of Molecular Cell Biology, The Weizmann Institute of Science, Rehovot 76100, Israel, the ‡Faculty of Life Sciences, Gonda-Goldschmied Center, Bar-Ilan University, Ramat-Gan 52900, Israel, the §Institute of Molecular Oncology, Showa University, 1-5-8 Hatanodai, Shinagawa-ku, Tokyo 142-8555, Japan, and the ¶Molecular and Cellular Endocrinology Branch, NIDDK, National Institutes of Health, Bethesda, Maryland 20892

    2018 retraction notice.
    This article has been withdrawn by Sanford R. Sampson, Motoi Ohba, Toshio Kuroki, Derek LeRoith, and Yehiel Zick. The withdrawing authors have become aware of several errors in the way images were presented in this manuscript. Because the original data are no longer available, the authors wish to withdraw the article in the interests of maintaining their publication standards and those of the journal. In the IRS-1 immunoblot in Fig. 3, lanes 1, 2, and 8 were duplicated. The PKCζ immunoblot from Fig. 4B was reused in Fig. 5. In Fig. 5, lanes 9 and 12 of the IRS-1 (IR-IRS-1 complex) immunoblot were duplicated. Additionally, lanes 1 and 8 and lanes 5, 10, and 11 of the IRS-1 (total) immunoblot were duplicated. In Fig. 6, lanes 2 and 9 of the IRS-1 immunoblot were duplicated. The withdrawing authors state that these presentational errors do not impact the underlying scientific findings of the article, which are also presented as quantitative bar graphs that essentially summarize data of a number of experiments. Therefore, the withdrawing authors stand by the original scientific results as described, which in the authors’ opinion has been confirmed in other laboratories (Moeschel, K., Beck, A., Weigert, C., Lammers, R., Kalbacher, H., Voelter, W., Schleicher, E. D., Häring, H.-U., and Lehmann, R. (2004) Protein kinase C-ζ-induced phosphorylation of Ser318 in insulin receptor substrate-1 (IRS-1) attenuates the interaction with the insulin receptor and the tyrosine phosphorylation of IRS-1. J. Biol. Chem. 279, 25157–25163 and Neid, M., Datta, K., Stephan, S., Khanna, I., Pal, S., Shaw, L., White, M., and Mukhopadhyay, D. (2004) Role of insulin receptor substrates and protein kinase C-ζ in vascular permeability factor/vascular endothelial growth factor expression in pancreatic cancer cells. J. Biol. Chem. 279, 3941–3948).

    1. Dear Dr. Pessoa

      Yes, this is true. Of >300 publications in my list over 50 years, this is the only paper that was retracted . Thank you for your mentioning it and providing an opportunity to explain a bit more about the background of this unfortunate event.

      On November 2000, I received a request mail from Dr. Yehiel Zick at the Weizmann Institute of Science, Israel, about the dominant-negative PKC (protein kinase C) zeta isoform for thier collaborative study with National Institute of Diabetes and Digestive and Kidney Diseases(NIDDK)of NIH. They needed it for additional experiments that were requested by reviewers for publication in J. of Biological Chemistry (JBC). I remember that we were very pleased to help the study done in these two prestageous institutes.

      At that time, we were interested in PKC-mediated signal transduction pathways. We established a series of adenovius expressing vectors of all isoforms of PKC including dominant-negative variants. Our vector played a crucial role in revising their study on possible involvement of PKC zeta in insulin-mediated signal transduction. The paper was published on January 29, 2001 in JBC.

      In February 2018, a mail advising retraction of this paper was forwarded from the editor of JBC because of the image falsification of five figures of electrophoresis. Why then it became an issue 17 years later? It started with the on-line watching site Pubpeer: seven of Zick’s papers already had been pointed out of image falsification. I sent an agreement email of retraction to the editor of JBC.

      Obviously, this was not our own error: our contribution to this work is only limited to providing the PKC vector but nothing more. To be honest, however, it is not very comfortable to see a paper that bears my name retracted. What we learned is that if you provide an experimental material to others, it is storongly advisable to confirm only a credit should be noted in acknowledgement only, not joinning as a co-author. I added this case in my book to be published from Oxford University Press as a lesson learnt.

      Toshio Kuroki

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.