The Good, the Bad and the Ugly: What retractions say about scientific transparency

photo by Cea via Flickr http://www.flickr.com/photos/centralasian/

Recently, Ivan has been invited to speak to two groups — the Danforth Center, in St. Louis, and CrossRef members, at their annual meeting in Cambridge, Mass. — about retractions and Retraction Watch. He gave variations on the talk below, “The Good, the Bad, and the Ugly: What Retractions Tell Us About Scientific Transparency.” In it, he discusses a number of cases we’ve uncovered at Retraction Watch, and offers some solutions for improving transparency. Use the arrows at the bottom of the slides to click through the presentation.

We welcome your feedback, and the opportunity to come speak to your group, too. Find our contact info here.

35 thoughts on “The Good, the Bad and the Ugly: What retractions say about scientific transparency”

    1. Well… of course the claims need to be checked. But I think his point is that just because someone makes anonymous claim, it is not untrue, while someone making claim in their own name does not make the claim true. What should matter in the content, not the person saying it. At least that’s what ideally should happen in my opinion.

      1. I was born in Soviet Union. “They” were very particular in respect of anonymous signals. Considering current state of scientific competition, I believe this is the road to Hell.

    2. Dear Elena,

      Guess what? The behavior you mentioned at 4:19 pm on the 22 November 2011 was probably not confined to the Soviet Union.

      The point is that people higher up the ladder always made, and still make, decisions about people lower down anonymously. You just didn’t know because of the simple fact that they didn’t even tell you. Now there is a possibility that people lower down can alter the fate of the people higher up and they don’t like it.

      You may come to the notice of the authorities, they must read this, as a “useful person”. Perhapsnotyou intention, but my prediction. I do not wish my comment to be personal, but the authorities do need “manageable” people. They recognize them.

  1. I think that this article is quite refreshing. It is about finding and analysis of duplicate publications in traditional Japanese herbal medicine. The authors found that 11 out of 384 ( 2.86 %) papers counted as duplicates.

    http://www.jcimjournal.com/en/FullText2.aspx?articleID=jcim20111003

    I do not know if any of the studies in traditional Japanese herbal medicine have been retracted. That may happen.

    Is there any reason to believe that what goes for traditional Japanese herbal medicine does not go for other forms of medicine?

    Does anybody have experience with duplicate publications?

    1. “…..Does anybody have experience with duplicate publications?…..”

      The continual republishing of my one and only self-plagiarised word salad is justified in the public interest. In my defence, I like to think that this serial re-publication as a temporal concatenate keeps me out of the bars and off the streets – and therein lies the public interest. I am loathe to use the word, ‘plagiarise’ in this case though – preferring the phrase, ‘reflective serendipitous academic convergence’ instead.

      1. Dear Bil,

        I am often a bit slow on the uptake. You comments reminded me of the hydraulic computer (it worked quite well, although not the U.K. economy) used to predict changes and steer the U.K. economy. I predict that recycling would reduce funds available to consumers and inhibit growth. It sounds like a simple question, but Iam never sure. How say you?

        http://en.wikipedia.org/wiki/MONIAC_Computer

    2. This 2011 study of published errata in 2 high impact oncology journals

      http://www.current-oncology.com/index.php/oncology/article/view/707/602

      concluded:

      “Error rates in high-impact oncology journals average 4%, which is likely an underestimate, because errors noticed by readers are not consistently reported. Propagation of serious errors decreases, but still continues, after publication of errata.”

      26 out of the 190 errors were classified as serious.

      Should we add this 4% error rate to the near 3% duplication rate (assuming that people in clinical oncology behave like those in traditional Japanese herbal medicine)?

      How to estimate the true rate of errors?

  2. Small thing, probably: On slide 19, you first write about events at University College London, displaying the logo of the University of Cambridge next to that. University of Cambridge shows up in the second paragraph, so using the logo on the slide is okay, but the layout of the slide could be perceived as a failure to distinguish between the two institutions. These institutions being old and proud, this could lead to complaints.

  3. To be honest, I don’t see a problem with “self-plagiarism”, if it only involves parts of background, introduction and discussion. All authors should hold the rights to what they publish, and if they like to use the same words somewhere else it should be their right to do so. Why should people waste a lot of time paraphrasing their own words so that another review paper does not have exactly the same words? You came up with the words in the first place, you can do whatever you want with them. The only technicality is that currently publishers typically hold the rights to your publications. But please, don’t make this into an “ethical issue”. It’s not an ethical issue. At least, that is how I see it.

    1. Dear Joseph Lakatos,

      1. Where do you draw the line?
      2. Are you aware of the ICMJE guidelines on overlapping publications?

      3. Why don’t the authors simply put the pen down if there is nothing new to say?
      4. The publishers have copyright. It is thus. I imagine that the deal is that they publish what you write with the understanding that belongs to them. If you do use exerpts you need to cite the original publication.

      Copying is an ethical issue.
      5. Copying from yourself is an ethical issue because you are giving yourself double points.
      These can be redeemed for positions, salaries, grants,reputation, leading to misdirection of resources.
      6. It is also an ethical issue because it confuses others. Scientists are interested in what is, not was has been copied. Copied from others, or from youself does, has the same effect, the copies will appear in the literature. Copies clutter the literature, degrading the picture.

      7. If you do copy from yourself there is a risk it will be retracted.

      Some recent examples of retractions, which include self-plagiarism as grounds.

      http://annonc.oxfordjournals.org/content/22/11/2536.full
      http://www.ncbi.nlm.nih.gov/pubmed/22024397
      http://jco.ascopubs.org/content/29/33/4471.full.pdf+html

      1. Dear dk, I think you are correct.

        The paper by Tazi et al appears to be plagiarised from Miyake et al:
        http://www.ncbi.nlm.nih.gov/pubmed/20686910

        The other already makes it abundantly clear in the retraction notice that it is not a matter of self-plagiarism.

        The first example is one where also the results were also copied, which isn’t exactly what Joseph Lakatos was referring to.

  4. “3. Why don’t the authors simply put the pen down if there is nothing new to say?”

    Very simple: 1. Because not everybody reads every journal. In fact, most people read very few journal regularly. Therefore, making the same points in multiple journals, even if the same language is used at times should be no problem. 2. Because we are not talking about copying an entire paper or misleading people into thinking that it is a new empirical study, but only parts of it in the context of other parts that are different.

    The idea of “double points” is bogus in this case. The goal of science is to spread knowledge and ideas and what settles in the long run. “Self-plagiarism” is a way to spread your ideas widely and in my opinion there is nothing ethically questionable with it.

  5. “6. It is also an ethical issue because it confuses others. Scientists are interested in what is, not was has been copied. Copied from others, or from youself does, has the same effect, the copies will appear in the literature. Copies clutter the literature, degrading the picture.”

    How could a copy “degrade” the picture? Fabrication and rushed paraphrasing of already clearly written text and concepts degrade, not copies. If one copies a review section where theory A is described and places it in a new context, how would that degrade anything? Please…

    1. What about point 2? Are you aware of the ICMJE guidelines on overlapping publications?

      In answer to “How could a copy “degrade” the picture?” I thought that was self-evident, but not to some.

      You have to give the source of the idea, concept, which would mean a citation, of yourself in this case.
      Otherwise people will think it is new, or an independent thing. It is like couting: is there a difference between 1 and 2? Not everybody will be familiar with your work.

      Most “new contexts” are just haging the wall-paper in another room.

  6. “1. Where do you draw the line?”

    If it’s intro, discussion, methods, self-copying is should be fine. The unacceptable part is passing an old study (with old data) for new. I caught a few of those in my field and had them killed after acceptance, but luckily before publication. That is where the double point argument comes in, at least in my field. For review papers and the like, I don’t see a problem.

  7. If you reuse a figure or data from your own previous papers – isn’t it not plagiarism and/or misconduct? How can one say that self-plagiarism is OK. Once the figures/data were used for a previous publication, in my opinion, they cannot be used for another publication.

    1. My understanding is that some data re-use is OK if the previous papers are clearly cited. This goes for review articles and also for papers with slightly different focuses that share some control or other data. The important thing is to cite the re-used data, and also include the info in the cover letter, so the journal and reviewers know which data is original and which is re-used.

      I thought this was how things worked… maybe it is different in certain fields?

      1. This seems to be exactly the issue with some papers by Gomes, the Brazilian Forensic Entomologist. Maybe you will find a solid case example form the last post on this blog for this discussion.

        Word has spread that Gomes et al. used lengthy amount of self-plagiarism and plagiarism is a paper-multiplication maneuver that ultimately led him into getting a university professor position at record-speed. Problem is, their papers are now crumbling under the scrutinity of many readers and the Abnormal Science Blog.

        To which extent should they get punished from reusing narrative and data, usually without clear indication of the practice? Happening right now, forensic case.

      2. Dear V,

        Here is the relevant section on the rules in biomedicine. They agree with what you wrote, but that

        “The footnote on the title page of the secondary version informs readers, peers, and documenting agencies that the paper has been published in whole or in part and states the primary reference. A suitable footnote might read: “This article is based on a study first reported in the [title of journal, with full reference].” ”

        Common practice is to cite the oringinal publication in the figure legends too.

        Here are the rules though.

        http://www.icmje.org/publishing_4overlap.html

        In particular these points:

        Secondary publication for various other reasons, in the same or another language, especially in other countries, is justifiable and can be beneficial provided that the following conditions are met.

        1. The authors have received approval from the editors of both journals (the editor concerned with secondary publication must have a photocopy, reprint, or manuscript of the primary version).

        2. The priority of the primary publication is respected by a publication interval of at least 1 week (unless specifically negotiated otherwise by both editors).

        3. The paper for secondary publication is intended for a different group of readers; an abbreviated version could be sufficient.

        4. The secondary version faithfully reflects the data and interpretations of the primary version.

        5. The footnote on the title page of the secondary version informs readers, peers, and documenting agencies that the paper has been published in whole or in part and states the primary reference. A suitable footnote might read: “This article is based on a study first reported in the [title of journal, with full reference].”

      3. The rules in biomedicine from:

        http://www.icmje.org/publishing_4overlap.html

        Acceptable Secondary Publication

        Certain types of articles, such as guidelines produced by governmental agencies and professional organizations, may need to reach the widest possible audience. In such instances, editors sometimes deliberately publish material that is also being published in other journals, with the agreement of the authors and the editors of those journals. Secondary publication for various other reasons, in the same or another language, especially in other countries, is justifiable and can be beneficial provided that the following conditions are met.

        1. The authors have received approval from the editors of both journals (the editor concerned with secondary publication must have a photocopy, reprint, or manuscript of the primary version).

        2. The priority of the primary publication is respected by a publication interval of at least 1 week (unless specifically negotiated otherwise by both editors).

        3. The paper for secondary publication is intended for a different group of readers; an abbreviated version could be sufficient.

        4. The secondary version faithfully reflects the data and interpretations of the primary version.

        5. The footnote on the title page of the secondary version informs readers, peers, and documenting agencies that the paper has been published in whole or in part and states the primary reference. A suitable footnote might read: “This article is based on a study first reported in the [title of journal, with full reference].”

  8. Here is a study from 1997

    http://www.bmj.com/content/315/7109/635.abstract

    Impact of covert duplicate publication on meta-analysis: a case study

    BMJ1997;315doi: 10.1136/bmj.315.7109.635(Published 13 September 1997)
    Cite this as:BMJ1997;315:635

    Design: Systematic search for published full reports of randomised controlled trials investigating ondansetron’s effect on postoperative emesis.

    Conclusions: By searching systematically we found 17% of published full reports of randomised trials and 28% of the patient data were duplicated. Trials reporting greater treatment effect were significantly more likely to be duplicated. Inclusion of duplicated data in meta-analysis led to a 23% overestimation of ondansetron’s antiemetic efficacy.

    It’s an old story, which does not look like it is going away any time soon.

  9. The ‘innacurate science’ message is very important for the entire community but isn’t a big issue that much of world science is also ” liv[ing] in obscurity in Medline and other databases.” http://bit.ly/vAxLVn

    If the valuable product of ‘accurate’ observation and insightful observation is not easily available to fellow scientists & innovators the rate of real knowledge creation does not match the effort. In an information age we must have this content easily available to researchers as part of their daily work. If it’s easier for researchers to read their Twitter feed on last night’s episode of their favourite show than than access the information they need to make their next advance then we have a bigger problem.

Leave a Reply

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