Retraction Watch

Tracking retractions as a window into the scientific process

P values: Scientific journals’ top ten plagiarism euphemisms

with 50 comments

labtimes1113The other day, we nominated a phrase in a retraction notice for the prize “of most-extra-syllables-used-to-say-the-word-plagiarism” because a journal decided to call the act “inclusion of significant passages of unattributed material from other authors.”

That lovely phrase can now be added to our list of best euphemisms for plagiarism, which we highlight in our most recent column for LabTimes. There, you’ll find such gems as “unattributed overlap,” “a significant originality issue,” an “approach,” and an “administrative error.”

As we write:

Science is supposed to be about cold hard truths, about tested hypotheses and convincingly small p values. So why are journal editors so afraid of the p word?

Add your own favorites!

 

 

Written by Ivan Oransky

November 29th, 2013 at 9:30 am

Comments
  • msaxschizzofunk November 29, 2013 at 10:10 am

    “We, the named authors, hereby wholly retract this Catalysis Science & Technology article, due to
    significant similarity with previously published work.
    Signed: B. Pawelec, R. M. Navarro, J. M. Campos-Martin and J. L. G. Fierro, October 2012.
    Retraction endorsed by Jamie Humphrey, Editor, Catalysis Sciecne & Technology

  • Rolf Degen November 29, 2013 at 10:17 am

    “The authors retract this article as it contains large sections of text duplicated from previously published articles, a without proper acknowledgement, including:”

    http://link.springer.com/article/10.1007/s12029-013-9542-2/fulltext.html

  • Rolf Degen November 29, 2013 at 10:27 am

    “The editors were made aware of partial overlap between already published material and the manuscript entitled “Ammonium transport proteins from Archaeoglobus fulgidus” by Daniel Cebo, Jörg R. Aschenbach and Martin Kolisek published online in the Journal of Physiology and Biochemistry.”

    http://link.springer.com/article/10.1007%2Fs13105-012-0205-8/fulltext.html

  • Rolf Degen November 29, 2013 at 10:32 am

    “This article has been withdrawn at the request of the Editor-in-Chief of European Journal of Nuclear Medicine and Molecular Imaging owing to the unexplained close similarity of some passages to parts of a previous publication [Rushing SE, Langleben DD. Relative function: Nuclear brain imaging in United States courts. J Psychiatry Law 2011; 39 (winter): 567–93].”

    http://link.springer.com/article/10.1007%2Fs00259-013-2613-6/fulltext.html

    • CR November 29, 2013 at 10:52 am

      “unexplained close similarity of some passages to parts of a previous publication” by far the best one

  • Rolf Degen November 29, 2013 at 11:17 am

    “We have been made aware of the fact that a large proportion of the Introduction section and corresponding references of the title paper [1] had been copied verbatim from an earlier paper by Hamdyand Gamal-Eldeen [2], and further, during our investigation it has also come to light that most of the text in question had been lifted unchanged from an even earlier review paper by a different group [3].”

    http://www.mdpi.com/1420-3049/18/9/11001

  • Matthew Hankins (@mc_hankins) November 29, 2013 at 11:34 am

    “a substantial part of the literature review (Section 2.1) contains extracted passages without quotes or sufficient changes to the original text and lacks the necessary acknowledgement in the reference section”

    http://www.sciencedirect.com/science/article/pii/S1071581907000043

    • JATdS November 29, 2013 at 12:23 pm

      Dear Rolf, please keep listing them as you have and as you find them, to expose, publically, the total hypocrisy of these editors (and publishers). On one hand, they claim peer review that promptly rejects a paper for data that is not statistically significant, usually a P = 0.01 or 0.05 value, but when it comes to plagiarism, no-one wants to put a clear value to the figure. Not even COPE or the ORI. What are these “ethical” bodies afraid of?

      I had already challenged a Taylor and Francis elite journal’s editor in chief, when he detected approximately 300 words of similar text (about 30-50 words were in fact accidentally identical) in a massive review of over 20,000 words. The manuscript was rejected immediately on the grounds that “significant plagiarism” was detected using “commercial software”. Elsevier has this delightful marriage with iThenticate, but it is evident that iThenticate (and Elsevier) are not doing the job right, either. I then indicated to the editor that this can on occasion happen, especially when one deals with multiple authors of multiple countries, and that immediate rejection is wholly unfair, and illogical. The detected plagiarism amounted to exactly 1.7%, or a p value of 0.017, and I claimed that the rejection was unfair because the editor brutally rejected the paper as if we were hard-core criminals who were intentionally trying to plagiarize (as if this is the intention of scientists, although I do admit that alot of scientists for whom English is not the native language commit alot more plagiarism, especially in the “predatory” OA journals listed on Beall’s http://www.scholarlyoa.com). I indicated to the editor that when such cases exist, that the correct and fair thing for an editor to do would be to inform the authors that plagiarism was detected, where it was detected, to provide the formal reports and to give the authors an opportunity to correct the submission. This would constitute the first step of the technical review even before peer review begins. Any author would be more than happy to do so to avoid a disastrous retraction down the road.

      So, in fact, I claim that many retractions are the result of BAD editorial work and scrutiny by the publishers, the enforcement of rules that are not fairly quantified, and the forceful implementation by the publishers of the retraction as “martial law” on scientists without any reason, or logic. Heibus corpus has been suspended in science, and scientists have only one option: to fight back in instances where there is proof that editors have been unfair, unprofessional or unspecific.

      Only a seasoned scientist will observe how the online submission systems of most journals published by Elsevier, Springer, Wiley, and Taylor and Francis are starting to sound more like interrogations by a spy agency then a manuscript submission. In some cases, it takes about 30 minutes or more of preparation to just submit your manuscript to a journal. In fact, last week, I submitted a paper to a journal (can’t reveal the name of the journal or publisher because the paper is now in review, so I do not want to jeopardize my review process) in which ONE of the submission pages had no less than 25 questions. It felt like harassment to me by the publisher. So, the p value related to rejections and to retractions is a serious plague right now, and will rapidly become an embarrassment to editors and publishers (as already wonderfully displayed by Rolf Degen above in a handful of cases). Although, I should point out to Rolf that it is usually not the editors who writes those retraction notices. It is the legal teams (with their “ethicists”, lawyers, and other “experienced” folk) of the publishers who are pushing this fake narrative down our throats. I suspect (I have no proof) that editors have to liaise with the publisher (i.e., the team I list above), reach a compromise on tactfully worded retraction note so that the FULL blame is passed onto the authors, who are then FORCED to sign these pathetic, illogical and unquantified retraction notices.

      IT IS TIME FOR SCIENTISTS WHOSE PAPERS WERE RETRACTED BASED ON THIS NONSENSE TO START FORCING THE PUBLISHERS TO RETRACT THE RETRACTION NOTICES THAT ARE VAGUE, UNQUANTIFIED, OR UNCLEAR.

      QUANTIFY PLAGIARISM. This will be the ONLY way to resolve this issue. I have proposed a strict p = 0.01 value, but this should not amount to an immediate rejection. Authors MUST be given a fair opportunity to respond to an editorial claim, and to correct the error. The current publishing model being enforced aggressively by Elsevier, Springer, Wiley, and Taylor and Francis, supported by “ethical bodies” like COPE are implementing rules that make sense to marketing managers and people without a sound scientific background and resume. Scientists are (in several cases) being treated as guilty subjects even before any proof has been provided. And, even if found guilty of “some significant level of plagiarism”, it is never quantified.

      If you are a scientist, and if you have been affected in one way or another been discriminated against by this p story, or the lack of it, it is time to step up to the plate and deliver your best shot to the editors and publishers.

      • R. Grant Steen November 29, 2013 at 12:54 pm

        “I claim that many retractions [for plagiarism] are the result of BAD editorial work and scrutiny by the publishers.” That is simply absurd! This “logic” amounts to blaming the victim. Plagiarism is the result of sloppy writing or intellectual dishonesty (or both). And to put some kind of a p value on the extent of copying is nonsensical.

        • JATdS November 29, 2013 at 1:24 pm

          Grant, can you give us an update on the special issue, “Misconduct in Scientific Publishing,” you’re editing for MDPI (http://www.mdpi.com/journal/publications/special_issues/scientific-publishing)?

          Never forget that there exists a darker, silent, and angry “Gotham city” of scientific “plebs” that have been abused by the system. I assure you that a war is taking place. An ideological war between those, on the one hand, that think they are ethical, and impose rules on others, sit back and glee on their pseudo-achievements when their pockets are being padded, and their repertoires are being supported by “charitable” ethical bodies (that charge members astronomical fees), all at the expense, of course, of the intellectual achievements of scientists. And, on the other hand, those who know that pseudo-ethics are being thrust upon them by the very same elitists. What I describe above is based on experience. Long, tough, brutal years of experience, of continuous abuse by editors and publishers (of course not all) who have fake value systems, artificially propped rules, non-sensical and unquantified value systems, and little respect for the opinion and the rights of scientists and authors. We pout the food on the table for the marketing managers and CEOs of these publishers, our intellect feeds the system that keeps their families warm, comfortable and safe. Yet, we, the average “Joe-the-Plumber” scientist, are treated with asbolute disrespect (on select occasions). Allow me to make a frank suggestion. If you want to have a nice, successful special issue, why don’t you include case studies? Case studies represent the sewer-base truth of abuse that takes place? Why can we, as I have already attempted to do, not submit evidence-based case studies to your “special issue”, if the facts are provided as clear documents?

          Going back to your comment, I disagree with your last sentence. I can assure you that the issue is simple. 99% of scientists around me would respect a rule that is quantified. If a publisher were to implement a clear rule, with two phases, as I keep suggesting on this blog over and over and over again, then 99% of the problems would be resolved. The two steps are:

          STEP 1: The publisher conducts, as a first stage of the peer review process, a “technical check” for duplication and plagiarism. That can involve commercial software like iThenticate, or it could involve something like Google or Google Scholar. Ideally, it should involve both, and also checks against the major data-bases. In the case where duplication is detected, then the case is simple: immediate rejection, with the possibility of contacting the authors’ institute.

          STEP 2: The publisher implements a clear, quantified rule. A maximum of 5% plagiarism is allowed at the time of submission. So, a p value of 0.05. However, where plagiarism is excessive, for exmaple, 10 or 20%, so a p value of 0.0 or 0.2, then rejection is immediate. However, where plagiarism, or self-plagiarism, is < 5%, the editor/publisher indicates clearly where such plagiarism occurs, provides the authors with an opportunity to explain why such plagiarism occurs, AND, most importantly, gives them a fair opportunity to correct it, so that a fair, balanced and impartial peer review can then take place.

          No publisher has a fair, quantitative and logical system as I am proposing here.

        • C. Clark December 2, 2013 at 9:17 am

          Agree with R. Grant Steen. Also that was a strange calculation: 300/20,000 = 1.5% (not 1.7), which has nothing to do with significance (p value). Altogether, JATds hurt his case rather than helped it. I’ve recently submitted several articles to Elsevier and Oxford University Press journals and the only onerous part was finding and providing potential reviewers’ contact information. As I don’t reuse previous work (mine or others’) to bolster my new submissions, I haven’t experienced any of the issues that he complains about.

          • Matt December 20, 2013 at 12:12 am

            Plus, the reasoning seems to be “I didn’t plagiarize the other 19,700 words so I should be given a pass.”

            This kind of misses the point. Science depends in large part on trust in the author’s scholarship and personal integrity. Lifting even ONE passage from another’s work and representing it as one’s own is a sig that that author cannot be trusted. Period.

      • Tim D. Smith (@biotimylated) November 29, 2013 at 11:51 pm

        You seem to be using “p-value” to mean “fraction” which is not consistent with any statistical method of which I am aware.

        Given a large corpus of biomedical literature presumed to be independently authored, it is probably sensible to design a bootstrapping test to determine the likelihood of a phrase of a given length occurring in two documents by innocent chance, and hence a genuine p-value. You could even get fancy and control for the frequency of subsets of the complete passage in the literature, since duplicates of “densitometry was performed in ImageJ” are probably far less suspicious than e.g. “rhinoceros serological Piltdown Man”. I’m sure someone at Turnitin or iThenticate has already built this. I don’t think I’d expect the results to be surprising, though.

  • Rolf Degen November 29, 2013 at 1:11 pm

    The problems with plagiarism could be remedied to a large part if some good working, easy to use plagiarism software were available for free. The programs I could test here in Germany were not up to the job. It must be a Google service, they have the best data available. Now, I know what people will say: The would-be cheaters would abuse that stuff to cover their tracks. But look, it would raise the bar for cheaters, because they would have to labor hard to change their sentences to escape detection. It would make plagiarism so costly that it would be easier for everone to formulate their own texts.

    • foobar November 29, 2013 at 2:19 pm

      No. As the author of copy, shape & paste has forcefully shown all over again, plagiarism software does not solve anything; at best, it provides only the initial alarm bell.

      As for the title of this entry, I am not sure whether I follow. We already went through this discussion of p-values and whatnot, and from what I can conclude, none of the participants were that eager to support the claims that there would be something inherently bad about critical values and such. Bad science is not plagiarism or misconduct; it is just bad science! And with all due respect, the evaluation of bad science is neither the topic of this blog nor in the expertise of the authors of this blog. And nor can I judge but only the tiny corner which I am familiar with.

  • Vincent November 29, 2013 at 2:06 pm

    “The contents of this chapter has [sic] been removed for legal reasons”
    Cambridge History of Medieval Philosophy

    http://universitypublishingonline.org/cambridge/histories/chapter.jsf?bid=CBO9781139095440&cid=CBO9781139095440A018

    • JATdS November 29, 2013 at 2:35 pm

      Dear Sylvain Bernès, thank you for that listing of three Elsevier papers, which serves as a perfect entrée for my next comment. And it also feeds into a response I have to Rolf Degen about free plagiarism detection software.

      Free plagiarism detection software is one solution, as I have long advocated. If so much free-ware is already available, then why can’t plagiarism-detection software also be made available? If such software existed, then scientists would have absolutely no excuse to submit a paper with plagiarized text to a journal. This would resolve a problem before it became cancerous. However, this would only resolve the problem fully if it also involved a fair element by the publishers that would involve a quantifiable system as I suggest above. A scientist that has access to powerful, industry-accepted free plagiarism detection software could only be considered to be a stupid scientist if they were to submit a paper without checking first. So, in that case, rejection (submission) or retraction (post-publication) are valid, and fair. However, since no free software occurs, and since not a single publisher has a quantifiable system in place to indicate an acceptable level of plagiarism (the elusive p value), the current system of rejections and retractions based on plagiarism is, I believe unfair and fraudulent.

      Allow me to give my personal interpretation why free software is going to face such an uphill battle and why I am going to get a lot of thumbs down). Free software is a serious challenge to the “ethics business model” now being implemented by most main-stream publishers that involves this rather incredible link: PUBLISHER + “ethics” software (e.g., iThenticate) + “ethics” society (e.g., COPE) + “ethical” editors (i.e., puppets) + “ethical” super-linked systems (e.g., Crossref) + “ethical” Big Brother control (e.g., ORCID, Elsevier’s “consolidated profile”, automatic online submission systems, etc.) + “ethical” special issues + “ethical” rules leading to retractions. All versus the “unethical” scientist meant to bedevil science and to squeeze as much profit from scientists and their gullible institutes and government ministries. It is a war taking place, as I have indicated already. An ideological war in which science is portrayed, eventually, as the evil, the scientist as the bearer of all evil, and the publisher, as the savior. An almost biblical parallel.

      I elucidate further. The plot becomes more complex.

      Currently (and I stand corrected here), the most powerful and most frequently implemented software available is iThenticate (http://www.ithenticate.com/). What iParadigms LLC will not inform you is that Elsevier was involved, in the first phases of development of this product. When I initially detected this on their web-site, I launched an official complaint. That information was wiped off the web-site almost immediately. Contact iParadigms LLC about this. Ask them why they don’t serve the scientific community

      And further. The plot now becomes sinister.

      So, why would Elsevier (and primarily its parent company Reed-Elsevier) want to be involved in developing (and implementing) the world’s most powerful plagiarism-detection software? Well, that is a long story, but let me summarize it with a few facts:
      a) LexisNexis, most likely the most powerful legal firm on the planet, would give Elsevier that legal and “ethical” framework to demand, by force, legally or course, a moral high ground globally. Of course, this is not without its fair of hiccups (http://www.independent.co.uk/news/business/news/fraud-affected-310000-customers-reed-admits-6148381.html);
      b) http://www.reedelsevier.com/OurBusiness/Pages/Home.aspx: A powerful grip on global science, global publishing, global legal resources, and dozens of other businesses would cement its business model and solidify its “moral” and “ethical” standing;
      c) The sudden branding of a very ethical publisher (as if this were ever a problem in their 300+ year history) in late 2012 (I claim in response to my claims of the lack of ethics by Elsevier and its parent company Reed-Elsevier) with plastic, glossy, brochure-like sites like this (http://editorsupdate.elsevier.com/);
      d) This revolving door of power meant to consolidate more power (http://www.elsevier.com/editors/perk/plagiarism-detection), all in the name of “ethics”;
      e) A fuzzy mingling of power, human “ethics” and charity: http://www.bookaid.org/ with the true objective of expanding its geopolitical power into the African continent, the future of global resources. To seed the idea of “cultural capitalism”, see Slavoj Zizek (http://www.youtube.com/watch?v=hpAMbpQ8J7g) (a little old, but never outdated)
      f) The use of PACs and SUPER-PACS, a legal way to bribe the US Government (http://www.opensecrets.org/orgs/summary.php?cycle=2012&id=D000028557), to ensure that the “ethics” model becomes the law, and that efforts like SOPA, PIPA, etc. become laws to stifle dissent by scientists against that imposed model. By padding the pocket of Barack Obama (http://influenceexplorer.com/organization/reed-elsevier-inc/2ba2255d154f4e3d9d04e8952ef32d62), it is quite obvious that a call for a legal implementation of the Elsevier “ethics” will be met with a hand-shake. Observe CAREFULLY the legal bills that Elsevier is supporting in the US Congress to understand that plagiarism, ethics and plagiarism detection is one way of cementing power.

      Free plagiarism detection is against the current corporate profile that has full government support. This is termed dirty ethics, or, in plain English, unethical.

      Astute scientists – not to be confused with smart scientists – unfortunately seem to be low in number. And those that are willing to stand up to this clear power grab, which treats scientists as “ethical” dispensables, are extremely few in number. Because their voices are silenced by the “system” that is being put into place.

      • tekija November 29, 2013 at 4:03 pm

        I checked some of the links that you thoughtfully included. What editorsupdate says about using plagiarism detection softwre, here iThenticate, has a ring different from your interpretation:

        “Tips for interpreting iThenticate results
        Human interpretation is crucial to differentiate between:
        paragraphs or sentences copied from properly referenced sources;
        text copied from the author’s previous works (often in the Methods section); and
        paragraphs or sentences copied from improperly or unreferenced sources.
        Similarities discovered in the Results/Discussion sections can be more concerning than those found in Intro/Methods.
        You should become suspicious if you discover:
        Similar strings of sentences or small paragraphs. One may not be an issue, but several could signify a problem.
        A couple of paragraphs containing identical material. This may indicate improper reuse and should be carefully checked.
        As much as a full page of matching material. Proceed with extreme caution!
        Ethics cases can be less obvious than they appear so whenever in doubt, check with your publishing contact to make sure you follow due diligence in any accusation of research or publishing malpractice.”

        • JATdS November 29, 2013 at 4:09 pm

          Tekija, I would view Elsevier’s editorupdate (http://editorsupdate.elsevier.com/) in light of the relationship between Elsevier and iThenticate.

          I forgot some important points with links:
          1) http://www.ithenticate.com/quote-request (so why can’t iThenticate just put up the price of their software for everyone to see?). Will an Indian get the same price as an American scientist, for example? How can a “ethics” software company not even practice transparency about its pricing? Also strange is how there is no “About Us” page, no indication of how the company started, who is on the executive board, nothing transparent. Only good profits and the power claim of this decade: “1 in 3 Scholarly Journals Has Access to iThenticate.”
          2) If iThenticate is such a loving company, then why not create a software that is FREE for scientists? This could make them an instant global leader. Their revenues could then be made by donations. But unfortunately the true objective is profit, not the need to serve the “ethical needs” of the scientific community. Their aggressive marketing campaign will prove this.
          3) Crossref and the “Carbon Fiber” project (1993-1996): (initiated by Springer-Verlag, Wolters, Wiley, Elsevier, Academic Press, Blackwell Publishing: see, for example Craig Van Dyck who jumped from Springer to Wiley, and who represents this powerful revolving door of elites in the global publishing world): http://www.crossref.org/08downloads/CrossRef10Years.pdf A very telling quote: “Eric Swanson recalled the immediate reaction to the presentation. “A number of us said, ‘This is very important — this must be done,’” he said. “As Pieter and I walked out of the room, we said to each other that if Nature didn’t achieve it, we had to find some other way to do it.”” “The four met for dinner in April 1999 at Monzu, an Italian restaurant then located at the corner of Mercer Street and Prince Street in New York’s Soho neighborhood. As it turned out, ACS only participated in the project through June, but the evening had given it a codename—thenceforth, Monzu.” The history of the power grab for the control of “ethics” is astounding, with the pressure applied by the elite publishers, forcing the hand of other smaller publishers. The most telling is, however, the current board: http://www.crossref.org/01company/05board.html (I don’t see a single scientists there). This company is not about meta-data it is about securing power through the control of data. Think about it, is a paper without a DOI a respectable paper anymore?
          4) Is it true that “members” of CrossRef only pay 29 cents for a check on iThenticate? OK, so how much does CrossRef membership cost? This sick revolving door policy of pay me peanuts but pay my partner your gold is the gradual corruption of the current publishing model. This carefully crafted marriage between the top half-dozen global publishing giants, COPE, CrossRef, DOI, ORCID and goodness knows what else is DANGEROUS. This, to me, as a scientist, does not breathe trust or confidence. It breathes distrust that the objectives are simply to secure profits, power and influence. Not in a million years will you ever convince me that your global elitist policies (I guess this could fuse in nicely with the New World Order thought-stream) are for the “ethical” good of scientists, or for the love of science.
          5) IEEE (http://www.ieee.org/publications_standards/publications/rights/crosscheckmain.html) which is “The world’s largest professional association for the advancement of technology”. Look carefully what this “association” is linked to: CrossRef and iThenticate.
          6) A quick update (also see http://journalseek.net/publishers.htm): Elsevier (>12 million articles on sciencedirect.com; they probably crossed the 12 million mark in October or November, much faster than I expected); Springer Science+Business Media (SpringerLink: “8,348,055 scientific documents at your fingertips”); Wiley (“Wiley Online Library delivers access to over 4 million articles from 1,500 journals, 8,000 books, and hundreds of reference works, laboratory protocols, and databases to a global audience of 16 million scientists and scholars.”); Taylor and Francis >1600 journals). The effort by “predatory” open access journals (e.g., at Beall’s http://www.scholarlyoa.com) is not only to break away from this clasp on publishing power globally, but to send a strong message that the grip on “ethics” is, ironically, not a successful business model (see OMICS group and Africa Journals Online (among others in tier 1 journals at http://journalseek.net/publishers.htm), which are listed as predators by Beall, making significant inroads into the elitist, traditional publishing model). Simply because as you implement more draconian forms of checking, revising and punishment, you create a wider and wider base of dissent, of grievances and of dissatisfaction simply because you lack three basic ingredients to keep your base: openness, honesty and transparency (fancy web-sites, glossy brochures, and carefully crafted, legally-assisted PR marketing-style responses should not be equated or confused with honesty and transparency).

          There is a lot more, of course, but we leave the continuation of this battle to wield power for another day.

          • Rolf Degen November 30, 2013 at 3:38 am

            iThenticate would immediately crash into nonexistence if Google would take on the job of offering plagiarism software at Google Scholar, for free as they use to. But there might be some irritation at first, as anyone could easily check the texts of everybody else. This might start with a phase of anarchism. And remember, Google has scanned 20 Million books from university libraries. There should be a lot of dissertations and other thesis among those. If Google would cross check those, we do not know what could happen.

          • Paul Brookes December 2, 2013 at 11:33 am

            Thanks JATdS, this is an excellent summary of the current power grab surrounding publishing ethics. The major problem as I see it, is no matter what the big-wigs in the corporate publishing boardrooms try to push through, the real battleground is at the level of individual EiCs and editorial boards. All this hot air means nothing if you have a rogue editor who doesn’t act according to the guidelines. Then there’s the problem that the guidelines don’t have “teeth”. Have you ever heard of a journal being thrown out from one of these organizations?

      • Debora Weber-Wulff November 30, 2013 at 4:15 am

        Your belief in the magical properties of software to be able to detect plagiarism is touching, but it just doesn’t work that way. Software can only give pointers to possible text similarities. The determination of plagiarism must always be left to a rational human being who is well-schooled in interpreting the results and continuing the investigation. As noted by a commenter above, I have been testing plagiarism detection software since 2004. My current, 2013 report, was published in September: http://plagiat.htw-berlin.de/software-en/test2013/report-2013/ and it is the same story:

        “So-called plagiarism detection software does not detect plagiarism. In general, it can only demonstrate text parallels. The decision as to whether a text is plagiarism or not must solely rest with the educator using the software: It is only a tool, not an absolute test.

        In particular, users must be aware of the false positive and false negative problems that all systems have. A university can and should make software available for their educators to use, but they should not use it as a general screening tool for all texts. If at all, general screening could only be reasonably used for first-year student papers.”

        The values generated by the systems are not absolute. Some systems add percentages, something that curls my toes. Percent of what? Characters, words, sentences, paragraphs? Some systems suggest a precision by giving 3 or even 4 decimal places. What do the numbers mean for changed words or phrases – which are still considered plagiarism. Other systems generate massive amounts of “possible sources” and when you begin sifting through them, you discover that most are not actually close to the text or even that the links given are no longer valid.

        Inculcating good scientific practice and public peer review (both pre- and post-publishing) are the most promising means of dealing with plagiarism. Oh, and using the P word where it is warranted would be a good step in the right direction.

        Shall we start a p-index for journals? What percentage of published articles is retracted? I’d love to see that data.

      • Matt December 20, 2013 at 12:24 am

        “If such software existed, then scientists would have absolutely no excuse to submit a paper with plagiarized text to a journal.”

        My perhaps naive expectation is that scientists CURRENTLY have no excuse for submitting a paper with plagiarized text. It seems pretty simple: don’t use others’ words as your own. Ever.

        Complaints about the cost of plagiarism detection software is a diversion.

  • Matthew Hankins (@mc_hankins) November 29, 2013 at 4:12 pm

    “In breach of copyright, the authors in their review article adopted phrases from an article previously published by T. J. Aitman in the British Medical Journal”

    http://www.smw.ch/docs/archive200x/2003/45/smw-10515.html

  • Richard Goldstein, Esq., Meyer Connolly, et al., Boston, MA November 29, 2013 at 4:47 pm

    This may be a minority view but as a lawyer who has represented many scientists and students accused of plagiarism, I urge readers not to conflate the “inclusion of significant passages of unattributed material from other authors” (yes, that it is a mouthful) with plagiarism. These kinds of awkward and wordy ‘euphemisms’ may be issued because there has not yet been (and there may never be) a determination the ‘copying’ or ‘duplication’ is, in fact, plagiarism. Plagiarism is research misconduct, which requires a certain state of mind. Not all unattributed copying or duplication is done with that level of intent.

    • Allison (@DrStelling) November 29, 2013 at 5:59 pm

      ” Not all unattributed copying or duplication is done with that level of intent.”

      Esp. when they’re PhD students who are still learning how to do proper literature searches. Here, the fault lies at the door of the PI or graduate school who failed in their educational duties and responsibilities to their students. (I was doing full lit searches in pubmed in undergrad, so don’t look at me.)

      Granted, half the time it’s because the PI is so busy writing grants; there’s no time to “mentor” properly.

      • Tom November 29, 2013 at 6:09 pm

        I might be willing to cut the student some slack if they are from another country or if they are not otherwise familiar with modern standards of publishing and attribution. In these cases, it might be expected that such a student would be unfamiliar with plagiarism as a concept, or something similar. But for many (most?) PhD students, it can reasonably be expected that they were taught to not plagiarize from a young age.

        I just can’t believe that a PI’s duty now encompasses such a basic (and trivial) task as teaching a student to not plagiarize.

        • Allison (@DrStelling) November 30, 2013 at 10:54 am

          I’ve TAed and mentored a fairly large number of undergraduate students at this point- and I have seen many who just have no clue how to even begin a literature search. Less so in Germany. From what I can tell, this might be an American undergraduate thing. Once I showed them how, they were fine! But it feels like they are not learning this in the lower-division science classes anymore……

    • Tom November 29, 2013 at 6:03 pm

      I absolutely have to disagree. Plagiarism is not, and never has been, “research misconduct.” I am sorry if it has come to be defined in that manner in legal-ese. Plagiarism is simply the lifting of significant passages of text or other published work from another source–exactly the type of behavior being discussed here. It is precisely this weasel-worded, legal mumbo-jumbo that muddies the water and obfuscates meaning. When (and why) did we abandon directness?

    • Debora Weber-Wulff November 30, 2013 at 4:24 am

      I beg to differ, Richard. Intent has nothing to do with plagiarism. A text is plagiarized if it includes significant passages of unattributed material from other authors. Full stop. There may be reasons why the text is this way that would invoke different levels of sanctions, but the text in and of itself is a plagiarism. Trying to prove intent (or rather, trying to excuse oneself out of plagiarism by stating that it was not intended) is not possible, as we can’t really see into people’s heads. Let’s understand: By submitting a paper to university or to a journal you are stating that you wrote the paper and attributed the words and ideas of others properly. If the text turns up plagiarized, it needs to be retracted. If a person has a number of retractions for plagiarism, maybe some sort of remedial education is warranted.

      • Richard Goldstein, Esq., Meyer Connolly, et al., Boston, MA December 3, 2013 at 12:32 pm

        Debra and Tom: I disagree (without trying to be disagreeable). This may sound overly technical (and ‘legalistic) but ORI defines plagiarism as a form of misconduct. 42 C.F.R. section 93.103(c). Misconduct, under ORI’s definition (which applies to every institution that accepts federal research money) has to involve intentional, knowing or reckless conduct. 42 CFR section 93.104. Honest error (or what I would call negligence or carelessness) is not misconduct. We may be arguing over definitions because I agree that plagiarism (no matter how it is defined) may warrant a retraction (or an acknowledgement of the author). All I am saying is that a scientist should not be disciplined by their institution or ORI (or have their errors exposed to the whole world) merely because there was an “unattributed copying.” Unfortunately, the institution (or ORI) must ‘see into people’s heads’ if they are going to impose a sanction of some kind. I also noted Debra’s comment on the limits of anti-plagiarism software and I couldn’t agree more. Candidly, I think it supports my point that copying an textual similarities don’t prove plagiarism. In closing, I would add that, under the current system, it is other scientists at the same institution (and not lawyers or non-scientist judges) who generally decide what’s plagiarism and what’s not.

        • frank December 3, 2013 at 8:27 pm

          ORI lacks the authority to define age-old words for the rest of us. You seem to be saying that “unintentional plagiarism” is a contradiction. I don’t think that’s true on the usual meaning, but I could be wrong.

          • Richard Goldstein, Esq., Meyer Connolly, et al., Boston, MA December 4, 2013 at 8:39 am

            This may be a definitional issue (legal v. scientific vernacular). For the purposes of a misconduct proceeding, I have argued successfully that plagiarism can never be unintentional. It may merit the approbation of one’s peers but it isn’t misconduct in the legal sense.

    • CR November 30, 2013 at 4:33 am

      I think that the legal issue with directly dealing with plagiarism is that it is so easy to catch, so common among many authors, and so easy do demonstrate that there is great pressure (bunch of people, including big shots) that want to make it an “innocent mistake”. Or should we call it “a white lie”. Should I remind that all other categories of crimes can be committed “without intention to harm”?
      Lawyers are interested in satisfying their colleagues, and not correcting society. Unfortunately science depends strictly on a working system, and not on happy clients, to continue generating knowledge on the real world.

    • Rolf Degen November 30, 2013 at 5:15 am

      If it is unintended, it is “Cryptomnesia”, kind of a neurological dysfunction. May have happened to George Harrison with his “My Sweet Lord”.

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

    • JATdS November 30, 2013 at 7:23 am

      If this be the case, as a lawyer, how do you defend the use of the word “significant”? How do lawyers ever know the true intent of their clients, or defendents? It’s not as if people don’t lie under oath, so the best way to equate the euphemism is with the word plagiarism. Scientists know what is wrong, fundamentally. If you copy a text, word for word (or almost word for word), and “copied” text exceeds 5% of the total text, this is not a coincidence. Set the limit at 5% of similar text and scientists will work this out themselves. At minimum it is sloppy attention to detail, and deserves a rejection. Where the editors fail to detect such a percentage of text overlap, well, that just shows that the editors, editor-in-chief and publisher were sloppy. And in this case, it is amazing, the authors get slapped with the bill: a retraction. There is no need for euphemisms with scientists. Things are pretty black and white for us. It’s obscure laws and skewed definitions offered by lawyers (who are not fundamentally scientists) and by “ethics societies” like COPE who are just puppets to the system, propping up an artificial economic skeleton to a system that does not require two skeletons. I call on a logical limit of 5% similarity using any software (free or commercial) to detect it, pre- or post-publication (the timing of the crime is not an issue). But, in this day and age, with the technology and tools that we do have available in our hands, there is now NO excuse by editors and publishers to NOT detect plagiarism. Smple. Therefore, any retraction that appears from now on (or in fact from about 2011 onwards) should also carry a note that the editors failed their duties and that the publisher also failed its responsibility. As I say, it’s so black and white to scientists. Is it not so clear for lawyers?

  • Praveen Chaddah November 29, 2013 at 10:19 pm

    This is from PHYSICAL REVIEW B71, 229902, 2005. “A portion of the text of our paper, including most of the second column of page 2, and the first full paragraph of page 3, is very closely correlated with text which also appears in reference 10.” Followed by “In addition, two other references to papers by those authors, which appear in the correlated portion of text in reference 10, were omitted from our paper.” And Physical Review (and American Physical Society) is not accepting my suggestions that Errata reporting such ethical misdemeanors be ATTACHED to the pdf file of the main paper, as the corrections are not read.

  • CR November 30, 2013 at 7:22 am

    http://www.tandfonline.com/doi/full/10.1080/00032719.2013.839299#preview

    “This constitutes a breach of warranties made by the author with respect to authorship” — also quite a pompous, long one!

  • ABrown December 1, 2013 at 4:12 am

    http://www.washingtonpost.com/world/the_americas/expanded-panama-canal-sparks-race-to-be-ready-for-bigger-cargo-ships/2013/01/12/f3c85d52-5785-11e2-8a12-5dfdfa9ea795_story.html

    Good one in the Washington Post…

    “An earlier version of this article included material borrowed and duplicated, without attribution, from Environmental Health Perspectives, a monthly scientific journal. In that earlier version, four sentences were copied in whole or in substantial part from “Progress and Pollution: Port Cities Prepare for the Panama Canal Expansion,’’ published on Dec. 3 and written by Andrea Hricko, a professor of preventive medicine at the Keck School of Medicine of the University of Southern California.”

    It was just borrowed though!!! We gave it right back afterwards!!!

  • Chris Siple December 1, 2013 at 5:24 pm

    To be fair, most of the examples aren’t really euphemisms. They lay it on the line as to exactly why the paper was retracted, describing the specific problem in detail. A euphemism would be an attempt to fudge or sugar-coat the offense, as some of these examples do. The term ‘plagiarism’ is less precise and has legal implications that may or may not apply in all cases too.

  • Post a comment

    Threaded commenting powered by interconnect/it code.