Retraction Watch

Tracking retractions as a window into the scientific process

Why plagiarism is such a problem for German PhDs: Q&A with Debora Weber-Wulff

with 15 comments

Debora Weber-Wulff, c. 2015 Nina Zimmermann

Debora Weber-Wulff, c. 2015 Nina Zimmermann

Why do so many PhD students publishing their medical theses in German resort to brazen plagiarism, even copying from people in their own research groups? We’re pleased to present a Q&A with Debora Weber-Wulff, based at the University of Applied Sciences HTW Berlin in Germany. She recently published a case study for the Council of Europe that shows a surprisingly high number of cases of plagiarism in medical PhD theses submitted to German universities, as well as a few in other European countries. Weber-Wulff is a member of the VroniPlag Wiki, a group of German-language scientists who have been scanning for — and publicly tracking — cases of plagiarism. They’ve published documentations on more than 155 cases so far, and begun investigations on over 200 more, including some very high-profile cases. We talked to Weber-Wulff about why plagiarism is such a problem in German medical PhD programs.

Retraction Watch: You note that German medical dissertations aren’t taken seriously, relative to many other countries. Why is that?

Weber-Wulff: Many of the medical dissertations are quite short, often simple statistical calculations on given data, or very shallow investigations. Scientists who have spent years in the lab or humanities academics who spent much time in libraries working on their dissertations tend to look down on medical doctorates as not being “real” research.

Perhaps it helps to understand that many German medical dissertations are written during medical studies and not after completion. Since the studies are grueling, students have little time to devote to research. Many look to save as much time as possible. Perhaps the advisor, also busy with taking care of patients, just says “have a look at what I’ve published and the medical dissertations I advised before you.” They brazenly misunderstand this as: Take those papers as a template, and just substitute your own data. It is, however, vital for medical research that students learn how to do research themselves, how to understand statistics, and how to communicate their results, giving credit where credit is due. They do a disservice to their readers if they do not make clear where materials — no matter if overview of the literature, methods, or discussion — was previously published.

RW: When looking for plagiarism as a member of VroniPlag Wiki, it’s not as simple as running a paper through a detection software, such as Turnitin. Can you briefly describe your method?

Weber-Wulff: Generally, someone will have already found a text parallel with a previously published text, so this text is compared to the paper in question in order to find more portions of text taken from this source. There is a simple tool that compares two texts and marks all of the text parallels. If a text is not available digitally, it can be easily scanned and the text extracted using standard tools.

Otherwise, a good place to start is to look at the bibliography. Are the references consistent? Is there a strange entry that one could google to find a source? Then one chooses a page more or less at random and reads it. Does the tone change? Is one paragraph in perfect grammar, the next one full of spelling errors? Start googling the perfect grammar. Google Books is another good place to look, if the snippets look promising, then a book can be ordered from the library. Or one can take all of the references given in the paper, digitize them, and compare them with the dissertation. It is strange, but many people put the sources for their plagiarism in the reference section, but do not use quotation marks, although they have taken text verbatim and extensively.

RW: What have your results been so far?

Weber-Wulff: One can’t say anything about the rates of plagiarism, as the sample was not representative and one can never determine that a paper is plagiarism-free. It can only be demonstrated that there is plagiarism by presenting the plagiarism together with a source. What one can say, is that far too much plagiarism has been found. So far, VroniPlag Wiki has documented plagiarism in 88 medical and dental theses, as well as PhDs in in pretty much all subject areas from chemistry to theology.

The sources for the plagiarism were surprisingly often published by people from the same research group, but also from dissertations defended at other universities. Three dissertations were found that had plagiarized
text on 100 % of the pages, and on close examination one could see that the data was also faked. The numbers had been changed, but the percentages or the standard deviations were not adjusted!

We also found quite a number of cases that took basic medical knowledge word-for-word from the Wikipedia or other online sources without reference.

RW: Was there a particular case that was most memorable for you?

Weber-Wulff: Many are memorable, each in a different way. I think that the chain of plagiarisms found — a plagiarism of a plagiarism of a plagiarism — was the most memorable, particularly as the “authors” were dentists researching ape eyes. Two of the plagiarisms had identical pictures, supposedly of retina cells in different ape species.

A very troubling thesis copied 9 out of 61 pages verbatim and without reference from one article in the Wikipedia, but the university didn’t feel that it needed to take action on the case, as it wasn’t proven that this was done “on purpose”. I can’t explain how so much of the Wikipedia can be found in a thesis except by intentional copy & paste.

RW: The same member of VroniPlag Wiki who discovered the plagiarism in former Education Minister Annette Schavan’s thesis also recently documented extensive plagiarism in the German defense minister Ursula von der Leyen’s thesis. Is it different when a team member discovers plagiarism by such a high-profile person?

Weber-Wulff: The only difference is that the press reports only on high-profile persons when they are politicians. The plagiarisms that trouble me more are those by people who are currently working in academia, either as professors or as researchers. How will these persons be able to teach good academic practice, when they obviously were not able to demonstrate it in their own theses?

RW: Do you suspect plagiarism occurs more commonly in medical dissertations from Germany than elsewhere? If so, why?

Weber-Wulff: I have given up making estimates or predictions, because I tend to be wrong. I did not think that the amount of plagiarism that has been documented to date would be found! And there are so many more that have not been documented, because no one felt like documenting them yet. All I can state is that there is plagiarism, and far too much of it, unfortunately.

RW: The paper notes that the European Research Council does not accept a German “Dr. med.” as equivalent to a Ph.D., which limits medical graduates’ job opportunities in other countries. What can be done to address this problem, so German students can be on more equal footing with their peers?

Weber-Wulff: The German Council of Science and Humanities (Wissenschaftsrat) has repeatedly pointed out the problem. German students have to prove that they can do research, either by additional research experience or by having a habilitation accepted, in order to be considered for such grants.

Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our new daily digest. Click here to review our Comments Policy.

Written by Alison McCook

February 11th, 2016 at 9:30 am

  • Miguel Roig February 11, 2016 at 9:57 am

    Thank you very much for this contribution, Debra, and for your wonderful blog on this subject.

    I hope we can all agree that plagiarism of data, which to me amounts to data fabrication, is a far worse offense than most cases of plagiarism of text. With that said, I wonder what the penalties have been for those who have been found to commit plagiarism of text vs. plagiarism of data (or both!) in their theses. Also, I’d be interested in some additional general commentary about the context in which these offenses occur. For example, whether any type of guidance regarding plagiarism (and other forms of research misconduct) has been traditionally offered by German universities and whether there is any formal instruction on responsible research conduct mandated like there is in, for example, North American universities.

    • Debora Weber-Wulff February 11, 2016 at 10:37 am

      Hi Miguel,

      there has been, to date, very little formal guidance about good scientific practice given at German universities in any field. Students are pre-supposed to come to university knowing how to write. There are now a few more seminars offered on the subject, and some universities are making them mandatory for students seeking a doctorate.

      I have been badgering the universities in question to give me the results of their investigations. Usually, their rules state that they must inform me (if I am the one informing the university of the potential misconduct). For fear of lawsuits I tend to get general answers (“We withdrew 3 theses and censured 5 others.”) or a refusal to answer. This quite concerns me, as I find it a valid question to know what amount of blatant plagiarism is acceptable at a particular university.

      I wouldn’t want to start arguments about which kind of scientific misconduct is worse. I find them all to be problematic.

      • Miguel Roig February 12, 2016 at 10:22 am

        I wouldn’t want to start arguments about which kind of scientific misconduct is worse. I find them all to be problematic.

        Thanks for your reply, Debora. I can understand not wanting to classifying misconduct according to the seriousness of the infraction. Too often cases are too complex for easy classification. It’s just that when it comes to punishment, I would hate to see pure text plagiarism (an activity that perhaps is better characterized as scholarly misconduct as opposed to scientific misconduct) be given the same punishment as, say, data plagiarism; a type of offense that seems to me to be much more damaging to science.

  • Ciaran February 11, 2016 at 10:02 am

    “We also found quite a number of cases that took basic medical knowledge word-for-word from the Wikipedia or other online sources without reference.”

    As a computer scientist / mathematician that worries me slightly. I’m guessing most maths-heavy papers would trigger a similar accusation: there are only so many possible ways of writing out definitions. I’m willing to bet that if you asked a mathematician to write out the definition of, say, a graph, and then searched for it, you’d find the same definition almost word-for-word elsewhere.

  • Miguel Roig February 11, 2016 at 10:08 am

    Oh, and sorry for the misspelling your name, Debora!

  • Dean February 11, 2016 at 10:20 am

    There was no mention of penalties for those students. Are the universities turning a blind eye?

    The European Research Council is right. Instead of getting upset, Germany needs to change the requirements for its “Dr. med” degree, to include intense hands-on research. If you don’t want to do that, then accept that this degree is NOT equivalent to a PhD.

    • Debora Weber-Wulff February 12, 2016 at 2:49 pm

      Hi Dean,
      unfortunately, many universities are embarrassed and indeed turning a blind eye. If you go to and scroll down to “Bezeichnung des verliehenen Grads” you’ll see in yellow those that have been rescinded, in grey those who have not been rescinded, and the ones without color that still haven’t been decided, sometimes even after 3 years.

  • M.V. February 11, 2016 at 11:51 am

    Debora Weber-Wulff is also the author of an excellent recent book on academic plagiarism: False Feathers: A Perspective on Academic Plagiarism (Springer: 2014). (

  • Steven St. John February 11, 2016 at 12:10 pm

    There would seem to be a massive failure on the part of the mentors here, and indeed, entire dissertation committees (assuming the process is analogous to a U.S. academic dissertation, with which I am more familiar). The students are busy, the advisor’s busy, the committee members are busy… and the world turns. Don’t do a dissertation, and don’t mentor a dissertation, if you are too busy to do it right.

  • PJTV February 12, 2016 at 10:13 am

    A question for Debora: so far it is all about copying text. However, there is also copying (better stealing) ideas. Are there any thoughts about detecting those?

    • Debora Weber-Wulff February 12, 2016 at 2:25 pm

      Hi PJTV,
      copying text is hard enough to detect, as small changes in word order, substitution of synonyms, insertion of words, etc, will confuse a program, although a human reader can see the similarity. Similar ideas cannot be detected, merely the expression of the ideas.

      • PJTV February 13, 2016 at 9:17 am

        Thanks for the reply. I had hoped for more ;). However, I also wanted to make the point that we should not ruminate on copy-paste work and then forget the worse crime of presenting ideas without due reference and credit. And yes, that is much more difficult to detect, maybe impossible without human readers.

  • Mike December 28, 2016 at 9:10 am

    Great research on such an important topic.
    I feel like many students have a poor understanding of plagiarism due to the lack of information about it. They underestimate the meaning and consequences of copying or plagiarizing someone else’s research. Moreover, like you mention, they may interpret the advice to look on other’s work wrongly and understand it like permission to use the data. There should be far more many explanations and guidelines on plagiarism issue and also some plagiarism detection tools provided, like plagiarism checkers (for example, or Turnitin you mentioned). They are not perfect, but at least they can found matches with other publication and calculate similarity index for the papers and make student pay attention to the problem, or prevent even unintentional plagiarism.
    Thank you for the interesting article.

  • Post a comment

    Threaded commenting powered by interconnect/it code.