Weekend reads: MIT professor accused of fraud, biologist who retracted paper suspended, and more

booksAnother busy week at Retraction Watch, featuring lots of snow at HQ and a trip to take part in a conference in Davis, California. Here’s what was happening elsewhere on the web:

11 thoughts on “Weekend reads: MIT professor accused of fraud, biologist who retracted paper suspended, and more”

  1. I like Gary Taubes, he does seem to be fighting a lonely battle against the ‘science’ of nutrition. I wonder how many NIH grants are wasted on studies that only test for correlation, or do not even have a testable hypothesis. My R01 is replete with ‘to test this hypothesis’. I would love to check a R01 from a nutrition lab.

    1. Perhaps Helene Z. Hill needs to change his/her/its strategy. We summarize our experience. Anonymous whistleblowers are being heard by Elsevier management, who take our reports seriously and follow through with the cases reported. We have a much worse track record with Springer management, and this may be related to the individuals who are in charge of investigating the cases. We have recently challenged the qualifications of those leading such cases and will publically report on our findings here to keep up the pressure and to maximize transparency. The greatest hurdle we have faced is with the editorial firewall. Our experience indicates that the pride of editors-in-chief and the lack of experience and depth of understanding for the need to correct the literature by editorial teams is weakening actual case reporting and a fair and correct response. We believe that publishers and editors are as accountable for the literature as authors. And unless actual misconduct on the part of authors can be proved (and proof must be publically provided), all three parties must be held equally accountable for correcting errors in the literature. Another great hurdle we face is the lack of understanding by editors that work published 5 or 10 years ago should be corrected. They believe that only recent work, or work under their watch is allowed to be scrutinized. Many believe that small errors to incorrect scientific claims should remain as such and do not merit errata. In essence, we are not only dealing with a psychological firewall by proud editors against valid whistle-blowers, we are dealing with gross negligence and lack of professionalism. So, how does one go about achieving success? We offer our advice to those seeking a way forward to academic justice and a firmer correction of the record.
      1) Make initial claims anonymously. Always support your claims with the opinions of at least three individuals, all of whom must be experts. Be confident that your claims are factually correct and make sure that only factual claims are made in the report.
      2) Never reveal the identity and never bow to pressure. Be insistent, not aggressive, and never illegal. When challenged, or ignored, widen the case and the complaint. This can be achieved by: a) mailing all editors; b) mailing a dozen or more professionals and peers; c) posting a report publically with an anonymous name.
      3) Tone. People are easily offended, especially with the words “misconduct”, “fraud”, or similar terms. To avoid offense and libel, stick strictly to the facts, but do not be afraid to label the class of misconduct, for example, figure duplication, plagiarism, self-plagiarism, in each and every case quantifying the problem as best as possible and indicating the problem accurately, and precisely.
      4) Form an anonymous group and seek additional support, with the guarantee of privacy. One will find more support from peers when they learn that their identities will not be revealed. This can make the case more difficult, but it lends you the power of truth based on fact, i.e., evidence-based claims. Not even the best lawyers in any publisher will be able to escape facts.
      5) Despite 4, seek to decentralize. Centralization is initially important to gather moss. Decentralization is important to get snowballs forming on different planes. For this, you need to identify motivated leadership with the same mentality.
      6) Respond always to queries that are essential for moving the case forward and that bring clarity. Make sure that the reason is always only academic. This is because in many cases, those who you are trying to report cases to will try to vilify you.
      7) If you are by chance identified, fear not. You have only facts to support you.

    2. I’ve seen an article from the BMJ Christmas satire edition edition actually tested in a US medical school’s nutrition class…

      Seperately, there are randomized diet/nutrition studies, but way too often the p value is worshiped without much concern given to the negligible effect size. While there is much, much, much need for improvement in this field, they may be fighting the most uphill battle possible of any research field with A) perhaps the most difficult organism… one that you just can’t force to follow a diet or blind to treatment B) an industry dedicated to making their organisms break trial protocol and C) the need for longitudinal studies that last many decades to really understand outcomes.

      I don’t know what you study, but if you’re like me and most folks I assume you’re referring to, you test hypothesis either in rodents, flies, worms, cell lines, bacteria, in silico or in some other organism amenable to relatively quick, controlled studies. The nutritionists might ask: how many cancers/psychiatric/obesity diseases can we cure in mice, the knowledge of which has provided no benefit to human health yet? Further, I think we might all agree that 1) cigarettes are bad for people 2) smoking rates have declined since this knowledge has become public 3) smoking related diseases have been declining with rates of smoking. How was this discovered? The same methodology that most nutrition researchers use: association.

      Conclusion: Should they do better? Yes. Can we do a randomized trial forcing half of the people to only eat big macs? No. Can we glean important information that benefits public health from association studies if effect sizes are consistent and big? Yes.

  2. The Trivers situation looks like a prickly old prof whose been calling “fraud” against a fellow professor, so now the administration is trying to force *him* out instead. Would like to know more.

  3. The article by Taubes scratches the surface. We need more articles like this.

    An open-access article by Lesser et al examine the connection between industry-funded nutrition research and findings. (Relationship between Funding Source and Conclusion among Nutrition-Related Scientific Articles (http://www.plosmedicine.org/article/info%3Adoi%2F10.1371%2Fjournal.pmed.0040005)) explains the (mixed) association between the consumption of specific foods and the prevalence of obesity. Having spent a brief period working at a nutrition research center, I was floored at the presence of the food industry in every aspect of research. Their presence is greater when NIH and other government sources for funding face budget cuts.

  4. The Trivers article is behind a paywall. Can anyone supply more information? He seemed like the good guy in that retraction story.

    1. Yeah, there is an older story on Chronicle of Higher Education that’s not locked. Just put Trivers name in the search box. And the free one is actually about the fraud. There are news stories on the Web about the current flap too.

      The paper with the fraud in sounds pretty crappy, it has to be said. Not to excuse fraud. And Trivers seems to both think he’s some kind of intellectual god and apparently goes around saying so. He pitches yelling fits at people and gets drunk in public. So whatever disapprobation Nature deserves (mainly for other stuff, like glitzhoundary) it sounds like Trivers is the author of his own misfortunes at Rutgers. Plus if he’s an anthropology prof (he keeps changing disciplines) then shouldn’t he know enough to teach Human Aggression? Is that not one of those standard undergrad anthro courses?

  5. Time for Retraction Watch to make a retraction! You state as part of a busy week:

    An article on minimum alcohol pricing in the BMJ “should never have been published in a respected academic journal, even as a feature piece, as it did not come close to the standards of scholarship that should be expected of such a journal,” says the Institute of Economic Affairs.

    The Institute of Economic Affairs is a political lobby group that promotes free market ideas not an academic or scholasitic organisation. Their views on minimum alcohol pricing reflect this. Would you publish the views of the Heartland Institute vis-a-vis PNAS in Retraction Watch? Maybe not.

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.