Scientific Retractions on the Rise

Friday, June 08, 2012

Transcript

When a paper released by a scientific journal turns out to be wrong, either due to human error or intentional fraud , the journal’s editors often will issue a retraction advising scientists to disregard the research. A Wall Street Journal study has found the number of such retractions to be soaring. New Yorker science writer Jonah Lehrer tells Brooke what he thinks is going on.

Guests:

Jonah Lehrer

Hosted by:

Brooke Gladstone

Comments [8]

Canary from New York

Oh, the irony.

Sep. 06 2012 06:40 PM
K. Wilhelm from NJ

In the law, it would be useless to provide court decisions without indicating if the decisions have been overturned, modified, reaffirmed, cited by other courts, etc. Lawyers would not subscribe to major legal databases like Westlaw or LexisNexis if they did not mark decisions with green, yellow and red flags to show that the holdings of the decisions are good law, bad law or something in between. Before the advent of electronic databases, providing such updates to the legal profession was an entire business plan for a company called Shepard's Citation Service; law students learned how to use Shepard's printed volumes and regular paper updates to confirm that a case they wanted to cite in a legal document was still good law. Science should take a leaf from Shepard's book.

Jul. 05 2012 01:06 AM
William Park from Silicon Valley

I would like to respond to Lila Ramaiah's comments. The scientific field is undergoing a slow and painful transition to the web. My company, DeepDyve (www.deepdyve.com), has partnered with many of the leading academic publishers to make their journal articles available for "rent" providing savings up to 90% compared to the cost of purchasing the articles. My apologies for the commercial interjection, but we believe that scientific publishing must continue to innovate to not only make research more accessible, but also to ensure, as Lila points out, that the research is accurate.

Jun. 19 2012 07:53 PM
Jen

Speaking of citation... I came to this page looking for links to the articles discussed in the piece (specifically the fisheries science study from the UW) but don't see any links or works cited list. For those of us wanting to know more, or to read the original source material and do our own analysis, this information would make it easier to avoid the "laziness" of failing to do our own due diligence.

Jun. 14 2012 12:11 PM
Lila Ramaiah from Jersey City, NJ

I would like to respond to a comment made during the show that very poorly and wrongly characterized scientists. Your guest stated that one reason for mistakes made in citations was "sheer laziness...lots of people use citations without fully reading the papers. They look at the title and maybe the abstract..."

I believe that the true answer to tho question is that we can only read the title and abstract because the full text version of an article simply isn't available. I truly wish I could confirm the findings stated in abstracts by verifying the full text. I am a scientist tasked with reviewing many submissions to journals. I have also written many texts in my field of expertise. I pay for yearly subscriptions to approximately 10 journals that are specific to my field. When a specific article is of particular importance, I do dish out the ~$30 cost to get the one article. However, there is no person, or institution for that matter, that can access full text versions of all articles published.

I took offense to your guest's comment. I believe that most scientists, including myself, have a very ethical bar and are my no means lazy.

Jun. 13 2012 01:11 PM
chris bender

I want to add to Matthew Leone's comments, which are on the mark, by mentioning some aspects of scientific publishing that might help explain some of the statistics discussed in the report and which seemed baffling to the reporters:
1. The role of the abstract: In my scientific training, one mentor emphasized the importance of crafting the paper's abstract. He explained that this is necessary because these are published in their entirety by the various databases, and few scientists read anything other than the abstract because of the sheer enormity of the science literature. As strange as this practice may sound, oftentimes the details of the study are not necessary to spur on ideas; the full papers are typically used as reference works for experimental protocols and/or experimental data when they are required (hence the importance of a research library).
2. Review articles: Often these abstracts are cited in 'Annual Reviews" or compilations that in some cases are revised annually rather than re-written; these annual reviews and compilations are also often used as the "go to" source for writing more papers, and this is one example of how one winds up with citations that never go away. But the purpose of the review article is simply to provide the reader with snapshots of science that MAY be something interesting and significant (review article authors rarely provide commentary), and then provide the requistite informaton for the reader to get the original paper.
3. Lastly, peer review does not mean that the experiments described in the paper and the resultant data have been replicated by the reviewers. It simply means that the paper was read, critiqued according to the state of the art in that field, and returned to the journal's editor with a recommendation (usually, publish as is, publish with - sometimes specified - revisions, or reject). Rejection often has nothing to do with the quality of the work, but rather whether the reviewer deems the results and conclusions to be 'significant'. It is possible and likely, as Dr Leone remarks above, that subsequent study will refute a published conclusion. In fact, the best indicator of a published work's significance is how often it repeated, refined, and refuted with the goal of getting that particular fact 'right'.
4. In short, scientists are just like everyone else, and one finds the full gamut of honesty, laziness, etc.

Jun. 11 2012 12:12 PM
Matthew Leone, MD from Yonkers, NY

I want to comment on your pieces “Scientific Retractions on the Rise” and “Retraction Watch”. I think the first piece lacked enough distinction between retractions and refutations. You do not present them as the same thing, but even to the careful listener, they would be easy to conflate. Retractions are due to fraud, failure to get IRB approval, etc.
The other class of error – due to scientific conclusions being proven wrong over time (refuted) – goes by another name: science. This is exactly how science is supposed to work when it works correctly and well. It bears mention that refuted papers are an important part of science and it would bother me if your listeners lumped them together with fraud. Fraud has no place in science but refuting each other is essential. One corrupts the process that is science and the other tries to move it forward.
The pieces are great and much in them was completely new to me.

Jun. 11 2012 12:49 AM
Kenneth Zahl MD from Morristown NJ

Brooke Jonah Lehrer could not be more on point Steve Schafer MD is a hero in terms of retractions of fraudulent and faulty research. Retractions reset the fast thinking jump to wrong conclusion junk science that is all too pervasive. Excellent coverage

Jun. 09 2012 07:40 AM

Leave a Comment

Register for your own account so you can vote on comments, save your favorites, and more. Learn more.
Please stay on topic, be civil, and be brief.
Email addresses are never displayed, but they are required to confirm your comments. Names are displayed with all comments. We reserve the right to edit any comments posted on this site. Please read the Comment Guidelines before posting. By leaving a comment, you agree to New York Public Radio's Privacy Policy and Terms Of Use.