It can’t happen here. That’s what most scientists will tell you about fraud in science. Science is magically self-correcting, fraudsters are isolated incidents, fraud is something that happens in those other professions. Well, they’re all wrong, as Horace Freeland Judson shows in his new book The Great Betrayal: Fraud in Science. While estimates of fraud — faking evidence, omitting or distorting evidence, and plagiarism — are naturally hard to come by, even very conservative studies place it as high as 10% — a staggering number to those who place their trust in Science.

Judson is at various times a historian, philosopher, sociologist, journalist, and student of science, but he combines all this into a detailed book that combines the best of each field — the journalism is thrilling and readable, the science accurate, the history and social causes analyzed, and so on. Judson seems to know everything about the subject — he’s at every major event, he interviews every major figure at their home, and so on. The result is a through book.

The book traces the cultural context of fraud, analyzes the history of fraud (Mendel, Darwin, Pasteur, Freud — all committed fraud to some extent), gives a very detailed description of many modern cases of fraud (including a whole chapter on the famed “Baltimore affair”), then discusses the problems of peer review and authorship, which most people think prevent fraud, then onto the future of science with open access publications on the Internet, and closing with how institutions can respond to end fraud.

Judson paints a picture of a scientific community that is trapped in its own sense of infallibility. Whistleblowers brings evidence fraud to the university president and he (almost always a he) brushes them off saying “it doesn’t happen here”. And anyway, science is self-correcting. The whistleblower goes public and gets fired — they’re inventing a fuss, tarnishing the name of the university. The government’s Office of Research Integrity investigates and concludes it is fraud but the case is appealed to a board of lawyers who don’t understand the science, are not allowed to look at the scientific evidence, and almost always overturn the case, making specious arguments like “if this data was fraudulent, it wouldn’t look so messy”. Even in the rare case when fraud is generally conceded, all the usual figures trot out the usual “few bad apples” claim — the rest of science is just fine, they say. When Congress dares hold hearings on the matter, the scientists being questioned rile up their colleagues by claiming that government is attacking scientific freedom.

So, in the end, the whistleblower ends up disgraced and unemployed, usually viciously attacked in public. The fraudster might have to go to another university or even retire early if it’s really bad. And the department head who let it happen under him gets no blame and so has no incentive to change things. And so fraud goes on, uninvestigated, unimpeded.

What’s the fraud like? A few examples:

That’s a small sample — the cases go on and on. Kudos to Judson for shedding light on a topic few know even exists.

posted March 14, 2005 11:54 AM (Books) (8 comments) #

Nearby

Stanford: You Really Don’t Have To Read This
What can you say to that?
The Republican Playbook
The Case Against Lawrence Summers
Blogshine Sunday: US Greenlights, Funds Genocide
Fraud in Science
How is Disney like the Soviet Union?
Summer Founders
The Truth About the Drug Companies
The Truth About Maryland

Comments

Seems to me that the detection of all these cases of fraud is a reasonable demonstration that science is self-correcting. The fact that people’s careers aren’t destroyed quite as thoroughly as you seem to hope for is not evidence against this view. When people say science is self-correcting, they mean the body of knowledge is constantly being improved because fraudulent and bad work is being rooted out. They don’t mean that fraudulent scientists are jailed for dishonesty.

posted by Michael at March 14, 2005 06:14 PM #

I don’t care what happens to the people involved, that wasn’t my point. (Personally, I don’t think they should be punished.) No, look at how these things were discovered: the majority were minor lab people who saw something odd and spoke up. Who knows how many things aren’t caught by minor lab people? How many lab people don’t speak up? (Anonymous surveys show that a huge number have seen fraud but haven’t reported it to anyone.) How many did report it but were ignored or threatened? And on and on. These are just the cases that were caught, who knows how many are there under the surface?

Does this level of fraud have a major impact on the general “body of knowledge” of sciences — I’d doubt it. But that doesn’t mean it doesn’t have an impact — the false neuroleptic data was really affecting mentally ill people, and there are many other examples.

posted by Aaron Swartz at March 14, 2005 06:59 PM #

I see two kinds of fraud here. In the first, the data and the results are bogus. In the second, the experimenetal methodology itself is fine (or so I hope), but the results are stolen from the actual researchers and published by someone else who claims all the credit.

posted by bi at March 15, 2005 03:37 AM #

How many of the fraud charges in the book were made in fields like physics or chemistry? It seems to me that most of these charges are related to fraud in biotech-related subjects. With the huge spending increase that has taken place in biotech over the last few decades, it should come as no suprise that pressure on results make some people do less than honourable things.

posted by Toby at March 16, 2005 12:55 AM #

Most of these cases involved researchers simply making up data. That shouldn’t matter. Other labs can follow up on that work and show it is inconsistent. Science is self-correcting in this regard. The NIMH was irresponsible to modify therapeutic practices with only one researcher’s study as a basis.

Furthermore, a lot of those cases involve medical researchers, and medical research, in my opinion, barely qualifies as science. Their sample size is often tiny, but they obscure that fact. I have read medical research papers with four patients that talk about a 75% success rate.

Unethical competition, however, is a serious problem. Everyone has stories of this. Science relies on the open exchange of ideas, and many people abuse that. It generally takes the form of “stealing” preliminary work and ideas from less established researchers. A new professor, with a small lab, has a promising start on a project. A more established professor takes the idea, and with a larger lab and more resources, gets a paper out first. This is a difficult problem to address, as it really comes down to a lack of professional respect, rather than outright fraud.

posted by Justin at March 16, 2005 12:23 PM #

Justin says “other labs can follow up on [fraudulent] work and show it is inconsistent”. But there are no grants given for replication studies so they’re very rarely done and studies that build on the fraudulent study generally assume its true and thus don’t show any inconsistencies.

I don’t think the problem with “unethical competition” is “a lack of professional respect” — the problem is with competition.

posted by Aaron Swartz at March 16, 2005 08:27 PM #

Cool review - thanks for sharing it. One nitpicky question: In the first example of fraud you provide, you state: “William T. Summerlin […] claimed he could transplant corneas, glands, and skin that would normally onto animals — sometimes even across species.”

“Would normally” what? Was it “be rejected”? Just curious. Thanks!

posted by Michael at March 16, 2005 08:46 PM #

An intriguing book; Toby wonders about physics and chemistry. An interesting case was that of Leo Paquette, who reviewed someone’s grant, then published the background material from the grant as his own in a paper in the highly regarded Journal of the American Chemical Society. In the end, Ohio State decided it was plagiarism, the chemistry department there, however, argued that it didn’t really matter (since it wasn’t the actual experiments that were lifted, just someone’s prose and research work). The NSF banned Paquette from getting federal funds for two years, but then funded him with a half-million dollars when the ban was up. So it didn’t kill his career, though I wonder if someone less eminent had done it, would it have? and if those who are more eminent in the field should in fact be held to a higher standard, rather than lower? or at least eat only their own young (aka grad students) and not the young of others!?

posted by Michelle at March 16, 2005 11:18 PM #

Subscribe to comments on this post.

Add Your Comment

If you don't want to post a comment, you can always send me your thoughts by email.


(used only to send you my reply, never published or spammed)

Remember personal info?


Note: I may edit or delete your comment. (More...)

Aaron Swartz (me@aaronsw.com)