Slashdot Mirror
MIT Professor Fired over Fabricated Data
karvind writes "CNN is running a story where MIT has fired an associate professor of biology for fabricating data in a published scientific paper, in unpublished manuscripts, and in grant applications. Luk Van Parijs, 35, who was considered a rising star in the field of immunology research, admitted to the wrongdoing. The revelations are a serious blow to MIT, which prides itself on its reputation as a scientific powerhouse. The announcement also serves to answer the rumors that have been swirling on the campus since Van Parijs vanished from the campus more than a year ago and had his lab disbanded without any comment from the university. Readers may remember the infamous Jan Hendrik Schön from Bell labs."
They fabricate data all the time. We should fire them. :)
"Imagination is more important than knowledge."
The revelations are a serious blow to MIT, which prides itself on its reputation as a scientific powerhouse.
Revealing a case of fraud strengthens their reputation. If they had let the case die in the darkness after dismissing him--that would lessen their reputation. But admitting that fraud has happened and that the school will not stand for it--that can only gain respect.
what's gonna happen to them? i'd imagine that's something you wouldn't put on a resume
"And I wouldn't have gotten away with it too, if it weren't for those meddling kids!" Scoooby-Dooby-Doooo!
84.2% of all college level professors fabricate data. I have a source for this -- I just can't find it right now.
See my journal for slashdot ID's by year. Mine created in 2005. http://slashdot.org/journal/289875/slashdot-ids-by-year
I'm sure they will try and twist it that way, but it's worth keeping in mind that we know about this precisely because scientists take things like making up data seriously, and try very hard to uncover those who would do it. The Intelligent Design side isn't quite so gung-ho in terms of caring about falsifying data (heck, or even providing data).
Huh? It is a "blow" to their reputation iff they knew about the misconduct and did nothing about it. In this case it is clear that they took swift action. I would give kudos to MIT for reacting swiftly. Recall the conduct of other organizations like NYT in such instances.
I can name one that I was in where I saw a ton of data being fabricated and work stolen - the department of Computer and Information Science at the University of Pennsylvania, and in general the school of Engineering, where I got my Ph.D. I have had my work ripped off by Professors, and seen the same people even do this with Undergraduate summer projects. The area that I found most of this sort of stuff, including fabricating data, is in computer vision and robotics. Everyone is so busy publishing, no one ever checks the papers. The truth is most robotics stuff you hear in the news is some nonsense cooked up for media attention. Then the actual papers never carry out the stuff that the researchers claimed in the news. You know stuff like "MIT professor makes robot with emotions". What a lot of crap.
Welcome to Science!
For going to the trouble of turning in the fraudulent research the tech had their phone tapped (which the lab later denied), was transferred out to a dingy little building in the middle of the desert to do menial tasks and just generally harassed until they eventually got another job.
There's so much pressure for getting grant money that producing the results that will get more grant money is pretty much the norm, espeically in contract research. Everyone likes to think science is pure, but you're deluded if you think that. It's all about making sure you've got enough charge codes to bill your time and supporting that 200% overhead rate.
That's our life, the big wheel of shit. - The Fat Man, Blue Tango Salvage
You mean like this? To sum up the case quickly, this is a tool for the automatic creation of fake but real-looking "science papers" (ironically enough, developed at the MIT), and one such paper ("A Methodology for the Typical Unification of Access Points and Redundancy") was submitted to the 2005 World Multi-Conference on Systemics, Cybernetics and Informatics, and actually accepted. At that point, of course, the authors of the tool wrote about it, the story hit Slashdot, and the organisers of the conference were quick to retract their acceptance...
Still, I think it goes to show that if someone is actively trying to dishonest in the scientific community, it's not hard to get past the safeguards. Fabricating data is something that is (I guess) comparatively hard to detect, compared to an entire document that was written without any human intervention and thus shouldn't really make any kind of sense at all, but even the fake document wasn't detected. It sure makes you wonder how many people fabricating data are actually not caught and instead get away with it.
quidquid latine dictum sit altum videtur.
Science is based on the belief that experimental results can be replicated in repeated experiments. I've always wondered why the global scientific community doesn't do more replication of data as part of peer review. A formal procedure for extracting the experimental specification so the experiment is performed without the prior data included in the knowledge of the experimenter. Then a comparison by another party not performing either experiment, so the data comparison is "clean". That seems a very valuable process, in validating the original, finding differences still covered by the same theory, but available for more precision, as well as training scientists - both new and old. It seem replication for the purpose of disproving flawed theories would be the most important, and most common, scientific activity. Is the relative lack of it due to the "efficiency" of the corporate science business? Shouldn't academics be spending more time replicating?
--
make install -not war
They removed his faculty page from mit.edu but it is still avaiable at archive.org.
Seastead this.
I've always wondered why the global scientific community doesn't do more replication of data as part of peer review.
Just try getting a grant for "Doing exactly what this other guy already did, just to make sure."
Yeah, it actually is important, but try explaining that to the bean counters. The best you can do is propose some sort of "continuation" and include the original experiment as a control group, and hope to verify it that way.
I worked in the Biology department at MIT when this happened. While Van Parijs' lab was under renovations, he took up space on my floor. After that, our department(mini-department? sub-department?, whatever) provided some computing resources for their lab. I was the network/systems guy, so I took care of our machines in their lab.
One day, I noticed that the Windows box in their lab wasn't responding and had been reported as haven been taken by the Cancer Center's sysadmin guy. I talked to a buddy of mine who sits across from me and did lab work for the Van Parijs. He called and asked about the machine. A couple of minutes later, the head of the Cancer Center called him and firmly told him to drop all inquries into said machine. He said it felt like the part of The Matrix where Neo gets the 'How are you going to talk without a mouth, Mr Anderson' line.
That's when the shit hit the fan. I was a weekly regular at the Plough and Stars in Cambridge on Wednesdays, and the Van Parijs members made it out there every other week or so. After six weeks or so, the guys who confronted Luk finally started talking about it.
It was quite the news in the department. I don't know about the rest of MIT, but all of Biology, and the CCR, Whitehead, and surrounding buildings knew about it since day one. It worked out well for the members of the lab. Everybody joined up with a different lab, except for one guy. He pretty much started working for himself. He's doing some post-doc work, and in light of what happened, the department just let him start doing his own thing until he finishes up.
What I remember about Luk Van Parijs(other than that he had the most gorgeous Russian administrative assistant. I could write for hours about her. I mean, she was hot and she said things like 'I think my phone just did a core dump' Hi Masha!) was that he was pretty much a jerk. Not that remarkable being that for MIT professors this is the rule and not the exception, but a jerk nonetheless.
Anyways, everybody thinks the New Scientist article was pretty scathing.
Do you see the sig? Do you have it in your sights? Why yes, Miss Moneypenny...
http://www.newscientist.com/article.ns?id=dn8230
Here is how they noticed a pattern:
Michael Borowitz, at the Johns Hopkins University School of Medicine in Baltimore, Maryland, says: "The shapes of the major clusters are often similar but in any system there is noise, and those noisy dots are in the same place too. That's hard to explain by biology. It is very difficult for me to believe that these were independent experiments." Borowitz is an expert in interpreting flow cytometry graphs, which he regularly uses to identity abnormal populations of cells in the blood and bone marrow of leukaemia patients.
Three other experts contacted, including Paul Robinson, a professor of immunopharmacology and biomedical engineering and Director of the Flow Cytometry Labs at Purdue University in West Lafayette, say that the graphs appear concerningly alike.
Except for ending slavery, the Nazis, communism, & securing American independence, war has never solved anything.
Noonian....Didn't he fabricate Data?
hahahahahhahha
Obscure Sci-Fi reference
People who lose sight of that, and who make stuff up to submit, are not only disrespecting their peers, they are stealing time and effort from them. For example, I lost about six months of my life because a senior colleague falsified data that I needed in graduate school. We were in the business of flying a rocket payload to look at the Sun in extreme ultraviolet light. We calibrated the photographic film at a synchrotron facility at Stanford. Our senior colleague (who later went on to become a bigwig at SPIE and in NASA's Astrobiology program) was in charge of developing the film that we exposed, at great effort, to calibrated amounts of ultraviolet light emitted by the synchrotoron. He forgot (or something) to write down which process he used on which piece of film. As a result, a year later when we were analysing our images of the Sun we couldn't make any sense of them. It took a good six months of concentrated effort to eliminate all reasonable hypotheses about what had happened, and to conclude that the film processing notes from that calibration run were simply made up. Once we knew that, we could get reasonable (if not-as-good-as-we-hoped) results from the rocket flight, using earlier calibrations. If my colleague had fessed up immediately we would have lost a few days' work rather than six months.
In the short term, the scientific refereeing process keeps out many honest mistakes or omissions, but anyone determined to deliberately slip fake results into a paper can probably get away with it. In the long term, though, there's no escape: anything made up will either be buried (because it turns out to be uninteresting or because no-one trusts it), or found out (because, if it is interesting, others will try to use or reproduce the result, and will niggle at it until the truth comes out).
Oh come on, you think this guy is the only one who did it? Let me ask you this: you have a hypothesis. You spend a ton of money from your grants and have your grad students spend a lot of their time trying to prove their hypothesis. The data you get is basically useless since it doesn't prove or disprove anything. Do you just say "New research into immunology finds nothing?" Of course not.
We VERY rarely hear of research actually failing, when in fact we should be hearing it ALL THE TIME since taking stabs at new ideas shouldn't be successful all the time. Failure should be a natural part of research, and there really shouldn't be an urge to have to make your research fruitful everytime. Unfortunately, no one would actually do this even if they agreed with the thought - people would only expect other people to follow the rule.
It's not like it matters too much regardless - 90% of research papers are bullshit wrapped in a myriad of technical jargon which makes it seem like they achieved something ridiculously important.
My 2 cents.
Ok, just to be clear here, the World Multi-Conference on Systemics, Cybernetics and Informatics is a well-known waste of academic space. It regularly spams people requesting submissions, has no obvious standards, and will accept pretty much any paper the authors are willing to pay to have published. It has no safeguards. It also has no respect. The MIT tool was developed specifically to prove that the conference in question was a sham.
We here at MIT find it quite humorous when someone suggests that this reflects badly on the Institute, given that the person in question was educated at Harvard and Caltech.
This space intentionally left blank.
What you say is true, however, this isn't really the greatest timing for a story to break on the fact that scientists sometimes fabricate their data. This provides a rather juicy opportunity for the various anti-science forces out there to point to this and say "See, scientists aren't the pristine investigators of truth that they would like us to believe! This one got caught, but how many others are doing the same thing right now? That's why we need to keep an open mind about {intelligent design, alternative medicine, bigfoot, global warming is a myth, etc.}."
You and I may see this story as evidence of the scientific system working the way it is supposed to. I suspect that the public will see this as evidence that science doesn't have a monopoly on the truth and maybe we ought to give those creationists equal time. Like I said, this isn't the greatest time for this story to break.
GMD
watch this
You completely missed the point of the article you linked to. The "World Multi-Conference on Systemics, Cybernetics and Informatics" is one of those academic "Conferences" that exist solely to make money. These conferences are a well known phenomenon used by organizers to make money on outrageous registration fees and by "attendees" to take vacations at their institute's expense. No reputable academic conference has such a thing as a "non-reviewed" paper. Even a cursory reading of the papers they submitted makes it obvious that it's random junk.
It's true that it's possible to sneak fabricated data past the peer-review process, but I think the damage is self-limiting in a way. If your results are significant, people will be interested in duplicating your results... either as a way of understanding them better or to compare against their own work. If nobody is able to duplicate your results, you are likely to have your fraud caught sooner or later.
If your results are not all that significant, it gets forgotten and nobody builds on your bad work so the scientific process itself isn't subverted although the dishonest researcher may have got an undeserved feather in his cap.
And if you got rid of government funding, you wouldn't have much left (or so the conventional knowlege goes). I'd actually agree that gov. funding should be eliminated, but I don't see how it "distorts the intellectual environment". If you're claiming that "money" in general is corrupting, I don't know what to say. People who reserach for the hell of it do it either way; money convinces the greedy bastards to start contributing. It seems you're more blaming shortsightedness than money itself.
Most great leaps in science have been governments funded/nobility patronaged. Most incremental improvement of exsisting technlogoies come from corprate enviroments. Corprate research goes to where the money soon will be, this leads to near sighted research into only a few fields. We get better cars, but we won't derive new energy sources. This is where academic government funded research is for. They research the things that aren't profitable but are interesting to science. Gov. funding should never be eliminated unless you want science to degenerate into nothing more then incremental improvements of consumer products.
"There are more things in heaven and earth, Horatio, than are dreamt of in your philosophy."
Did your physics prof understand that his 'science' grew out of Philosophic Inquiry?
Sure there are idiots in Philosophy, but it gave birth to all modern sciences.
$30 Off All Plans: Use code TRIPLESAWBUCK
So Biologists get fired for publishing bad data.
How many physicists have knowingly published work that they found themselves to be wrong, but just didn't include or played dumb about the parts with blatant mistakes or systematic problems, or used a derivation that just doesn't work?
I've seen papers from nobel prize winning physicists that upon actually trying to apply the theory one finds that it's totally inconsistent--almost as if the writers never actually bothered to try to check their work, or did and decided to hide the problem rather than throw it away, fix it, or qualify it as problematic.
I understand that there's a lot of pressure to publish, but unfortunately I can't publish my own work because it deconstructs work of more prominent people--it would literally shift the foundations of the theory. The problem is that if I've made a mistake, it's my reputation that's shot.
It can be really heartbreaking, too: I went over my friend's doctoral thesis just before he was to send it for publication, and verifying it through computational proof I found a small, seemingly insignficant problem about 130 pages into 900 pages of work compiled over years that propagated across the entire edifice he'd costructed, rendering the much of the theory implausible--and since it was group theory, difficult to salvage. He hasn't talked to me since (probably because he's busy trying to fix it).
My point is, if you're going to persecute people for publishing bad data, how about publishing people for bad proofs? Sorry about the AC: this is the first time I've ever felt the need to use it, but I've already received enough flak for my criticism of some very brilliant people.
I think including alternative medicine with the other topics you mention is pretty short sighted. To think that we have all the medical answers, and that there aren't other medications or treatments that western medicine might not know about is ignorant. Take for instance pressure points: no western doctor or treatment explains or addresses them.
This is almost precisely the same argument used by the Intelligent Design folks to get creationism taught in schools! I'm flabergasted that this got modded up. Yes, yes, yes -- neither evolution or western medicine has all the answers. Scientists fully admit this. However, the fact that those fields don't have all the answers doesn't mean that we should start relying on creationism or alternative medicine. You need to provide verifable evidence that sticking needles into people can cure ailments and not rely on "well, you guys don't have all the answers" arguments. Alternative medicine most definitely belongs with the other things I listed. Not because it's all crap. Because all of those things openly (almost pridefully) reject the scientific method.
However, I'll tell you right now that there are a great many instances where accupressure/puncture can make huge differences in a number of maladies.
Oh boy, take a guess what my next question is going to be. Can you provide references to multiple peer-reviewed studies verifying your claim? You "can tell me", huh? And who the hell are you? Some guy on slashdot? I'm going to take medical advice from User 549286?
redfieldp, I think you misunderstood my post. Maybe I should have left UFOs out of the list. My point was that alternative medicine is anti-science. Alternative medicine practitioners apply their techniques to the public at large without scientific evidence that these methods work or are even safe. Having scientists publically outed for falsifing data is only going to provide more ammunition to those who claim they deserve equal status and recognition. Alternative medicine is welcome to use the scientific method to verify their claims. Until that time, it belongs squarely in the "anti-science" camp.
GMD
watch this
You're painting with far too wide a brush. Many alternative medicine practitioners and researchers are using the scientific method and expanding our knowledge of medicine. Take a look at the National Center for Complementary and Alternative Medicine - part of the U.S. National Institutes of Health. http://nccam.nih.gov/
I've already taken a look. I recommend you look at this.
GMD
watch this
My philosophy professors said that about philosophy too, and they didn't dismiss physics or mathematics while doing so, because they fully understood how they worked. Philosophy is more about logic than mathematics or physics are. If you don't undertand that, you didn't understood anything about philosophy. In fact, mathematics and physics are a concequence of philosphy, like C and lisp are concequence of computer science, but i won't even try to argue about that because you damn 'scientists' guys are so entranched in your way of thinking that you wouldn't be able to admit it anyway. ...
And that's what's so ironic: you can have a philosophy guy admit anything if you can proove it, even things like "you don't exist". Why ? because if the logic behind the argument is irrefutable, it is by all means true unless you can proove otherwise. But scientists, who believe they have the finest logics that exists, can't be bothered with all this stuff because because you can't measure it, look at it, or quantify it. They're a subset of philosophy which deals whith tangible things, but they forgot it, and believe the science which gave them birth is crap because they don't understand it anymore. Total nonsense
Being a researcher myself, I have seen big lies and little lies. The big ones sometimes become public, like the present case. But I have seen so many little lies and they do have an impact on research, too. For example, people find a peak in a distribution and overestimate the statistical significance of the peak. All the sudden a few counts become a new discovery. Lots of money goes into further investigations until the case is settled. Or take the systematical uncertainty of a measurement. A scientific result is not just a single value. Normally we also have to specify an error of an experimental or theoretical result. Of course everybody likes to do a good job. But to get data published people sometimes attach unreasonable errors to their numbers just to ensure their results get published. Collaborations accept those numbers because nobody has the time to deal with those issues. Everbody has their own little project and it costs a lot of time to proof someone wrong, even if everybody knows the numbers are wrong. Eventually nobody asks to be taken off the author list of a publication. Other researchers then take the published results, fit models, make comparisons and find, of course, that their calculations are dominated by those results with the smallest errors. Again, it takes decades until the falsified results get filtered out. The situation is caused by funding agencies. They want to see positive results, fast. No publication, no money. It is that simple. Just because someone spends the time needed to evaluate carefully all aspects of an experiment or theory does not mean the researcher gets funded. On the contrary! The quick-and-dirty jobs are being awarded with precious resources such as research positions. Jobs are the critical resource here. Funding agencies spend a lot of money into equipment, labs and other infrastructure. Life is good. But there are just not enough positions for researchers available. My older colleagues tell me that things have gone a lot worse over the past few decades. You can also see this when you look at the distribution of the researchers' age. Things are worse in the EU than in the US. The EU is putting a lot of money into projects, but far too little into jobs. The taxi driver with doctor title is no joke. The US has traditionally welcomed foreign researchers, but that has changed some 5 years ago dramatically (long before 9/11) and things get worse for everybody. Science and research is not just end in itself. It is part of our culture and our future. Keep an eye on researchers. Make sure they do their job. But also make sure they can do their job. My research is funded by DOE, thanks.
He can always get a job at the Whitehouse....
- Probably brilliant biologist with
- a PhD from Harvard, who is now
- Discredited among his peers and unemployed, and who
- Apparently doesn't have much of a grasp of ethics.
I can see it now...Those bastards, we'll see who has the last laugh now... Soon my army of Super Mutants will TAKE OVER THE WORLD! Muahahah!
It sounds like the origins of some kind of cheap comic book super vilain. Except, its not really that funny. This guy might be developing biological weapons for terrorists in the near future. If all he cares about is money and isn't too concerned about right and wrong, he's going to go work for somebody who might overlook his past mistakes as long as he can deliver what they want.
I feel really sorry for this guy and I think that a good part of the blame should be passed on to the universities which granted this guy his academic credentials yet failed to beat enough ethics into his head in the process to prevent him from making this tragic career mistake.
Hopefully he can take this in stride, and find some "good guys" who are willing to give him another chance. You can't get that far along without knowing something about biology, and it would be a shame to loose a valuable scientist of that caliber. This is going to be a painful lesson in professional ethics for Dr. Varijs which he will wish he had learned a lot earlier on.
Clickety Click