Friday, December 03, 2010

Another fine mess Stanley

UPDATE: Wiley Coyote requests a hint on the Stanley matter:



Medical Writing Editing and Grantsmanship brings word to Eli of another fine mess, one that makes the fluffy kerfuffle (Eli DOES hate that word) about the Wegman Report, various theses, and the response of George Mason University, look an introductory course. In this case,
Based on the findings of an investigation by Columbia University (CU) and additional analysis conducted by the Office of Research Integrity (ORI) during its oversight review, ORI found that Bengu Sezen, former graduate student, Department of Chemistry, CU, engaged in misconduct in science in research funded by National Institute of General Medical Sciences (NIGMS), National Institutes of Health (NIH), grant R01 GM60326. Specifically, ORI made twenty-one (21) findings of scientific misconduct against Dr. Sezen based on evidence that she knowingly and intentionally falsified and fabricated, and in one instance plagiarized, data reported in three (3) papers* and her doctoral thesis.
Seven papers were earlier retracted unilaterally by the adviser, Dalibor Sames and the claims were big news at the time. A good summary can be found in an presentation for an ethics class by Julia Wang. One of the dirty littles in this pile of laundry is that Sames ran through five other graduate students who could not reproduce Sezen's results (as in fired) and that Sezen has continually defended her results.

UPDATE: M. points to Sezen's defense which is JSE class. Ed Wegman will NEVER top this
Sezen now also claims that Sames did not use the proper catalysts when trying to reproduce her work. In an e-mail to C&EN, she writes: "It is as simple as this: You can not make espresso without coffee beans. Prof. Sames and coworkers claimed in their retractions that they could not reproduce my recipe for espresso. And later (when I asked which brand of coffee beans they used), they stated that they did not have (and never had) coffee beans. Without having coffee beans, how can one try to reproduce the recipe?"
More to the point, it turns out that after leaving Columbia when this first broke (in 2006, it looks like John Mashey is going to have to learn patience) Sezen went to Germany, got her PhD and now holds a group leader's research position in Turkey.

There is lots of comment on this and will be more

Janet Stemwedel had opined on this mess in early days with the question of whom should you trust and when

5. So can you ever trust your collaborator, or should scientists all author their papers alone? Can advisors ever trust their graduate students, or graduate students their advisors?

The alternative to trusting other scientists, whether by way of collaborations or by consulting the scientific literature, is doing all the science you care about all by yourself. And if that's your plan, you really have too much to do to be reading blogs. Scoot!

On the other hand, since scientists have some acquaintance with the idea of backing their beliefs with facts, it's good to base your trust on facts, too. This would be easier if collaborators took the time to get to know something about the pieces of the project contributed by their collaborators, if PIs still got involved in conducting experiments themselves, if grad students worked up their data with their advisors at least some of the time rather than only delivering the finished product with a pretty bow on top.

25 comments:

Sou said...

This is not just an academic issue, it is a risk in any workplace.

Successful cheats are good at bluffing. Some people are more easily bluffed than others, sometimes because they are lazy or sloppy managers, sometimes because they don't have a good 'nose' for sussing out con artists. Work-based politics plays a part in this as well. Cheats are often good at sucking up to those in positions of influence.

John Mashey said...

Yes, patience ... but I will admit this is a serious entry in the academic misconduct sweepstakes.

but I'm not sure there is a linear metric for comparing two cases, and if you are trying to stir me to harder work, I actually think we're ahead, but it could be arguable. However:

For example, Sezen one has 21 cases and 1 person involved, arguably 2.

1) I think DC and I can beat the 21 cases, if you count each source plagiarized separately. Counting only textbooks (with substantial amounts) we have 3, then there are the 17 Summarized papers, and then surely Wikipedia counts, and arguably the 2 PCA texts.
But that's just for the WR, and I think there are some fabrication/falsification possibilities as well.

Then we have at least one article and 2 PhDs that re-used the social networks stuff from the WR that was already plagiarized.

Then we have Said's PhD that used the ethanol web page.

SSWR had at least Wegman plus 4 students (Said, Sharabati, Rigsby, Hezazad) involved to some degree or other.
===
Two days ago, plagiarism by Wegman(or coauthors) of an earlier GMU PhD has popped up, re-used multiple times in articles, a 4th dissertation, and a patent. This just appeared at Deep Climate, H/T Andrewt.

Then there is at least one more article DC has found, not yet written up, I think. I just stopped looking a long time ago, but this stuff just keeps coming.

Oh, and then we have the unconnected latecomers, Rapp and McShane&Wyner.

I think, when the dust finally settles, this will turn out to be bigger :-), will probably ensnare students who were badly supervised (like, what was postdoc Said doing supervising Sharabati?) But then, there is a full professor involved, which is pretty rare.

But I will admit, Sezen is fine mess.

Jim Bouldin said...

"The alternative to trusting other scientists, whether by way of collaborations or by consulting the scientific literature, is doing all the science you care about all by yourself. And if that's your plan, you really have too much to do to be reading blogs. Scoot!"
____________________

Correct.
And it ranges from extremely difficult to impossible as well, especially if you want to do something relevant.

a_ray_in_dilbert_space said...

While it is always disturbing to be confronted with dishonesty among scientists, I would point out here that the system worked: the cheat was caught. One of the things about science is that nature keeps giving consistent results when we ask the same questions, so it is to our advantage to make sure our responses concur with those of nature. Cheats are rare in science precisely because most scientists of above room temperature IQ realize they won't get away with it.

And wrt collaboration, I would contend that this is more likely to enforce honesty in papers than the reverse. Having said this, one should trust one's collaborators...but verify what they tell you. Scientists are human. Even the honest ones make mistakes. Our colleagues are our first line of defense against such errors.

Anonymous said...

If you do it all yourself it is far to easy to miss the blindingly obvious (coworkers howling with laughter at your paper is not a good sign). Besides having some well accepted names as co authors must increase the chances of publication, particularly if your results are unexpected or counter intuitive.

ittle Mouse
who did some stuff on Public Administration

John Mashey said...

Well, OK, Eli, thus is ahead on papers withdrawn, so far. :-)
But mire seriously, I repeat a comment from RC a few weeks ago:

"Folks expect modern computer hardware to work reliably. It does, despite the fact that individual components are not perfectly reliable. Some systems (like telephone switches) have long had elaborate hierarchies of fault-recovery software. Memories, disk and communication channels use Error-Correcting Codes … or else none of this would work for more than few seconds.

Science is the ECC of human discourse, it tends to detect errors quickly, fix some and keep them from propagating too far. However, it runs slightly slower as a result.

Sadly, some other areas of human discourse seem to lack even simple parity bits.
They can be cheaper and faster, but then they crash."

Now, in this case, it took 4 years for the Correction, but at least it worked.
I'm also pleased to see that ORI has a long attention span...

Jim Bouldin said...

Good points Little Mouse; I am in fact in that latter situation right now, having just about finished the development of a new paleoclimate (tree ring) analysis method that may have high significance. Paleoclimate's not my specialty (though I do know about tree growth), the algorithm is very complex, I'm a programming moron who had to essentially learn R from scratch to do the work, and I did the entire thing alone. So, it HAS to be 100% airtight or it will get shot down, and given the overall situation (including the nature of the public discussion of climate change), whoever reviews it SHOULD/BETTER do so with a definite critical eye. I expect absolutely nothing less, which is one reason I've sent a preliminary draft to two highly respected experts in the field.

When you work alone, you do the work itself alone, but you have to make absolutely certain that you do not ignore the literature and the informal thoughts of others, and that you do in fact know the topic inside and out. Nevertheless, you can still spend a lot of time reinventing wheels, getting temporarily mired in swamps, etc., depending on exactly what you're working on, your background etc.

The best solution overall, in terms of efficiency and accuracy, is to find people that you solidly trust to be on their game, and then work with those people. Makes a number of things a helluva lot easier too.

Anonymous said...

Has McIntyre jumped the shark with his "Chladni Patterns"? Did the reviewers do him a favor by making him remove references to them in his paper?

Anon(1)

Horatio Algeranon said...

It's like everything else in life.

If you choose your collaborators carefully, chances are you won't have problems with this stuff (plagiarism, cheating, fabrication, lying etc).

These things usually don't just appear "out of the blue". There are recognizable patterns associated with this kind of "questionable" behavior -- signs that things are not right.

And it is often "repeat behavior". People who do this stuff usually have been doing it for some time and therefore have a history (sometimes even a reputation). (Anyone who now would work for or with Sezen deserves what they get, in Horatio's humble opinion.)

That's why it also should really come as no surprise that "this stuff just keeps coming."

In fact, usually, such behavior will not cease until there is a negative incentive to do so (ie, until the person engaging in it gets caught and is forced to "face the music")

Most of us have encountered such behavior in folks that we work with at one time or another and gone out of our way to avoid them like the plague(That's not always possible, but usually it is)


And incidentally, the USPTO might have some interest in the patent issue, partucularly if any of the uncited material has a direct bearing on the "novelty" (and hence patentability) of the invention.

Anonymous said...

Tip & Ring to you too, Jim Bouldin.)

Anonymous said...

A character called Jan Hendrik Sch\"on managed a Science and also Nature - careful how you say it - faked-up paper before he got sprung badly. Suspicion existed but the guy knew how to talk his way out of a jam. Eventually - take a deep breath - the ol' duplicated graphs in utterly different papers on different topics but with different captions and axis labels and variables but the same noise exactly trick (h/t Maxwell Smart) got him in the end. Or maybe it was the ol' duplicated data for quite different experiments in different papers trick, I forget. Anyhow, look up "Plastic Fantastic" as a booktitle on google and that'll fill you in.

Anonymous said...

In a major oversight, you forgot to mention the "coffee bean" defense: http://pubs.acs.org/cen/news/84/i25/8425papers.html

More seriously, there seems to be some serious questions raised about Prof Sames qualifications as an advisor. The (alleged) fraudulent papers are not sufficient in and of themselves - I did have days as a grad student when I realized that I could just make up data and my advisor would never know (or, more tempting, when I realized I'd done something wrong, I could have left my first results standing rather than admit to my error and correct them), and there have certainly been other cases of fraud where the advisor was clearly not at fault* - however, the (alleged) firing of FIVE students for inability to reproduce these results is really inexcusable - firing even one student without probing WHY they were unable to get stuff to work in lab is bad, but when a second student couldn't do it, that should have raised serious red flags, and WHY THE F*** would any advisor not realize what was up by the time student number 3 couldn't do it??? Also, at least one of your linked websites claims that there are rumors about a possible "romantic entanglement" between advisor + student... but perhaps it is best not to repeat such gossip without evidence.

*I refer here to the 'Enantioselective Reactions in a Static Magnetic Field' paper (http://onlinelibrary.wiley.com/doi/10.1002/anie.199404541/abstract) - when I was an undergrad, grad students in my lab worked to try to reproduce the results from the paper (because my advisor was rightly dubious that these results were real, and also because the result would have been fairly earth-shattering not only in the enantioselective catalysis field but also in the bio-genesis field... a lesson perhaps that one should not fake results that are so big that other people will pay attention): not only did Zadel apparently falsify results, but apparently when other people in his lab were working on reproducing them he would go in and spike the samples so that they would also get the same results. Apparently, he has defended himself variously by claiming that he "vaccinated" his samples with a tiny bit of pure enantiomer in order to get more of that enantiomer, and that some people in his lab managed to repeat his results once when he was on vacation. Four years later, unable to get a good job, he sued to get compensated for lost income due to his lost reputation...

http://syntheticenvironment.blogspot.com/2006/12/top-5-scandals-in-medicine-and.html has a list of a top 5 frauds: in addition to the Zadel fraud, it mentions the case of one peer reviewer who rejected a paper, then copied and submitted that paper as their one to another journal... now THAT is chutzpah.

-M

Anonymous said...

While it is always disturbing to be confronted with dishonesty among scientists, I would point out here that the system worked: the cheat was caught.

But you still run loose, free to post anonymously on obscure internet blogs, after advising your boss to run through $10 billion dollars and five years on his fraud advised by a_ray_in_dilbert_space. Why is that?

Anonymous said...

Aw, fudge. I lost a long post. The key points though were:

1) You missed a chance to discuss the "coffee bean defense": http://pubs.acs.org/cen/news/84/i25/8425papers.html

2) Advisors can be scammed by their grad students, and it can be hard to figure it out: eg, the infamous "magnetic enantioselective catalyst" of Zadel, wherein he would (allegedly) slip in and doctor the samples of other people in his lab who were trying to reproduce his results so that they (and only they, in the whole world of people trying to reproduce this experiment, including people in the lab I worked in) would get his results.

2b) Lesson learned: if you really feel the need to fake something, don't fake something that is such big news that everyone wants to duplicate it.

2c) While an advisor can be scammed, they really should figure out that something was wrong long before they run through 5 grad students trying to duplicate it. And you should never fire a grad student without trying to figure out _why_ their experiments aren't working. And you especially shouldn't fire a 2nd grad student (and 3rd, and 4th, and 5th???!) for failing on the same reaction without wondering if maybe it isn't your students but rather the reaction at fault. This suggests that Sames may be... a less than stellar advisor.

3) Other fun frauds, including the guys who peer reviewed a paper, recommended rejection, and then submitted it as their own work to a different journal: http://syntheticenvironment.blogspot.com/2006/12/top-5-scandals-in-medicine-and.html

-M

Wiley Coyote said...

Dear Mr. Dr. Professor Lagomorph, Sir,

Who exactly is Stanley?

Sincerely,
Wiley

Wiley Coyote said...

...and was he the copy-er or the copy-ee?

The group here is concerned about this copying thing. It's things like that that lead to defective rocket skates. We also urge that "copy" be used instead of that big word starting with p.

Sincerely,
Wiley

Wiley Coyote said...

...which leads to stressful things like in my name link, which we are not fond of, being as we are canids accustomed to roam freely...

EliRabett said...

Eli gets the long ones, although they don't appear and he has superbunny powers from the Akame
-------------------------
In a major oversight, you forgot to mention the "coffee bean" defense: http://pubs.acs.org/cen/news/84/i25/8425papers.html

More seriously, there seems to be some serious questions raised about Prof Sames qualifications as an advisor. The (alleged) fraudulent papers are not sufficient in and of themselves - I did have days as a grad student when I realized that I could just make up data and my advisor would never know (or, more tempting, when I realized I'd done something wrong, I could have left my first results standing rather than admit to my error and correct them), and there have certainly been other cases of fraud where the advisor was clearly not at fault* - however, the (alleged) firing of FIVE students for inability to reproduce these results is really inexcusable - firing even one student without probing WHY they were unable to get stuff to work in lab is bad, but when a second student couldn't do it, that should have raised serious red flags, and WHY THE F*** would any advisor not realize what was up by the time student number 3 couldn't do it??? Also, at least one of your linked websites claims that there are rumors about a possible "romantic entanglement" between advisor + student... but perhaps it is best not to repeat such gossip without evidence.

*I refer here to the 'Enantioselective Reactions in a Static Magnetic Field' paper (http://onlinelibrary.wiley.com/doi/10.1002/anie.199404541/abstract) - when I was an undergrad, grad students in my lab worked to try to reproduce the results from the paper (because my advisor was rightly dubious that these results were real, and also because the result would have been fairly earth-shattering not only in the enantioselective catalysis field but also in the bio-genesis field... a lesson perhaps that one should not fake results that are so big that other people will pay attention): not only did Zadel apparently falsify results, but apparently when other people in his lab were working on reproducing them he would go in and spike the samples so that they would also get the same results. Apparently, he has defended himself variously by claiming that he "vaccinated" his samples with a tiny bit of pure enantiomer in order to get more of that enantiomer, and that some people in his lab managed to repeat his results once when he was on vacation. Four years later, unable to get a good job, he sued to get compensated for lost income due to his lost reputation...

http://syntheticenvironment.blogspot.com/2006/12/top-5-scandals-in-medicine-and.html has a list of a top 5 frauds: in addition to the Zadel fraud, it mentions the case of one peer reviewer who rejected a paper, then copied and submitted that paper as their one to another journal... now THAT is chutzpah.

-M

Wiley Coyote said...

Dear Mr. Dr. Professor Rabbit,

I confess I am more confused than ever. I am still unsure as to who Stanley is, where the guys at the top came from, why they are singing and crying at the same time, why the mess is fine rather than coarse, why the others are apparently now and forever out of coffee beans for their espresso (whatever that is), why everybody got fired for it, and just exactly who copied who.

This is indeed a mess. I am sorry but I am going back to reading the ACME manual, which though not a perfect description of engineering realities, does inspire the imagination without undue confusion.

Sincerely,
Wiley

BillD said...

One key point about scientific scandals is that they usually involve only one person and extremely rarely involve two or three people. The fact that scientists are very dedicated to the truth and not at all shy about criticizing other scientists means that conspiracies involving groups of scientists are just about impossible.

Horatio Algeranon said...

You can not make espresso without coffee beans

...and you cannot make Congresso without copy means.

cynthia said...

Jennifer Washburn, investigative journalist and author of the report "Big Oil Goes to College," has reaffirmed my suspicions that private corporations are hooking up with publicly-funded scientists in order to have scientific research skewed in their favor as well as to get "a research lab on the cheap"...

http://www.democracynow.org/2010/10/18/big_oil_goes_to_college_bp

This adds to a growing body of evidence that Big Business has become so intertwined with Big Government that corporate scientists and government scientists have become mirror images of each other. So it's rather pointless for anyone to be arguing that government scientists are part of the fat and bloated world of bureaucratic socialism while corporate scientists are part of the lean and mean world of free-market capitalism. These two groups of scientists now coexist together in the corrupt and fraudulent world of crony capitalism.

And about the only way to crack down on this type of corporate crime is to shut down K Street. But don't expect this to happen anytime soon. And anyone who thinks otherwise has got their head buried in the sand!

a_ray_in_dilbert_space said...

Cynthia,
How many gummint scientists do you actually know? For that matter, how many corporate scientists do you know? Do you know that Exx-Mob's scientists were telling the behemoth to quit opposing the science of climate change?

The thing about science is that nature keeps giving you the same answers when you ask it questions. Nature doesn't change its story. And us scientists, we're really just stenographers who try to take down what science tells us and make sense out of it.

There is no "corporate science." There is no "gummint science." And while we're at it, there's no "liberal science" or "conservative science". There's only science, and if you depart from the truth or spin your results, you ain't doing it. Where the K-Street criminals do have influence is in what does and does not get funded.

cynthia said...

I gather that you are a pure and uncorrupted government scientist, who hasn't sold out to the corporate kleptocrats on Wall Street, making you a rare breed among your peers. If what I'm saying is true, more power to you, Dilbert.

a_ray_in_dilbert_space said...

Cynthia,
I know of very few scientists getting rich. Most are in it purely because we want to understand our subject matter--no more than that. Yes, it's a rat race, and competing for grants is a pain and gets in the way of doing science, but nearly every scientist I know is happiest in the lab. Now, granted, I work in the physical sciences, not pharmaceuticals, but I still say if you stray one iota from the truth, you ain't doin' science.