We have a few follow-ups from stories we’ve recently covered:
Terry Elton case initially chalked up to “disorganization,” not misconduct
Ohio State University (OSU), which along with the Office of Research Integrity (ORI) recently sanctioned a pharmacy professor for image manipulation, “failed at first to recognize his deception,” according to an investigation by The Columbus Dispatch based on university documents.
The piece, which quotes Ivan, reveals that OSU needed some prompting from the ORI before it concluded that Terry Elton was guilty of misconduct, and not just unintentional errors that he at one point blamed on a research technician who lost her job in October 2011:
After allegations surfaced in 2010, the OSU pharmacy department committee concluded that “irregular” images in journal articles were caused by disorganization, not “intentional malfeasance,” on the part of professor Terry Elton.
That July 2011 report might have ended the OSU probe into Elton’s research.
But a year later, the university found that Elton had intentionally misstated figures in several journal articles and in a grant application to the National Institutes of Health.That was only after the federal Office of Research Integrity urged the university to reconsider the case, using a PowerPoint presentation to highlight a pattern of falsified images in Elton’s publications over the past decade.
The newspaper’s request for public records also uncovered two other misconduct investigations over the past five years.
The story is accompanied by a good roundup of the effects fraud can have on science, also quoting Ivan, along with ORI director David Wright, Ferric Fang, Arturo Casadevall — two co-authors of an October PNAS paper on retractions — and others.
Give both pieces a read. Reporter Ben Sutherly does a nice job of digging up new material that highlights the conflicts of interest inherent in university investigations, and of broadening the story.
Resveratrol researcher loses job for good
We’re several months behind on this, but we thank a commenter for pointing out last week that Dipak Das, the resveratrol researcher whom the University of Connecticut found guilty of more than 100 instances of misconduct, has lost a bid for his job back. UConn told Das he had lost his job on May 18, according to The Hartford Courant, which reported this summer that he was appealing the decision. On August 8, UConn’s Board of Trustees voted unanimously to uphold the termination, as noted in their meeting minutes.
As we’ve reported, Das has retracted 19 papers so far.
Fastest retraction ever? Ivan on Science Friday
Finally, NPR Science Friday host Ira Flatow interviewed Ivan, Scientific American editor-in-chief Mariette DiChristina and MAKE Magazine editor-in-chief Mark Frauenfelder for an hour-long segment last week recapping 2012. The audio is available here.
Don’t miss the opening moments, in which Flatow jokingly set the record for quickest retraction ever.
It appears to me that the ORI may be guilty of a rank injustice towards Professor Elton:
“His research technician, Mickey M. Martin, lost her job in October 2011 after Elton’s grant funding ran out. Elton blamed Martin for irregularities in his lab in a December 2011 letter to a journal in which he had published some research that has since been discredited”
How typical of the federal flatfoot to get the wrong end of the stick entirely. They should be going after the technician.
“His research technician, Mickey M. Martin, lost her job in October 2011 after Elton’s grant funding ran out”
Another instance of jumping to unwarranted conclusions without knowing the facts. When you are a research assistant type on a grant, your job ends when the grant runs out, unless the principal investigator has another grant you could be put on. This is simply because there is no more money to support your role, your job depends on the PI getting funding.
Peter are you an administrator? I remember your saying that you were not a scientist.
Mr Pessoa, perhaps you know only one thing (I already said I recognize one-trick ponies quickly), so I’m sorry to strain your brain.
I know how money flows in academia and I know a lot about how finances work in academia.
Mr Pessoa please don’t mock Mr Bloomberg.
His parents were both scientists who drank heavily and yet his family is also fabulously wealthy who thinks nothing of plonking a couple of million dollars on suitably nerdy looking Ivy League professors
He is a valued addition to our community.
The ORI is not above screwing up. If you would like to know more, I refer you to my website http://www.helenezhill.com.
I had a quick look, but because I know nothing of radiation biology and there was no plain language precis of the core issue that I could find, I afraid I am none the wiser. Aside from the fact you appear to have spend a lot of energy and money for no result.
Unfortunately fraud is an inevitable outcome when very smart and very competitive people fight for limited resources in a system that has at its core an honor code and self-regulation. No other outcome could be expected. Scientists are no more inherently ethical than lawyers, accountants or bankers and self regulation didn’t work out for them either, the only difference is when ethical failures hit these professions the cost is immediately felt and apparent, whereas with science the cost is insidious and not immediately apparent. Getting angry at scientists who cheat is as pointless as getting angry at bankers who pay themselves large bonuses.
Some areas of science we felt was important enough to regulate properly – drug development – and so we have instituted a system where cheating is very difficult and does not get rewarded. Off the top of my head if I was to describe the key elements they would be: blinding, multi-site replication, meticulous documentation and auditing When society is ready we will take the lessons learned there to the rest of science. Until then, it is best to sit back and enjoy the insanity.
Unless I misunderstand, the issue is that readouts from the Coulter Counter were recorded by hand with output recorded in such a way that suggested the data recorded were not recorded randomly.
To use an example, if people use a spreadsheet to record their clock-in/clock-out times, if someone is making things up they are likely to write 9:00 AM a bunch of times, versus the random distribution (eg 8:57, 8:48, 9:05, 9:02).
A better example would be the older Beckman DU spectrophotometers. I used to use ones where we got dot matrix printouts with abs listed out to the fourth decimal (or was it fifth? Been a while). I opted to use the 3.5″ disk module and export data to excel instead.
“[…]In the laboratory, the numbers of cells in the volume appeared on an LED display and were copied from there by hand onto a chart in a notebook. […]. The terminal digits vary from 0 to 9 and, in both Coulter and colony counts, are expected to be random or uniform throughout the 10 possible digits. A test known as the Chi Squared test is used to determine the randomness or uniformity.”
I hope this is sarcastic not a serious comment. The ORI did what it is supposed to do and should always do, making sure your tax money and mine is not spent supporting somebody doing photoshop science.
Fish, why do you assume that it’s the technician who was responsible?
The PI should know exactly what the tech is doing. The tech is the hands of the PI. As such, the PI is absolutely culpable. If the fabrications are all the result of a single tech, the PI is culpable for not better keeping track of the work the tech produces.
To suggest that the buck stops with the tech is amateur…
The tech did not lose her job because of retaliation. The grant was over, ergo it was the end of the contract.
Peter Bloomberg, now you state you know alot of how money flows in academia! May I suggest, sir, you know nothing?
And you then say the tech didn’t lose his/her job because of retaliation. You cannot know that.
Maybe his/er contract was allowed to ‘run out’ because he raised concerns to his/her boss. That would be the PI.
That was the end of him/her.
The rewards go to the boss (PI), the buck stops with the PI.
The PI is responsible, legally and ethically, you may not like it, but thats how academia works, clearly, you have no idea what you are talking about.
Hey stewart, here is the quote again, for your limited attention span, please focus: “Mickey M. Martin, lost her job in October 2011 after Elton’s grant funding ran out.” Nothing else to discuss here.
I’m shocked by the following sentence in The Columbus Dispatch report:
“But Whitacre described Elton’s sanctions as “severe.”“He’s essentially cut off from doing research,” she said”.
Actually, Elton is rewarded with a 3-year sabbatical leave, with $130,146 / year earning. Not so bad!
Please, at least let’s be honest about it!
Peter Bloomberg wrote:
‘Hey stewart, here is the quote again, for your limited attention span, please focus: “Mickey M. Martin, lost her job in October 2011 after Elton’s grant funding ran out.” Nothing else to discuss here.’
The grant may have been allowed to run out if the technician suggested the work that was published is fraudulent.
It has happened before:
http://www.timeshighereducation.co.uk/story.asp?storycode=411229
The doc in question, a world renowned expert in bone analysis, whose contract was allowed to run out, had raised concerns about the validity of her bosses work.
She was right, she got fired, and now….the boss is living it up in first class travel-land!
“It is clear from the PowerPoint that Dr. Elton has a long-standing convention of reusing figures to represent both control and experimental conditions,” wrote John Dahlberg, director of the federal office’s investigative oversight division, in an October 2011 letter.“It would also appear that he has copied, resized/stretched/shrunk, darkened and flipped images (horizontally and vertically) … to conceal similarities.
It looks as though the ORI did exactly the same thing that Paul Brookes was doing on the Science Fraud web site to convince OSU to investigate Elton more fully..
@ PB and Stewart,
Based on my reading of the Dispatch articles and RW post, it is unclear whether Ms Martin’s job loss was retaliation (stewart’s position) or not (PB’s position). In fact, it’s not even clear whether she was the anonymous source of the allegations. All that IS known is that she was the one Elton blamed. So, we can connect whatever dots we want but it would all be assumption/presumption.
However, I admit that the timing of the job loss (October 2011) relative to Elton’s letter to the journal in which he blamed Ms. Martin (December 2011) is suspicious. Was the loss of grant funding a convenient excuse to remove Ms. Martin, so that she could be an absentee scapegoat? This, of course is just speculation based on incomplete information…no conclusions drawn.
In general, research staff have fixed term appointments that are contingent on extramural funding. It seems unfair but its completely normal procedure for them to be terminated when funding runs out. In my lab this includes several very accomplished professional researchers who contribute a great deal to research at my institution. Seeing these kinds of people’s livelihood at risk when other people (like for example Terry Elton) are benefiting or attempting to benefit from misconduct is one of the reasons why this is such an emotive issue.
Ah yes, but we also move people from grant to grant as needed, and I have known PIs to move someone to a grant *just before* it runs out. “Oh, so sorry, the grant that was funding you has run out!” The lab is running several other grants, but those “cannot accomodate you right now.” Then the person is “let go” and an advertisement gets posted for a new tech. Now, mind you, I have seen these techniques used with poorly performing staff, and not with whistle-blowing staff, but I imagine the same tactics could easily be deployed.
In this case, we don’t know whether the technician was “let go” because there really were no funds to keep her in the lab, or because it was an easy way to remove her from the lab without raising additional questions in the administrative ranks. Perhaps that information will be available at a future point, but right now everyone is speculating.
DrDoo – we agree…nothing is clear, just suspicious, and the timing, as you correctly said was convenient.
But, what is not speculation are the FACTS: Terry was a flipper, both vertically and horizontally, a squisher and a stretcher.
Damn! He would have made a great science fraud article.
May still do.
These miscreants are actually being paid to sit at their desks and design fraud!
Its a bloody disgrace!
Noah-
You are correct of course but in the current climate many labs (like mine) that have been fortunate to operate with more than one reasonable sized federal research grant for the past 20 years are now functioning with one grant under conditions where very low priority scores are needed to assure funding. If you look at Elton’s funding record he was clearly running into difficulties around the time that the misconduct came to light. Having said that, I think the onus should be on the institution to go out of the way to avoid scapegoating lab workers of all kinds who are caught up in this kind of mess. In the Eric Smart case discussed elsewhere on this blog the whistleblower and other lab members were supported by the institution for a year to give them an opportunity to find new positions.
A smart whistleblower would try to find another job before blowing the whistle, obviously. Even if this guy had another grant she could have switched to, did she really want to continue working for him? Why? Once you realize that the PI you work for is a fraud, you should start planning to leave (and if you choose to blow the whistle, do so AFTER securing another job). Probably she was young and naive, no idea.
Ideally yes, but in the present funding climate these kinds of jobs can be hard to come by.
You mention, correctly, the current funding climate. My worry about improperly managed publicity about these relatively infrequent fraud cases by the media is that the anti-science forces in this country (there are plenty in Congress and outside) will use them to make the case that science is all corrupt and a waste and to crack down even more on funding and increase micromanagement.
I have been astonished about how the level of discourse on this site has deteriorated in the last few days – since the closure of the Science Fraud site. I have followed both sites for some time and have used examples from both in lectures to graduate students about what not to do and why it is wrong. The current state of affairs is embarrassing. The emotional personal attacks and counter-attacks do not reflect the expected level of communication among thoughtful persons. The conjecture and accusations are not reflective of useful scientific discourse.
If this level of behaviour were encountered in a kindergarden class, I expect some “time-outs” would be assigned. Perhaps the moderators of RW should take this approach until some “big-people behaviour” can be re-established. Good grief, I’ve been recommending these sites to students for a good example of how to question behaviour in scientific publication.
ROB
There’s more! This is the culture at Ohio State University. For example, Robert J. Lee, In the same academic unit, obtained multiple NIH grants under pretense. In spite of the NIH ruling (see NIH report: http://www.ohiostcopharmacademicsflawed.com see page 2. ), Lee was promoted to full professor!
I am not associated with OSU, but when you say “This is the culture at Ohio State University.” you are making a broad statement about a large institution, about (at least) the majority of administration, departments, professors and implementation of policy. I have had some contacts with various people there, and I don’t think that generalization holds, but perhaps you can supply more data to be more convincing. I can certainly believe that in many schools, there are pockets of problems, and different levels of skill and willingness by the school to handle them, but blanket claims about a whole university require strong evidence or ought to be reframed to cover a smaller subgroup. From experience, it is clear that misconduct handling can vary widely among different departments in the same university.
I have had occasion to document much more pervasive alleged bad handling of misconduct, here, but I also said “No over-generalization should be made about the … faculty.”
I agree with John Mashey. I taught as OSU in the 80s. I was in a very different area than pharmacy. There are perhaps 10-15K faculty there. There are thousands of staff. There is the B-School, Arts and Sciences, the Med School, Engineering. There are probably persons who cheat, and admittedly these cases look very bad. But for the thousands and thousands who do not and who are committed to the research in their discipline, this is a very inappropriate statement. “This is the culture at Ohio State University.” This is a pretty huge stereotype.
R.J. Lee was promoted by the provost (Alutto). He’s the chief academic officer of the entire Ohio State University. In his letter, page 2 of the document: last page, Alluto states: “The ‘ethical lapses’ raised by the committee are without merit.”
The link leads to a pdf that does not seem to include a reference to Robert J. Lee. Could you please double check the link?
Leo: the very last document on page 2 is about the “Promotion decision of Associate Professor Robert Lee”.
Back to the R. J. Lee stuff mentioned by Wright, yet, looks like a fake laboratory was created, people were paid to work there with federal grant dollars. Also found this link discussing the promotion of Lee to full professor (www.ohiostateucoverup.com ).
1) I agree, but this actually raises some other issues where people may be jumping to conclusions, perhaps based on insufficient knowledge of academic misconduct inquiry/investigation/adjudication processes. For student issues, OSU has this committee, whose outputs actually include public annual reports about the frequencies of cases, their nature, That is far better public documentation than I’ve seen at most places.
2) Of course, that is for students, not faculty, who are also covered by OSU research misconduct policy.I’ve read dozens of such things, this is pretty vanilla and as best as I can tell, follows the Federal rules from OSTP, etc. Among other things, it says:
‘The second objective served by this Policy is to protect the integrity and reputation of the University and its scholars from false or unproven allegations of research misconduct. For this reason, the University assumes that a person accused of research misconduct is innocent of any allegations until the contrary has been established by a final decision reached under this Policy and the applicable disciplinary rules or procedures.
‘Time Requirements. The investigation shall commence within 30 days after completion of the inquiry. The investigation shall conclude within 120 days of its commencement or such time as required by federal law. If an investigation cannot be completed within this time period, the Coordinator shall submit a written request for an extension to the relevant oversight agency or funding entity, if required to do so by law or contract. The request shall explain the reasons for delay, and include an interim report on the progress of the investigation and an estimated completion date.’
3) I recommend the presentation from 2006, by Prof Kinghorn, interestingly at OSU pharmacology at OSU</a, especially the useful chart on p.5, adapted from Nylenha and Simonsen, which I found very useful when writing this, where I noted:
‘Honest errors happen. Incompetence can be hard to discern from intentional actions,2 but the likelihood of intent increases as one discovers:
Pervasive errors that almost always skew in support of a desired view. Each might be honest error, but en masse, seem increasingly intentional.
Silent suppression/omission of data/text inconvenient to desired claims.
Plagiarism of credible text to fake expertise, but with removal of key phrases, persistent weakening or inversion of key conclusions.’
Put another way, ~via Ian Fleming: “Once is an accident. Twice is coincidence. Three times is an enemy action.” Some actions may be obvious in isolation, others may require enough data to establish a pattern of misbehavior.
4) If a university investigation committee (usually composed of professors without conflict of interest, some of whom may have other things to do, and whose staffing may be difficult if particular technical expertise is needed, and quite often have people who have not done it before) investigates, but the preponderance of evidence does not rule out error or others’ responsibility, then they must say that, given presumption of innocence.
If someone else finds more evidence later, that shows a clear pattern strong enough to reopen the case [which seems to be what happened here, maybe], then that happens. But there are a range of interpretations possible about the university in question, ranging from very negative to reasonably positive. Anyone who expects large organizations to *always* perform perfectly, with hindsight … jsut doesn’t have much experience with large organizations.
a) The evidence was actually overpowering, but a conspiracy amongst the investigation committee and administration wanted to sweep it under the rug.
b) The original evidence was insufficient, and the committee failed to do due diligence in searching for obvious related evidence.
c) The original evidence was insufficient, but the committee really did try hard in its few months to find more, but just missed finding enough. ORI sees a *lot* of cases, is rather experienced at this, and does not have the same time limits … and sometimes may get complaints that the university might not see first. It would be astonishing if ORI did not occasionally find things that even a competent university investigation did not.
People may not realize how much effort it can take to find purposefully-obscured patterns. Even for blatant plagiarism, which is obvious once found, one may have an experienced “feel” that something with a few paragraphs of such is mostly mosaic plagiarism, but simply not be able to find the other sources. Not everything is on-line. Then, maybe someone else recognizes it, and it turns out that multiple papers copied big chunks from an obscure book. That’s simple compared to these murkier cases with multiple investigators, grad students, techs, etc.
5) Just as with researchers, but even more so for universities, one may need much more evidence to say much about a specific university’s diligence and competence to handle cases, and it can be fairly hard to find public documentation about them. See this, pp.68-69 for summaries of 2 messy cases that were actually well-documented, one returning misconduct verdicts and the other not, helpfully illustrating fact that universities must for their own reputations both rule misconduct when it happens and clearly rule it not so to protect faculty reputations when needed. I especially sympathize with people at UC-Boulder who ran the former investigation, generating a 125-page report, with 2 outside members, including have several people agree to serve, then drop out.
6) In the OSU case, there isn’t much data to go on, although I see ORI’s John Dahlberg was quoted:
‘Dahlberg said in a statement issued late Friday that “Ohio State University has a long history of working effectively with ORI to ensure that OSU maintains the integrity of its research programs and that ORI’s regulatory requirements are met.’
Dahlberg is quite careful in public statements, so that is some evidence for 5 c), but then Ivan is right to say it may spur people to look closer at the process, and it is probably a good idea to figure out whether this was closer to 5 b) or 5 c). On the visible evidence so far, I think the answer is “no clear conclusion about OSU’s handling.” Personally, I would wish for more public documentation of misconduct cases, so that universities and others could compare their performance.
It’s interesting that you mention Kinghorn. I noticed that he was a member of the committee that cleared Brueggemeier despite the editor’s letter (page 1) that included sanctions for Bruegemeier’s publishing practices. Did some searching and also noticed that Kinghorn reported to Brueggemeier and he also published papers with Brueggemeier. Another member of the committee D Vandre also published with Brueggemeier. If Ohio State University holds such high standards of moral integrity, then why did they put together a committee with such blatant conflict of interests. This seems to be a habit with Ohio State University as Dahlberg from the Office of Research Integrity had to ask that “anyone who had a personal or working relationship with Elton be removed from the panel investigating the matter.” This was after the first investigation “concluded that ‘irregular’ images in journal articles were caused by disorganization, not intentional malfeasance.”
John Mashey provides a very interesting link to presentation regarding research integrity by Prof. Kinghorn. Pg 14 of the linked document presents a true case history of research misconduct and its consequences i.e. termination of employment. So let’s consider how these 2 cases are different
1. The accused in the earlier case was a post doctoral student and not an esteemed faculty member
2. Better still was a foreigner!!
So here we go – in keeping with the high ethical standards he/ she was terminated!!!
Did any one question how despite being in a supposedly supervised/ mentored trainee position the said post doc managed to submit a manuscript twice without the PI/ lab head/ mentor being unaware?? No we don’t bother with such questions, we proudly present the case of how much we value scientific integrity and punish the guilty!!
Just a suggestion – in future presentations regarding scientific integrity and research ethics, cases of Elton, Lee etc can also be included alongside with the true case of the post doc so that the students know that as long as they are student/ post doc all the risk and blame is entirely theirs but if they are successful in landing a faculty position then the university will put all its might behind them to cover up any allegations. In case the fraud is so blatant that it cannot be covered despite all attempts then you will just have to spend some time without doing research but will still be paid your salary!!
BTW, also noticed that Platz and Brueggemeier had administrative ties. Now back to the Elton case, there were 3 faculty members from the OSU College of Pharmacy who co-authored these publications with Elton: CA Carnes, DL Knoell, and TD Schmittgen. As a follow up to Dahlberg’s comment about the committee composition and the investigation redo, makes one wonder if any of these co-authors or the investigators from Brueggemeier’s case were involved in the first investigation of Elton?
Amen, WB. High ethical standards dictate that underlings be fired and supervisors be exonerated or rehabilitated. It has been ever thus.
I’ve just sent this message to a number of commenters:
“Thanks for the spirited discussion on Retraction Watch. There are legitimate questions about how alleged fraud should be handled. The content and tone of some comments has, however, degenerated into name-calling and personal attacks, and that’s neither useful nor what we’re trying to accomplish with Retraction Watch. We won’t tolerate more, and will be moderating comments much more closely from here on in.”
[“denigrated” edited to “degenerated, thanks to “puzzled monkey” and “RW fan” for the correction.]
Thank you, Ivan. (Rather than “denigrated” I might say “deteriorated.”)
Indeed, thanks, and fixed. “RW fan” had the same suggestion on another thread.
I do see a fundamental problem with a University (or a Journal) investigating one of their scientists or publications. They do have a motive to keep things quit and undercover, which is a great conflict of interest. In my humble opinion, these investigations should always involve at least one outside party, ORI is preferred.
I didn’t see this one on the orignal list of retractions.
J Biol Chem. 2006 Jul 7;281(27):18277-84.
http://www.jbc.org/content/281/27/18277.long
Figures 2 and 3. Why does the notion of consistency cross my mind?
Can you have too much of a good thing?
Look at Fig 5B in the same paper, upper panel. Enlarge it and you will see that all 3 lanes minus angiotensin are triplicates and the lanes plus angiotensin for the mock control and anti-miR124 are duplicates.
In the lower panel, lanes 1 and 4 are mirror images.
Similar band duplication is going on in the western in Fig 2.
Irre,
you mean Fig. 3 not Fig. 2. Was just preparing a comment to Fernando:
Fernando
You mean that some of the bar charts look a bit similar? May be something in that.
But really this is not up to your usual sleuthing standards 😉
Figure 3C is an “immunoblot” and you surely know what that means.
In the pERK1/2 strip, the three negative controls are actually the same. The tell tale is the vertical stripe crossing the upper band at 1/3rd in. Two of the positives are obviously superposable but without such a striking extra feature.
In the ERK 1/2 strip, bands 1-3 and bands 4-6 exhibit a quite striking mirror symmetry, including the blemishes. Not sure I’ve ever seen this before.
Excerpt from the figure legend: “A representative immunoblot is shown. Results are representative of four independent experiments.” Blimey! In the figure, the data don’t amount to one independent experiment.
Thanks for drawing attention to this paper. Perhaps there is unfinished business in Columbus, Ohio?
I was a bit tired. I was just thinking about the error bars. Sometimes you can miss other things.
Is it a lost Breugel? But which Breugel?
Bruegel
Regarding Dipak Das, just came to know that he has passed away in India. At this point we all can forgive the past if possible.