Today brings two journal editorials about misconduct and retractions. They take, if we may, a bit of an optimistic and perhaps even blindered approach.
In an editorial titled “Scientific misconduct occurs, but is rare,” Boston University’s Richard Primack, editor of Biological Conservation, highlights a Corrigendum of a paper by Jesus Angel Lemus, the veterinary researcher who has retracted seven papers:
The authors Gerardo Jimenez, Leandro Melendez, Guillermo Blanco and Paola Laiolo provide in this issue a corrected version of the paper ‘Dampened behavioral and physiological responses mediate birds’ association with humans’ published in 2011 by Biological Conservation, 144/5:1702–1711. Data on bird physiology that could not be verified were removed from this corrected version. The results and discussion regarding bird behavior and habituation, community organization, and interactions, which constituted the main part of the original paper, are presented without change in this corrected Version.
Other journals took a harder stance on Lemus’s work. Primack writes his editorial:
Readers of this journal and other scientific journals might be concerned that this example and others reported in the press and scientific outlets suggest that scientific misconduct may be both widespread and increasing (Steen, 2010; http://retractionwatch. wordpress.com/), perhaps due increasing competition for jobs and research funding. However, we at Biological Conservation come to a very different conclusion.
In fact, this is the first case of serious scientific misconduct that we have seen over the past 9 years of the journal, during which time around 2000 papers have been published. Consequently it appears that scientific misconduct in this area of biology is actually quite rare. In fact, analysis by Steen (2010) suggests that retractions of scientific papers occur at a rate of 1–3 papers per 10,000 published. It is also possible that there are undetected cases of misconduct that were never uncovered. However, we think that over the years, these would have been discovered if they existed.
That’s a reasonable assumption, but based on the sorts of reactions to such discoveries that we see — short version: swept under the rug — we have to politely disagree. That’s true even though we were paying attention to the fact that Primack “used the words ‘serious misconduct,'” as he asked readers to note.
Primack also described — without giving any details — two hair-raising examples of alleged misconduct:
In one paper, the authors stated that a reviewer was threatening physical violence and may have been engaged in inappropriate cyber-stalking. In another paper, the author asserted that the two reviewers were deliberately trying to wreck his career. We carefully investigated these charges, but we did not find any evidence for the accusations.
He concludes:
While we at Biological Conservation continue to be concerned and vigilant over the issue of scientific misconduct, we also remain impressed by the high ethical standards of the vast majority of the scientific community, including both authors and reviewers.
Ferric Fang, who wrote a 2012 paper with Steen and Arturo Casadevall finding that most retractions were due to misconduct, said he is “in agreement with Dr. Primack that the vast majority of scientists have high ethical standards.” However, he tells Retraction Watch:
it would be incautious to conclude that the approximately 1 out of 10,000 papers that have been retracted reflects the true prevalence of research misconduct, and that most fraudulent papers have been uncovered by now. What we have learned from surveys by Brian Martinson and others (reviewed by Fanelli) is that approximately 1 in 7 surveyed scientists claims to have witnessed serious misconduct and about 1 in 50 admits to having committed serious misconduct. These numbers suggest that misconduct is likely to be substantially more common than 1 in 10,000. Moreover, it can take years for a fraudulent paper to be retracted from the literature, and our ability as a scientific community to detect falsified or fabricated data in publications is far from perfect. Misconduct is often uncovered because of whistleblowers who have inside information not available to the reviewer, editor or reader. In assessing the problem of research misconduct, I would seek the middle ground between sensationalism and complacency. While it would be wrong to try to discredit science as a whole because of a few bad apples, it would be dangerous to ignore the seriousness of the misconduct problem and what it may be telling us about stress in the scientific enterprise. We should all be able to agree that any degree of research misconduct is too much, so it is appropriate to try to understand why it occurs and what can be done to prevent it.
We found a lot of the same “it’s not a big problem” notes in an editorial out today by G. Richard Holt, editor of JAMA Facial Plastic Surgery. Holt writes:
To its credit, JAMA Facial Plastic Surgery has an exemplary record of keen oversight and selection of well-conducted research articles.
As far as I can determine, retractions of articles with misconduct in the surgical disciplines are less frequent than in other disciplines and the biosciences, although some recent retractions for scientific misconduct can be identified in the literature.
Holt listed three; we have a bunch more to add.
Retractions in the cosmetic and reconstructive surgery disciplines are fortunately very infrequent, although a couplet of recent retractions, in which the authors attempted to compare facial procedures experience during residency training among selected surgical subspecialties, deserves our attention. As noted in the journal, the reason for the retractions of these 2 articles was given as “(1) published proprietary information of Accreditation Council for Graduate Medical Education (ACGME) without permission from ACGME, and (2) incorrectly compared sets of information in the article(s).”18,19
Those two retractions were for Experiential Learning in Aesthetic Surgery Training: A Quantitative Comparison among Surgical Subspecialties and Reconstructive Surgery Training: Increased Operative Volume in Plastic Surgery Residency Programs, both of which appeared in a competing journal, Plastic and Reconstructive Surgery.
In June, there was in fact another retraction from the same journal, of “Histologic Results of Neuronal Anastomosis of the Microvascular Latissimus Dorsi Transplant,” because it “shares many duplicated items with another article published
in 2000 in another journal.”
But if you’re not looking for evidence of misconduct, you’re not going to find very much, are you? After all, as Holt concludes:
…the public must be made aware that the profession is acting on its behalf through oversight, fiduciary responsibility, and accountability. Korenman et al have stated the obligation very well, indeed: “The professional integrity of scientists is important to society as a whole and particularly to disciplines such as medicine that depend heavily on scientific advances for their progress.”
There are two fundamental issues here.
One is scientific misconduct that outrages the individual scientists because we are all trying to make careers and honest science is difficult. Those who cheat with duplicate publications, faked Western blots, etc, are hated simply because they are attempting to get an unfair advantage over those who struggle honestly to make a contribution, publish in a high quality journal, get a grant, tenure, etc. It’s a further outrage when journals, universities, and other organizations sometimes seem to turn a blind eye.
The other issue is scientific progress. Here it is not obvious that misconduct is as harmful as might be thought. Honesty and integrity in any individual publication is actually not important in the long term. A paper is valid not because it can be proven to have been done honestly, but because further work, by other people in related fields, yields results that are consistent. Your integrity matters not at all in determining whether the theories you are working on now are later considered to be sound or if they are rejected.
Conversely, it is quite possible for a fraudulent paper to be fundamentally correct, if only by coincidence. And it will be correct not only because no one ever finds out it was false, but because it is consistent with further work published later by other workers.
We do not know if Mendel did his work honestly in the way described. But in terms of scientific progress it doesn’t matter. His claims have been verified as fundamentally correct by later scientists.
It’s doubtful to me whether science fraud even delays or obstructs progress. If important, new claims are tested immediately. If less important they can’t really claim to be holding anything back.
Dan writes: “It’s doubtful to me whether science fraud even delays or obstructs progress. If important, new claims are tested immediately. If less important they can’t really claim to be holding anything back.”
This would be true if research wasn’t funded primarily by industries and governments that invest in predictable outcomes. As long as Pope Urban funds the astrologers, Galileo doesn’t have much of a voice.
Fraud is a serious crime but it’s not the most serious problem impeding science: http://blogs.discovermagazine.com/neuroskeptic/?p=126#.USZr-mfZb3A
During the 1970s, the Catholic Church claimed that child molestation was a rare occurrence, even as pedophile priests were shuffled between vulnerable congregations and their academics turned seminaries into homosexual brothels.
Since the 1990s, pseudo-policemen like William Bratton used COM-STAT to artificially reduce crime stats in ways that discouraged crime reporting, which helps mayors to reassure nervous voters, taxpayers and tourists.
In 2005, Congressman Barney Frank dismissed “alarmist” regulators who warned of the looming housing collapse, while financial managers who questioned Wall Street’s creative portfolios were silenced or fired.
In all of these cases, a silent majority acquiesced to a morally-confused leadership that retaliated against those who ask questions or make noise.
While Martinson’s assessment that one-in-seven has witnessed corruption may be closer to reality, that raises more questions. Why would six “highly ethical” scientists not notice what a seventh observes? Did the seventh remain silent, or did he speak out while the six ignored him? Was the seventh silenced or fired? What happened to his career, grants and publishing after ringing the alarm? In an academic vacuum where such questions are not asked, a retraction rate of one in 10,000 should be as reassuring as Gotham’s “safe streets” and America’s $16 trillion “investment.”
The progress of science (bio- and med-) was stopped by this:
1. Wrong direction: molecular, statistical and stochastical approach.
2. As the result of the first, previous research is forgotten.
3. Sloppy work that goes to the details here and there.
No scientist is answering far-reaching, fundamental questions. He does what makes papers fast. No wonder, he often cheats, he needs short answers to small questions. Every time his game is just for one-two months. If he had invested years to reach farther, he would care much more about answers. He was taught self-esteem and lost self-respect.
By the way, I made a proposal to do something about cheating:
https://www.retractionwatch.com/transparencyindex/
Don’t propose. Create. Don’t wait for others to do it. Do it yourself. Lauch the TI yourself if it is indeed your own brain-child, get some equations going, and make sure it is free and open access, or prepare for a serious backlash for promoting ghost ideas and solutions.
“we have to politely disagree”. I’d like to politely point out that without retractions and the associated, media-grabbing claim that fraud is endemic and that scientists are stealing from the public, Retraction Watch would not exist. So, of course the blog owners have to politely disagree, pretty much the same way in which a heart surgeon would disagree that diet can prevent heart problems: no heart surgeries = no business. In other words, the owners say that these editors have blinders on, but the reality is that EVERYBODY has blinders on (including the owners), determined by their own interests.
I understand this argument, but I don’t think it represents the situation in science. RW is not a heart surgeon – RW is more like a public health doctor saying that smoking causes lung cancer… and these editors are more like people who say ‘My mother smoked and lived to 92′.
The evidence that I am aware of is pretty clear on this – research misconduct is common and distorts in a substantial way the flow of research funding and other scientific resources; it destroys entire university departments and influences careers with many people giving up in science after working in a lab involved in misconduct.
There are no easy solutions but pretending the problem doesn’t exist at all, as these editors seem to be doing, can’t be the right approach, can it? Sounds to me like another step along Homo sapiens’ road to ruin…
It’s my understanding that RW is a labor of love, and does not bring in the big bucks. Or any bucks at all. Conversely, heart surgeons get paid. Also, I’m sure heart surgeons wouldn’t mind people eating better. There will always be other surgeries to do and as people get older, organs get weaker regardless. Plus, I’m not sure there is a glut of unemployed heart surgeons like there is of biomedical scientists. That seems like some sort of terrible analogy overall.
It looks like a hell-of-alot-of-work.
Try science-fraud-hunting!
I’ve done so much recently I have muscles growing in my brain!
to Dan who wrote “Conversely, it is quite possible for a fraudulent paper to be fundamentally correct, if only by coincidence” – do you have an example Dan?
In Biological Conservation, Primark wrote
“It is also possible that there are undetected cases of misconduct that were never uncovered. However, we think that over the years, these would have been discovered if they existed.”
I’d be a bit sceptical about this (as someone who works in ecology). Some would have been discovered, but the nature of fieldwork is that it’s often difficult to replicate it: biological systems have their own peculiarities, which can always explain away an odd result. So, whilst I’m confident that most (or almost all) ecologists and conservation biologists are honest, the few who are not might be difficult to spot. This does worry me, but unfortunately I don’t know the solution.