What if scientific fraud were illegal?
August 24, 2024 3:51 AM Subscribe
Legal consequences needn't be jail. You can lose your right do drive a car or work as a doctor or lawyer for doing those things badly or unethically, maybe you should be able to lose your right to work in research or academia similarly.
posted by Dysk at 4:18 AM on August 24 [33 favorites]
posted by Dysk at 4:18 AM on August 24 [33 favorites]
Reputations should matter; how-we-know-true-what-we-believe-true needs to be protected. Journalism used to have standards and reporters kept their reputations in good standing, so should scientific investigators.
posted by k3ninho at 4:23 AM on August 24 [1 favorite]
posted by k3ninho at 4:23 AM on August 24 [1 favorite]
Legal consequences needn't be jail. You can lose your right do drive a car or work as a doctor or lawyer for doing those things badly or unethically, maybe you should be able to lose your right to work in research or academia similarly.
Those are things that already require a license that can then be revoked. I don't think we want to create a science license.
Wakefield lost his medical license. Poldermans was diagraced and fired and should have lost hisedical license for breaching ethics regarding patient consent. None of this requires legal penalties beyond the maintenance of licensing standards by existing regulatory bodies.
And non-medical scientific fraud, while obviously bad, is much less concerning than the other kind.
posted by pattern juggler at 4:46 AM on August 24 [6 favorites]
Those are things that already require a license that can then be revoked. I don't think we want to create a science license.
Wakefield lost his medical license. Poldermans was diagraced and fired and should have lost hisedical license for breaching ethics regarding patient consent. None of this requires legal penalties beyond the maintenance of licensing standards by existing regulatory bodies.
And non-medical scientific fraud, while obviously bad, is much less concerning than the other kind.
posted by pattern juggler at 4:46 AM on August 24 [6 favorites]
The first is likelihood of abuse. It is easy to imagine repressive governments using this to make science and scientists they don't like go away.
Not even repressive governments. This is a tool that shouldn't be in the toolbox of the graybeards that already gatekeep academic discipline. You can find something you can convince a jury is "academic fraud" in any line of research, especially if you have the Esteemed Science People up there on the stand talking about how evil it is.
If this had been the case back in the 1950s, the people who were advocating for plate tectonics would have all been imprisoned or drummed out of geology or whatever the punishments were. The ulcers-are-from-stress people would have just driven the ulcers-are-from-infections people out of the research world.
Or, weird coincidence, all the prominent women keep committing fraud. Weird.
posted by GCU Sweet and Full of Grace at 4:53 AM on August 24 [26 favorites]
Not even repressive governments. This is a tool that shouldn't be in the toolbox of the graybeards that already gatekeep academic discipline. You can find something you can convince a jury is "academic fraud" in any line of research, especially if you have the Esteemed Science People up there on the stand talking about how evil it is.
If this had been the case back in the 1950s, the people who were advocating for plate tectonics would have all been imprisoned or drummed out of geology or whatever the punishments were. The ulcers-are-from-stress people would have just driven the ulcers-are-from-infections people out of the research world.
Or, weird coincidence, all the prominent women keep committing fraud. Weird.
posted by GCU Sweet and Full of Grace at 4:53 AM on August 24 [26 favorites]
This is a really difficult and multifactorial issue that we are getting Wrong; as noted by pj, there are issues with roping in the legal system. Elisabeth Bik is one of my heroes, and she is doing great work. The pollution of the scientific literature with fakes of all kinds is maddening, and a net harm against society, but it comes in significant part down to "who watches the watchers" and goodharts law.
There is a huge tension for "get more science out there, make it novel" and not just from the classical incentive error: bean-counters looking at how many pubs you have for funding, or paper/citation mills, or straight up confected data/image fraud.
There is a nasty overlap with a lot of the e/acc people who are tired of the slowness of publishing and their (justified) criticism about peer review being broken, and how how journals gatekeep good science based on prestige rankings...... I think that their efforts to break those things will lead to a net of more shit being spewed into the world, and a perilous reduction in signal:noise.
......especially because i) AI writes paper-shaped prose and is used for reviewing badly now and the volume of that 'work' is increasing....., and that ii) AI will be writing real papers soon. How will we tell the difference between i) & ii) - we are doing so poorly with our efforts at looking at HUMAN bullshit.
posted by lalochezia at 5:05 AM on August 24 [8 favorites]
There is a huge tension for "get more science out there, make it novel" and not just from the classical incentive error: bean-counters looking at how many pubs you have for funding, or paper/citation mills, or straight up confected data/image fraud.
There is a nasty overlap with a lot of the e/acc people who are tired of the slowness of publishing and their (justified) criticism about peer review being broken, and how how journals gatekeep good science based on prestige rankings...... I think that their efforts to break those things will lead to a net of more shit being spewed into the world, and a perilous reduction in signal:noise.
......especially because i) AI writes paper-shaped prose and is used for reviewing badly now and the volume of that 'work' is increasing....., and that ii) AI will be writing real papers soon. How will we tell the difference between i) & ii) - we are doing so poorly with our efforts at looking at HUMAN bullshit.
posted by lalochezia at 5:05 AM on August 24 [8 favorites]
What can happen when the government has the ability to prosecute scientists? Lysenkoism
posted by TedW at 5:05 AM on August 24 [7 favorites]
posted by TedW at 5:05 AM on August 24 [7 favorites]
.... and in the edit window, pj also notes that medical fraud is significantly different. See below.
How Should We Fund Scientific Error Detection?
"I have been asked many times how I would start a foundation to pursue scientific error detection.
My previous answers were too diffuse. After consideration, I have a better answer:
I would try to stop research fraud from killing people."
--
One of the things that isn't generally understand is in populations of hundreds of millions of people, that bad policy recommendations, that derive from corrupted processes kill hundreds of thousands of people EVERY YEAR . This is more than any war or genocide over a similar time period. See the above link for evidence of that number.
This isn't even in the contested realm of "how we treat poverty" or "tax avoidance" or "austerity" or "redistribution" vs. "individual freedom/private property"(sic) but in anodyne public health decisions that don't appear to have any of the hallmarks of a hot button, corruptible issue at all.
Science has saved hundreds of millions of lives (billions if you include ag. science) over the last century. But we are also cursed by knowledge: now we know things, and how to gather good evidence, we have a duty to gather the evidence well and act. And we aren't.
posted by lalochezia at 5:06 AM on August 24 [24 favorites]
Tagging on to what lalochezia just said: even the PRINCIPLES aren't fully there, much less the processes.
There's a Committee on Publication Ethics. It publishes guidelines for journal editors. They're voluntary and there is no certification program for them. And if you look, they're milquetoast as fuck, and they only address sins against other authors -- plagiarism, citation farming, and the like.
They don't address sins against research subjects. (What THE FUCK is "consider retracting unethical research," COPE. WHAT.) They only glancingly address fraud, which is a sin against scholarship. This is ridiculous. Without transparent standards and guidelines for this stuff, and transparent processes for judgment and penalty, how is anybody supposed to ride herd on it effectively without turning into the greybeards or bias launderers rightly decried above?
Editors skate because they can; until Retraction Watch became a thing they could just shut up about the shit that got past them and nothing would happen. Retraction Watch, however, isn't nearly big or funded enough to be more than reactive about research fraud. Bik and deevybee and others like them are amazing and I have near-infinite respect for them -- but they're not funded either. In the US, the Office of Research Integrity does good work, but guess what's also reactive because it's totally underfunded for the breadth of its mandate?
That's the problem. It's always money, innit.
posted by humbug at 5:45 AM on August 24 [7 favorites]
There's a Committee on Publication Ethics. It publishes guidelines for journal editors. They're voluntary and there is no certification program for them. And if you look, they're milquetoast as fuck, and they only address sins against other authors -- plagiarism, citation farming, and the like.
They don't address sins against research subjects. (What THE FUCK is "consider retracting unethical research," COPE. WHAT.) They only glancingly address fraud, which is a sin against scholarship. This is ridiculous. Without transparent standards and guidelines for this stuff, and transparent processes for judgment and penalty, how is anybody supposed to ride herd on it effectively without turning into the greybeards or bias launderers rightly decried above?
Editors skate because they can; until Retraction Watch became a thing they could just shut up about the shit that got past them and nothing would happen. Retraction Watch, however, isn't nearly big or funded enough to be more than reactive about research fraud. Bik and deevybee and others like them are amazing and I have near-infinite respect for them -- but they're not funded either. In the US, the Office of Research Integrity does good work, but guess what's also reactive because it's totally underfunded for the breadth of its mandate?
That's the problem. It's always money, innit.
posted by humbug at 5:45 AM on August 24 [7 favorites]
There is a monument to Giordano Bruno at Campo de' Fiori square in Rome, Italy, to commemorate the italian philosopher and astronomer, who was burned there for heresy in 1600.
posted by Winnie the Proust at 5:47 AM on August 24 [4 favorites]
posted by Winnie the Proust at 5:47 AM on August 24 [4 favorites]
The article suggests that “research misconduct may be hard to distinguish from carelessness.” I consider this attitude part of the problem. Carelessness is a form of misconduct— when providing medical care, when driving a car, and when publishing research. When the right kind of carelessness leads to funding, advancement, and tenure, there is going to be a whole lot of carelessness in the system. The same people who precisely audit their lunch receipt for correct specification of sandwich toppings are somehow unable to properly attribute raw data files in major publications.
These are solved problems. The principles of accounting and quality management are well designed to minimize these risks, but a whole lot of people do not want that kind of accountability.
posted by Vox Clamato at 6:22 AM on August 24 [14 favorites]
These are solved problems. The principles of accounting and quality management are well designed to minimize these risks, but a whole lot of people do not want that kind of accountability.
posted by Vox Clamato at 6:22 AM on August 24 [14 favorites]
These are solved problems. The principles of accounting and quality management are well designed to minimize these risks, but a whole lot of people do not want that kind of accountability.
I don't think they are solved problems in academia. Everyone can generally agree on best practices and what levels of care are required when deiving a car, or performing highly standardized medical procedures. And even there, there is some room for understandable lapses
Research is much more complicated. Not only is it collected in manu different ways, there are also numerous modes of analysis and interpretation of data. To detect errors or malfeasance tequires expertise in the techniques usef and the subject matter, and even then it isn't always clear cut.
posted by pattern juggler at 6:35 AM on August 24 [9 favorites]
I don't think they are solved problems in academia. Everyone can generally agree on best practices and what levels of care are required when deiving a car, or performing highly standardized medical procedures. And even there, there is some room for understandable lapses
Research is much more complicated. Not only is it collected in manu different ways, there are also numerous modes of analysis and interpretation of data. To detect errors or malfeasance tequires expertise in the techniques usef and the subject matter, and even then it isn't always clear cut.
posted by pattern juggler at 6:35 AM on August 24 [9 favorites]
That's the problem. It's always money, innit.
Yes, I think you've got it there.
The essay is a serious reflection on the issue - and she's very good about seeing the nuances and pitfalls of trying to regulate scientific publication.
But in the end I think she's overlooking the root cause: it's about money. About an academic system under severe economic stress, with decades of insufficient funding. We as a society (as nations, not corporations!) must fund more research, not less, in order to see scientists focus more on the discovery of knowledge and less on the competition over dwindling resources.
posted by Flight Hardware, do not touch at 6:38 AM on August 24 [9 favorites]
Yes, I think you've got it there.
The essay is a serious reflection on the issue - and she's very good about seeing the nuances and pitfalls of trying to regulate scientific publication.
But in the end I think she's overlooking the root cause: it's about money. About an academic system under severe economic stress, with decades of insufficient funding. We as a society (as nations, not corporations!) must fund more research, not less, in order to see scientists focus more on the discovery of knowledge and less on the competition over dwindling resources.
posted by Flight Hardware, do not touch at 6:38 AM on August 24 [9 favorites]
> Research is much more complicated. Not only is it collected in manu different ways, there are also numerous modes of analysis and interpretation of data.
Yes, but when raw data are collected, those unedited data need to be archived, indexed, and locked. All the modes of analysis and interpretation leading to published findings need to be fully documented and traceable and transparent to reviewers and auditors and anyone reading a publication.
Certain projects may not be able to meet those standards, and I would argue those may be exploratory or hypothesis-generating exercises, but the findings should not be publishable research.
If these standards were applied, we would have fewer publications, but they would be more reliable.
posted by Vox Clamato at 7:05 AM on August 24 [6 favorites]
Yes, but when raw data are collected, those unedited data need to be archived, indexed, and locked. All the modes of analysis and interpretation leading to published findings need to be fully documented and traceable and transparent to reviewers and auditors and anyone reading a publication.
Certain projects may not be able to meet those standards, and I would argue those may be exploratory or hypothesis-generating exercises, but the findings should not be publishable research.
If these standards were applied, we would have fewer publications, but they would be more reliable.
posted by Vox Clamato at 7:05 AM on August 24 [6 favorites]
These are solved problems. The principles of accounting and quality management are well designed to minimize these risks, but a whole lot of people do not want that kind of accountability.
I don't think they are solved problems in academia.
I would add to this point that scientists are just not trained in these principles as part of their education. In my own profession (environmental consulting) I work with a bunch of biochemists, fish toxicologists, and other environmental scientists. It's not research oriented, but it does involve smart people trained as scientists who are called upon to manage sometimes multi-million$ projects in a business setting. And I can tell you that, other than informal workshops we conduct internally, scant few of my colleagues have any training in project management, budgeting, contractual law, etc. Some people pick it up through experience, and some get really good at it (or at some part of it, while still being deficient in others), but many never do and are just terrible people to work for, even though they rise to high positions within the company by dint of good marketing abilities or other qualities. It's not that they're dumb or terrible people - it's that their formal education never gave them (or helped them build) that skill set.
posted by Pedantzilla at 7:14 AM on August 24 [14 favorites]
I don't think they are solved problems in academia.
I would add to this point that scientists are just not trained in these principles as part of their education. In my own profession (environmental consulting) I work with a bunch of biochemists, fish toxicologists, and other environmental scientists. It's not research oriented, but it does involve smart people trained as scientists who are called upon to manage sometimes multi-million$ projects in a business setting. And I can tell you that, other than informal workshops we conduct internally, scant few of my colleagues have any training in project management, budgeting, contractual law, etc. Some people pick it up through experience, and some get really good at it (or at some part of it, while still being deficient in others), but many never do and are just terrible people to work for, even though they rise to high positions within the company by dint of good marketing abilities or other qualities. It's not that they're dumb or terrible people - it's that their formal education never gave them (or helped them build) that skill set.
posted by Pedantzilla at 7:14 AM on August 24 [14 favorites]
Science, generally, is more like making music than making airplanes. If you're building airplanes, the worst thing you do matters and can kill a bunch of people. With music, you can ignore the mediocre stuff and enjoy the best parts that rise to the top.
So, IMO, there's no problem with having lots of results, lots of papers... More is generally better, because you've got more chances to produce a hit. It's a discovery problem, not a foundational problem.
The place where there's issues is when there's wholesale fraud, especially if it gets into high-cost, hard-to-reproduce studies.
posted by kaibutsu at 7:26 AM on August 24 [4 favorites]
So, IMO, there's no problem with having lots of results, lots of papers... More is generally better, because you've got more chances to produce a hit. It's a discovery problem, not a foundational problem.
The place where there's issues is when there's wholesale fraud, especially if it gets into high-cost, hard-to-reproduce studies.
posted by kaibutsu at 7:26 AM on August 24 [4 favorites]
pattern juggler: The first is likelihood of abuse. It is easy to imagine repressive governments using this to make science and scientists they don't like go away. Can you imaginea researcher on the effects of abortion access getting a fair hearing in Texas or Alabama if they are accused of fraud?
First thing that sprang to mind is knuckledragger dipshits using this against Fauci for his imaginary crimes against humanity during the pandemic.
posted by dr_dank at 7:26 AM on August 24 [5 favorites]
First thing that sprang to mind is knuckledragger dipshits using this against Fauci for his imaginary crimes against humanity during the pandemic.
posted by dr_dank at 7:26 AM on August 24 [5 favorites]
there's no problem with having lots of results, lots of papers
lit. review
posted by HearHere at 7:28 AM on August 24 [4 favorites]
lit. review
posted by HearHere at 7:28 AM on August 24 [4 favorites]
Lit review is fine, though.... Find the stuff that's on topic and has >N citations. If that's still too much, it's time to write a survey paper, and harvest a bunch of citations by making everyone else's lives easier.
There's increasingly good tools which help make sense of the heap. And talking to people consistently helps surface surface interesting things. No one expects your lit review to include bad papers. Just like music, really.
posted by kaibutsu at 7:39 AM on August 24 [1 favorite]
There's increasingly good tools which help make sense of the heap. And talking to people consistently helps surface surface interesting things. No one expects your lit review to include bad papers. Just like music, really.
posted by kaibutsu at 7:39 AM on August 24 [1 favorite]
kaibutsu, i agree with you: lit [quora]
fta: Scientific fraud kills people
words kill people? idk. application of words to action is a different story. for a helpful approach to medical situations, see e.g. Atul Gawande's Checklist Manifesto
posted by HearHere at 7:53 AM on August 24 [1 favorite]
fta: Scientific fraud kills people
words kill people? idk. application of words to action is a different story. for a helpful approach to medical situations, see e.g. Atul Gawande's Checklist Manifesto
posted by HearHere at 7:53 AM on August 24 [1 favorite]
Scapegoating is also a huge concern, both for basic fairness and the inevitable chilling effect of prosecutions. Why would anyone want to become a practicing geologist in Italy, for example?
posted by mubba at 7:55 AM on August 24 [5 favorites]
posted by mubba at 7:55 AM on August 24 [5 favorites]
......especially because i) AI writes paper-shaped prose and is used for reviewing badly now and the volume of that 'work' is increasing....., and that ii) AI will be writing real papers soon. How will we tell the difference between i) & ii) - we are doing so poorly with our efforts at looking at HUMAN bullshit.
For now, AI writing entire (published) papers is pretty rare. But researchers using AI as a tool in writing or editing papers is probably accelerating. Either that, or researchers suddenly love the word "delve" as much as ChatGPT does.
posted by a faded photo of their beloved at 7:57 AM on August 24 [1 favorite]
For now, AI writing entire (published) papers is pretty rare. But researchers using AI as a tool in writing or editing papers is probably accelerating. Either that, or researchers suddenly love the word "delve" as much as ChatGPT does.
posted by a faded photo of their beloved at 7:57 AM on August 24 [1 favorite]
I think the scientist is the wrong focus here. Scientists are generally responding to pressures within their institutions. Those institutions should be the ones to pay the price - one of your profs does a fraud? You have to pay the grant - and the overhead - back.
I think this would lead to more severe career penalties for scientists who commit fraud and incentivize universities to support careful and honest research. Which they sure don’t do right now.
posted by congen at 8:46 AM on August 24 [4 favorites]
I think this would lead to more severe career penalties for scientists who commit fraud and incentivize universities to support careful and honest research. Which they sure don’t do right now.
posted by congen at 8:46 AM on August 24 [4 favorites]
I think it would incentivize universities to create structures making it impossible to legally identify responsibility, personally.
posted by clew at 8:55 AM on August 24 [7 favorites]
posted by clew at 8:55 AM on August 24 [7 favorites]
I work in medical research. My team does data collection software, so we're very much behind the scenes people, but I still had to take a class in medical ethics. We got to learn about all of the egregious research done in the past, and how that informed the regulation environment we live in now. There were a lot of cases of people losing their PhDs, and a couple where their PhD advisor also lost theirs.
Back in the mid 2000s, I got to watch this play out. Some doctors where I work were interested in research being done at Duke, and asked some statisticians in my department about it. They looked at the research, and found problems. They told the doctors not to use Duke's research, and contacted the researchers with their concerns. The researchers made some, but not all, corrections. (And this is research on people, not animals.) Meanwhile, someone at NIH noticed the same issues, and started raising red flags as well.
No one at Duke or in the medical community cared. My coworkers published a paper about their use of "forensic statistics" to figure out how the Duke people got their results. At one point something like 15-20 heads of Biostatistics departments wrote an open letter to Duke asking them to shut down the research. Duke eventually opened an investigation, which determined that the researchers had done their paperwork correctly, and that was that.
Eventually, someone determined that the head researcher had lied about something inconsequential on their CV, and the hammer came down. The lead researcher was eventually convicted of fraud for all this. (And my coworkers appeared on 60 Minutes.) So, yay?
posted by Spike Glee at 8:58 AM on August 24 [9 favorites]
Back in the mid 2000s, I got to watch this play out. Some doctors where I work were interested in research being done at Duke, and asked some statisticians in my department about it. They looked at the research, and found problems. They told the doctors not to use Duke's research, and contacted the researchers with their concerns. The researchers made some, but not all, corrections. (And this is research on people, not animals.) Meanwhile, someone at NIH noticed the same issues, and started raising red flags as well.
No one at Duke or in the medical community cared. My coworkers published a paper about their use of "forensic statistics" to figure out how the Duke people got their results. At one point something like 15-20 heads of Biostatistics departments wrote an open letter to Duke asking them to shut down the research. Duke eventually opened an investigation, which determined that the researchers had done their paperwork correctly, and that was that.
Eventually, someone determined that the head researcher had lied about something inconsequential on their CV, and the hammer came down. The lead researcher was eventually convicted of fraud for all this. (And my coworkers appeared on 60 Minutes.) So, yay?
posted by Spike Glee at 8:58 AM on August 24 [9 favorites]
This isn't academic to me. Were these laws extant twenty years ago, I'm sure I would have been in court numerous times by now. I work in an area that has been high profile a few times in my career, and I am certain that people with passionately held beliefs would be taking me and my colleagues to court for the advice that we've been asked to provide.
Think of the reaction to COVID and the public health advice. Do you think Dr. Fauci would not be in a courtroom now if such a law existed? Many have tried already with existing laws. These would just give more tools to those who would use them in bad faith and for political advantage. I think of what happens to climate change scientists too. I'm neither of those, thankfully, but have had to walk through riots more than a few times in my career and been called all sorts of names to my face. I've seen part of that first-hand.
I do think scientific fraud is a major problem, and not just in the medical area. Any time there's a commercial interest, in tobacco science, in climate science, there have been found evidence that science had been not just fraudulently done, but suppressed for decades because the results were unfavorable to moneyed interests.
But I don't think criminalizing individual cheats and frauds is necessarily the right way to do it.
posted by bonehead at 8:59 AM on August 24 [6 favorites]
Think of the reaction to COVID and the public health advice. Do you think Dr. Fauci would not be in a courtroom now if such a law existed? Many have tried already with existing laws. These would just give more tools to those who would use them in bad faith and for political advantage. I think of what happens to climate change scientists too. I'm neither of those, thankfully, but have had to walk through riots more than a few times in my career and been called all sorts of names to my face. I've seen part of that first-hand.
I do think scientific fraud is a major problem, and not just in the medical area. Any time there's a commercial interest, in tobacco science, in climate science, there have been found evidence that science had been not just fraudulently done, but suppressed for decades because the results were unfavorable to moneyed interests.
But I don't think criminalizing individual cheats and frauds is necessarily the right way to do it.
posted by bonehead at 8:59 AM on August 24 [6 favorites]
So one of the reasons hormone replacement therapy (HRT) is still incredibly difficult for peri/menopausal and post-menopausal folk to timely access is due to a flawed 2002 “Women’s Health Initiative” study discussed here. The medical community continues to deprive folks en masse of lifesaving HRT based on this flawed study.
Does the behavior of those WHI researchers fit the generally understood definitions of “fraud” or even “negligence”? I’m not so sure. But my god, the consequences of the WHI’s decisions have been catastrophic, and relatively silent. They’ve quite literally fucked up the health of millions of people over the last 22 years.
I’d guess I’d have more confidence in the force of criminal statutes in cases of more bright line standards like “knowingly falsified data.”
posted by edithkeeler at 9:27 AM on August 24 [2 favorites]
Does the behavior of those WHI researchers fit the generally understood definitions of “fraud” or even “negligence”? I’m not so sure. But my god, the consequences of the WHI’s decisions have been catastrophic, and relatively silent. They’ve quite literally fucked up the health of millions of people over the last 22 years.
I’d guess I’d have more confidence in the force of criminal statutes in cases of more bright line standards like “knowingly falsified data.”
posted by edithkeeler at 9:27 AM on August 24 [2 favorites]
It’s hard for me to see how cases of blatant fabrication funded by government money aren’t theft.
posted by bq at 9:31 AM on August 24 [5 favorites]
posted by bq at 9:31 AM on August 24 [5 favorites]
It's a good, thoughtful article. But I'm a hard "no" on the idea of turning the decision of what constitutes scientific fraud over to the courts. As others have pointed out, that is opening a real can of worms for anti-science radicals to further target scientists whose work they don't like.
I think cases like the one described in the article, where fraudulent research led to the adoption of new clinical standards that worsened patient outcome, are at least as much a reflection of a poor culture of how to treat scientific evidence in clinical decision-making as they are a consequence of fraud. Clinical practice should not be rooted in reseach outcomes from only one researcher or research group. Even if the research is conducted honestly, there are far too many ways for unknown factors to influence results. Clinical practice should be based on work that has been replicated. This is the best defense against fraud, sloppy work, honest mistakes, and bad statistical luck.
I do agree that consequences for scientific fraud need to be severe, reliable, and timely. But this is a problem that needs to be solved by the scientific community, not courts.
posted by biogeo at 10:32 AM on August 24 [8 favorites]
I think cases like the one described in the article, where fraudulent research led to the adoption of new clinical standards that worsened patient outcome, are at least as much a reflection of a poor culture of how to treat scientific evidence in clinical decision-making as they are a consequence of fraud. Clinical practice should not be rooted in reseach outcomes from only one researcher or research group. Even if the research is conducted honestly, there are far too many ways for unknown factors to influence results. Clinical practice should be based on work that has been replicated. This is the best defense against fraud, sloppy work, honest mistakes, and bad statistical luck.
I do agree that consequences for scientific fraud need to be severe, reliable, and timely. But this is a problem that needs to be solved by the scientific community, not courts.
posted by biogeo at 10:32 AM on August 24 [8 favorites]
Wary of repression, since there is already state repression of scientists, at least in the environmental field, especially when it comes to documenting impacts of oil and gas extraction.
LSU gives direct curricular control to Exxon and Shell; and if you don't want the board of regents promoting you into a closet you will not cross their donors
posted by eustatic at 10:53 AM on August 24 [6 favorites]
LSU gives direct curricular control to Exxon and Shell; and if you don't want the board of regents promoting you into a closet you will not cross their donors
posted by eustatic at 10:53 AM on August 24 [6 favorites]
It seems like some thorough peer review could be a better mitigation here, preventing publication of incorrect or fraudulent research in the first paper. Start requiring line data and fully provenanced human subject data be provided along with these papers, the same way it’s required for regulatory drug and device submissions. Especially in these cases where medical decisions could potentially cite the research. Maybe all those big subscription fees we talked about in another thread could support this review….
posted by Tandem Affinity at 12:18 PM on August 24 [2 favorites]
posted by Tandem Affinity at 12:18 PM on August 24 [2 favorites]
After an earthquake, there was a prosecution of scientists in Italy who had said that earlier tremors didn't imply any elevated risk. I considered it blatant scapegoating. The legal ordeal lasted seven years before they were cleared.
As a general "we should be doing more" type request, I agree with the article. But government has trouble with this too, as anyone familiar with the times the Office of Research Integrity has been in the news will remember.
Not sure what to recommend, but I'd this is the classic sort of thing that will be prosecuted or policed when it's high profile, rather than when the misconduct is well establish or egregious. No one is going to prosecute a chemist for using white out on an NMR spectrum to boost the purity estimates (as was once, apparently, common practice) even though it's blatant fraud. Misinterpret some lab reports on vaccine research and a lot of prosecutors will be lining up to take a shot at you.
posted by mark k at 12:20 PM on August 24 [2 favorites]
As a general "we should be doing more" type request, I agree with the article. But government has trouble with this too, as anyone familiar with the times the Office of Research Integrity has been in the news will remember.
Not sure what to recommend, but I'd this is the classic sort of thing that will be prosecuted or policed when it's high profile, rather than when the misconduct is well establish or egregious. No one is going to prosecute a chemist for using white out on an NMR spectrum to boost the purity estimates (as was once, apparently, common practice) even though it's blatant fraud. Misinterpret some lab reports on vaccine research and a lot of prosecutors will be lining up to take a shot at you.
posted by mark k at 12:20 PM on August 24 [2 favorites]
Do we actually need new laws? Or would faking research already constitute wire fraud? Like, if you say you will do research X with NIH grant money, instead you do something fraudulent, is that currently okay?
posted by pwnguin at 12:30 PM on August 24 [1 favorite]
posted by pwnguin at 12:30 PM on August 24 [1 favorite]
more laws are always good for the legal profession.
posted by philip-random at 12:31 PM on August 24 [2 favorites]
posted by philip-random at 12:31 PM on August 24 [2 favorites]
My prior would be that the vast majority of incorrect science is not fraudulent science.
posted by kickingtheground at 1:18 PM on August 24 [9 favorites]
posted by kickingtheground at 1:18 PM on August 24 [9 favorites]
Duke eventually opened an investigation, which determined that the researchers had done their paperwork correctly, and that was that. Eventually, someone determined that the head researcher had lied about something inconsequential on their CV, and the hammer came down. The lead researcher was eventually convicted of fraud for all this.
I don't know which of the (many) Duke cases you're referring to, but it sounds like it could be Erin Potts-Kant, who falsified data on truly a ton of federal grant applications, or Anil Potti, who did the same fucking thing and also in publications as well, but it could just as well be Dan Ariely, who was just making up data in Excel to fit his hypotheses. As a Duke alum, it's really infuriating the shit that they let bigshots get away with in high profile fields.
All that said, and however, I come from a long academic line of whistle blowers who tell the truth who malicious prosecutors could have totally gone after. Gene Likens was one of the first scientists to present evidence of acid rain due to industrial pollution, including testifying before Congress. If industry leaders could have found a way to destroy him, they would have. He was the doctoral advisor for both my master's and PhD advisors.
My doctoral advisor's current research focuses on the effects of mountaintop removal mining on streams. Joe Manchin personally would love to destroy her.
Last year, Michael Mann finally managed to win a civil suit against the people determined to destroy him for telling the truth about climate change. If they could have prosecuted him back when the Hockey Stick Graph was first published, they absolutely would have.
As incredibly infuriating as scientific fraud is to me as a scientist, the right way to handle it is to fire people and permanently remove federal funding for provable cases of fraud. Prosecuting will be used to attack real scientists.
posted by hydropsyche at 1:23 PM on August 24 [23 favorites]
I don't know which of the (many) Duke cases you're referring to, but it sounds like it could be Erin Potts-Kant, who falsified data on truly a ton of federal grant applications, or Anil Potti, who did the same fucking thing and also in publications as well, but it could just as well be Dan Ariely, who was just making up data in Excel to fit his hypotheses. As a Duke alum, it's really infuriating the shit that they let bigshots get away with in high profile fields.
All that said, and however, I come from a long academic line of whistle blowers who tell the truth who malicious prosecutors could have totally gone after. Gene Likens was one of the first scientists to present evidence of acid rain due to industrial pollution, including testifying before Congress. If industry leaders could have found a way to destroy him, they would have. He was the doctoral advisor for both my master's and PhD advisors.
My doctoral advisor's current research focuses on the effects of mountaintop removal mining on streams. Joe Manchin personally would love to destroy her.
Last year, Michael Mann finally managed to win a civil suit against the people determined to destroy him for telling the truth about climate change. If they could have prosecuted him back when the Hockey Stick Graph was first published, they absolutely would have.
As incredibly infuriating as scientific fraud is to me as a scientist, the right way to handle it is to fire people and permanently remove federal funding for provable cases of fraud. Prosecuting will be used to attack real scientists.
posted by hydropsyche at 1:23 PM on August 24 [23 favorites]
>I don't think they are solved problems in academia...
>I would add to this point that scientists are just not trained in these principles as part of their education.
@Pattern Juggler, Pedantzilla-- Agreed that these are not solved problems in academia, and training and education are major factors in these failures.
posted by Vox Clamato at 3:01 PM on August 24 [3 favorites]
>I would add to this point that scientists are just not trained in these principles as part of their education.
@Pattern Juggler, Pedantzilla-- Agreed that these are not solved problems in academia, and training and education are major factors in these failures.
posted by Vox Clamato at 3:01 PM on August 24 [3 favorites]
We know for-profit scientific publishers are fraudulent parasites, so their journals retaining any quality remains more inertia than anything else.
After the initial work and writing, we should've distribution of papers, code, and data via pre-print servers, github, etc, followed by scientific discussions, reply papers, etc, and eventually slower auditing-like work.
Auditing work should've crystal clear declared biases, like the IPCC avoids over estimating warming and consequences, and skirts hard questions like tipping points, because they'd risk legitimizes deniers if wrong.
An auditing layer could theoretically bear some liability of course, but maybe this should occur within (re)insurance or similar: We could've semi-preditory insurance agencies who employ investigators and scientists to discover horrible things being done by buisnesses, and the pseduo-extort those buisnesses through offers of insurance going forward, and profit sharing from aiding victims bring cases for past miss-behavior. In other words, scientists would've no direct liability, but their employer wants worst & average case damage estimates. At present we always change the legal rules so that capital v capital fights never stop number from going up, so this sounds unlikely, but just dreaming up one example.
posted by jeffburdges at 3:35 PM on August 24 [2 favorites]
After the initial work and writing, we should've distribution of papers, code, and data via pre-print servers, github, etc, followed by scientific discussions, reply papers, etc, and eventually slower auditing-like work.
Auditing work should've crystal clear declared biases, like the IPCC avoids over estimating warming and consequences, and skirts hard questions like tipping points, because they'd risk legitimizes deniers if wrong.
An auditing layer could theoretically bear some liability of course, but maybe this should occur within (re)insurance or similar: We could've semi-preditory insurance agencies who employ investigators and scientists to discover horrible things being done by buisnesses, and the pseduo-extort those buisnesses through offers of insurance going forward, and profit sharing from aiding victims bring cases for past miss-behavior. In other words, scientists would've no direct liability, but their employer wants worst & average case damage estimates. At present we always change the legal rules so that capital v capital fights never stop number from going up, so this sounds unlikely, but just dreaming up one example.
posted by jeffburdges at 3:35 PM on August 24 [2 favorites]
We know for-profit scientific publishers are fraudulent parasites, so their journals retaining any quality remains more inertia than anything else.
Sure, that's why many of us choose to only publish in non-profit society journals. They still have page charges, but having served on the board of my society, I understand where that money goes. And I know the editor and all the associate editors, and they are definitely not fraudulent parasites.
posted by hydropsyche at 5:45 PM on August 24 [5 favorites]
Sure, that's why many of us choose to only publish in non-profit society journals. They still have page charges, but having served on the board of my society, I understand where that money goes. And I know the editor and all the associate editors, and they are definitely not fraudulent parasites.
posted by hydropsyche at 5:45 PM on August 24 [5 favorites]
bonehead: This isn't academic to me. Were these laws extant twenty years ago, I'm sure I would have been in court numerous times by now.
Facts-in-law aren't always the same as the scientific consensus, SLAPP exists and the courts would have to use your scientific epistemology to assess the claims made against you.
posted by k3ninho at 9:27 AM on August 25
Facts-in-law aren't always the same as the scientific consensus, SLAPP exists and the courts would have to use your scientific epistemology to assess the claims made against you.
posted by k3ninho at 9:27 AM on August 25
Speaking of using existing laws: DoD sues Georgia Tech for fraud. In this case, the alleged misrepresentation isn't so much in the publications but in the security of the surrounding process.
posted by pwnguin at 10:27 AM on August 25 [3 favorites]
posted by pwnguin at 10:27 AM on August 25 [3 favorites]
It is easy to imagine repressive governments using this
Almost every tool created by progressives to restrain regressive behaviour ended up being overtaken by regressives, spun 180°, and used on progressives.
The sentient interia burned in this infinite struggle is the main reason why we are not incandescent orbs flitting among the redwoods right now.
posted by CynicalKnight at 11:25 AM on August 25 [4 favorites]
Almost every tool created by progressives to restrain regressive behaviour ended up being overtaken by regressives, spun 180°, and used on progressives.
The sentient interia burned in this infinite struggle is the main reason why we are not incandescent orbs flitting among the redwoods right now.
posted by CynicalKnight at 11:25 AM on August 25 [4 favorites]
FWIW, in the US if you fabricate data in FDA pharmaceutical trials you absolutely can go to prison.
Also, the analogy to medical malpractice is interesting. I have experience working in a city where a nurse was criminally convicted and sentenced for a single medication error. I think it's human nature to make errors and it's the role of constructed systems to prevent those errors, and I have found the ability for criminal prosecution to change human nature is pretty lackluster. But it sure as hell made everyone exponentially more anxious.
posted by midmarch snowman at 7:02 AM on August 26 [4 favorites]
Also, the analogy to medical malpractice is interesting. I have experience working in a city where a nurse was criminally convicted and sentenced for a single medication error. I think it's human nature to make errors and it's the role of constructed systems to prevent those errors, and I have found the ability for criminal prosecution to change human nature is pretty lackluster. But it sure as hell made everyone exponentially more anxious.
posted by midmarch snowman at 7:02 AM on August 26 [4 favorites]
I was referring to Potti, who was experimenting on people with Cancer. He was supposed to find the best treatment for a given cancer, given certain biomarkers. (We've done similar studies.) What he actually did was find the worst treatment, and treated patients with that, instead.
posted by Spike Glee at 9:58 AM on August 26
posted by Spike Glee at 9:58 AM on August 26
> Ctrl+F "profess"The professions originally were bodies of knowledge-and-practice that had to be self-governing because of the nature of their knowledge-and-practice. You profess, you take an oath, to uphold the standards of your field.
> Phrase not found
Sometimes that means throwing a malicious or incompetent member out of the field. This isn't a recourse to the legal system, it's a statement by one body of practitioners. (The legal system might have opinions about the way in which the body does it, ofc.)
It's kind of interesting seeing which professions still try to do this and which have totally given up. I'm taken aback that scholarship, research, the profession we call "professors", does seem to have totally given this up at every stage *except* granting the PhD.
posted by clew at 1:20 PM on August 31 [2 favorites]
« Older The Bicycle of Payments | Lavinia Fontana, the Self-Fashioned Painter Newer »
The first is likelihood of abuse. It is easy to imagine repressive governments using this to make science and scientists they don't like go away. Can you imaginea researcher on the effects of abortion access getting a fair hearing in Texas or Alabama if they are accused of fraud?
There is also the decision of who determines what is fraud and what do we do with the guilty? Scientific fraud is often very hard to prove and often turns on subtle distinctions. A lay jury would probably need to be trained to even understand some of the issues involved, like p-hacking, much less determine whether they occured. And setting up science courts seems like a nad plan.
And even if you have someone dead to rights, putting them in prison in most countries would be unethical. It is also unlikely to be effective. Greater punishment seldom dissuades people from crime. Greater likelihood of being caught does. If we want to reduce academic fraud, forcing institutions to be more rigorous in catching and reporting it seems more ethical and effective.
posted by pattern juggler at 4:15 AM on August 24 [41 favorites]