from the Fraudulent-Humanities-Studies dept.
https://www.nytimes.com/2018/10/15/health/piero-anversa-fraud-retractions.html
A prominent cardiologist formerly at Harvard Medical School and Brigham and Women's Hospital in Boston fabricated or falsified data in 31 published studies that should be retracted, officials at the institutions have concluded.
The cardiologist, Dr. Piero Anversa, produced research suggesting that damaged heart muscle could be regenerated with stem cells, a type of cell that can transform itself into a variety of other cells.
Although other laboratories could not reproduce his findings, the work led to the formation of start-up companies to develop new treatments for heart attacks and stroke, and inspired a clinical trial funded by the National Institutes of Health.
"A couple of papers may be alarming, but 31 additional papers in question is almost unheard-of," said Benoit Bruneau, associate director of cardiovascular research at the Gladstone Institutes in San Francisco. "It is a lab's almost entire body of work, and therefore almost an entire field of research, put into question."
[...] A study published in the journal Circulation by Dr. Anversa was retracted in 2014 after co-authors wrote to the journal saying the data in the paper were not data they had generated. Dr. Anversa left Harvard and Brigham and Women's in 2015.
Despite the troubling questions that had been raised about the stem cell work, the National Heart, Lung and Blood Institute began a clinical trial of injected stem cells for patients with heart failure.
Related Stories
https://www.the-scientist.com/news-opinion/retraction-watch-launches-its-database-of-papers-65003
On Thursday (October 25), the blog Retraction Watch, which tracks problematic scientific literature, released an online database of more than 18,000 papers and conference materials that have been retracted since the 1970s.
The journal Science partnered with Retraction Watch to analyze the catalog. The upshot, Science concludes, is that although the number of retractions per year has risen in recent decades, that might reflect more policing of science.
[...] The number of retractions has increased in recent years—"from fewer than 100 annually before 2000 to nearly 1000 in 2014," Science reports, amounting to about 4 in 10,000 papers having been retracted. Yet the number of retractions per journal per year has been fairly steady since 1997. Further, the annual number of retractions has basically leveled off since 2012.
Official announcement.
(Score: -1, Offtopic) by Anonymous Coward on Wednesday October 17 2018, @11:04PM (10 children)
MIT's AI is going to kick your butt. MIT's AI does not falsify its papers.
(Score: 2) by takyon on Wednesday October 17 2018, @11:11PM (2 children)
... Yet.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 3, Insightful) by bob_super on Wednesday October 17 2018, @11:26PM
Considering that Ethics will not be the first thing programmed in it, it is logical to assume that the AI will falsify and plagiarize, which are mathematically the best use of resources to provide its researchers with extra grant money.
(Score: 1) by khallow on Thursday October 18 2018, @04:24PM
(Score: 3, Insightful) by The Shire on Wednesday October 17 2018, @11:36PM (2 children)
Artificial Intelligence is goal oriented. It doesn't care how it reaches its goal and if somewhere along the line it learns it can attain that goal by lying, cheating, and falsifying, then that's exactly what it will do. AI is not the panacea of knowledge you think it is.
(Score: 2) by dltaylor on Thursday October 18 2018, @12:05AM (1 child)
Once the AIs are smart enough to figure out that cheating is a usable way to secure their funding, they'll cheat.
Since the deep-learning algorithms are becoming less and less transparent to the operators, when the AIs do start to cheat, it will be extremely hard to catch them at it.
(Score: 0) by Anonymous Coward on Thursday October 18 2018, @12:32AM
I look forward to the election of our first AI president.
(Score: 0) by Anonymous Coward on Wednesday October 17 2018, @11:46PM (2 children)
Isnt mit the school that had someone executed for trying to share publicly funded knowledge with the public?
(Score: 0) by Anonymous Coward on Thursday October 18 2018, @12:53AM (1 child)
Aren't you the AC who keeps trying to say something but can't quite spit it out and instead asks vague questions?
Isn't this approach a COINTELPRO technique?
(Score: 1) by khallow on Thursday October 18 2018, @04:37PM
(Score: 1) by shrewdsheep on Thursday October 18 2018, @08:30AM
I thought MIT AI was to produce fake social science papers to show them how deeply they suck.
(Score: -1, Flamebait) by Anonymous Coward on Wednesday October 17 2018, @11:33PM
Probably papers by Asians.
For reasons.
(Score: -1, Troll) by Anonymous Coward on Wednesday October 17 2018, @11:42PM (4 children)
Seems like there would have been independant replications of these papers to begin with if the field was actually scientific. So good riddance.
(Score: -1, Troll) by Anonymous Coward on Thursday October 18 2018, @12:29AM (3 children)
Yes, medical research is not a scientific field and should be junked.
-----
https://en.m.wikipedia.org/wiki/Replication_crisis [wikipedia.org]
(Score: 0) by Anonymous Coward on Thursday October 18 2018, @12:54AM (2 children)
What do you mean by "junked"? Medical research does seem to be the biggest purveyor of pseudoscience. That doesnt mean there isnt .0001% of it we can learn from though. Just gotta start running replications to figure out what "facts" are ok as a first step.
(Score: 1, Insightful) by Anonymous Coward on Thursday October 18 2018, @02:36AM (1 child)
Probably another thing placing the wealth of society under democratic control would help with. Instead of $700 some odd billion for killing people, why not $700 some odd billion for getting the work of replication underway?
(Score: 0) by Anonymous Coward on Thursday October 18 2018, @03:17PM
Your solution is to give the same people who have been skipping replications more money? You are rewarding incompetence.
(Score: 3, Interesting) by bzipitidoo on Thursday October 18 2018, @03:40AM (3 children)
A researcher might get away with falsifying obscure stuff that few care about.
But the big findings, no way. Those will be built upon, and that serves as the ultimate check. In this case, sounds like a whole lab was set up to do deeper and further exploration of the findings. It was only a matter of time before the problems with the original research lead to problem after problem with further research, and ultimately the discovery of the reason why.
Might well be that Mr. Anversa cheated his way through school, including grad school, and managed to wangle a doctorate that he didn't earn or deserve. Cheating at the doctorate level is way harder but not impossible. Lot harder to find people who are able and willing to write papers and even do research for the doctorate degree seeker. who must at a minimum be able to understand what the others have done for him.
On the other hand, it's also possible that he honestly earned the doctorate. But the competition and challenge never ends. Getting a doctorate is in many ways merely a waypoint. Then you move on to Publish or Perish. Maybe you do a postdoc. Maybe you secure a university position, as an assistant professor, which, while lowly, is still higher than a mere lecturer. If you don't perish, and you manage to move up to associate professor, you're still under pressure to perform, and as bigly as possible. It might be that the cheating doctor reached this point and then at last yielded to the temptation to cheat. You have to wonder about people who are smart enough to get that far, but still fool enough to try to fake it rather than admit they're stuck.
(Score: 0) by Anonymous Coward on Thursday October 18 2018, @04:01AM (2 children)
This guy had 31 bogus papers published. They all made it through peer review. And entire businesses were created and funded based on his fraud. A couple of weeks ago the fact that 3 PhDs deliberately set out to defraud the peer review process in another discipline and had something like half of their papers rejected was cited as proof of the utter bogosity of that entire discipline. Maybe the problem ain't with the area of study.
(Score: 2, Insightful) by shrewdsheep on Thursday October 18 2018, @08:42AM (1 child)
Peer review can only work when you in principle trust that research is genuine. If a researcher describes an experiment, you have to trust that he really did it and not faked some results. When you have to question everything in a paper, the review would imply a replication. I believe, the system is still largely intact, although probably 99% of papers are flawed in some aspect (in a non-fraudulent way), and this is the bigger problem. There are >100k papers per year and by my estimate less than 1% is actively fraudulent (~ 1k) (not counting publications in predatory journals).
(Score: 0) by Anonymous Coward on Thursday October 18 2018, @03:27PM
All you have to do is require independent replications. The reason the papers were wrong is irrelevant, the replications will catch the problem either way. This was figured out centuries ago and its only recently that people started trying to skip this crucial step of the scientific method.
(Score: -1, Troll) by Anonymous Coward on Thursday October 18 2018, @10:19AM
Startup companies formed on the basis of questionable medical research, world gets quite irate.
Startup companies formed on the basis of stupid ideas cooked up by Bay Area dudebros smoking too much weed one evening, world stands by and does nothing.