From the past few weeks at Retraction Watch:
University of Washington (professor in the dentistry school attempting to dismiss a scandal involving conflict of interest in a promotion of homeopathic treatment)
Stanford University (professor of engineering who sued a journal for publishing a criticism of his research)
National Institutes of Health (immunology researcher who published an image that had been “inadvertently duplicated”)
Inadvertently duplicated.” I hate when that happens.
And more:
University of Aix Marseille (biologist who “attempted to silence critics of his work . . . using legal threats and harassment”)
University of California, San Francisco (brain tumor researcher who retracted a paper and has other papers with apparent issues of image duplication)
Jining First People’s Hospital in Shandong, China (“at least 35 scientists [sanctioned] for research misconduct”)
University of Stuttgart (materials physicist who “has refused to investigate strong claims of misconduct by an anonymous whistleblower”)
Johns Hopkins University (biologist who retracted 4 papers from PNAS after issues such as “there is a concern that the first and second lanes of the HIF-2α panel show the same data, and that the first and second lanes of the HIF-1α panel show the same data, despite all being labeled as unique data . . . there is a concern that the second and third lanes of the HIF-1β panel show the same data despite being labeled as unique data”)
Extra credit to the Johns Hopkins dude for saying, “We believe that the overall conclusions of the paper remain valid.” Kinda makes you wonder why these guys do experiments in the first place, given that the overall conclusions of their papers always seem to remain valid, no matter what. It’s stone soup all over.
But wait, there’s more:
National Cancer Institute (former postdoc “faked 15 figures and a movie”)
Faked a movie! I’ve never heard that one before. Except for the moon landing, of course.
And more:
University of São Paulo State in Brazil (veterinary researcher who reused images of different species of animals)
Columbia University (mouse researcher whose published results can’t be replicated by outside groups)
Columbia University?! Oh noooooooo . . . Dude in question is the “Paul A. Marks M.D., Professor and Chair of the Department of Genetics and Development at Columbia University Medical center.” I wonder what Paul A. Marks, M.D., would’ve thought of his money going to support unreplicable research. No problem, though; I’m sure the overall conclusions of the paper remain valid.
Still more:
Open University of Sri Lanka in Nawala (psychosocial counselor listed as an author on several papers outside her areas of expertise)
SUNY Downstate in Brooklyn (biologist suing the university for retaliating after finding that she committed research misconduct)
University of Pennsylvania (pharmacy researcher who had to retract “a 2017 paper on induced brain injuries in piglets”)
Induced brain injuries in piglets! I guess it would be hypocritical of me to shrink in horror at that one, given that I eat pork all the time. Still, it does seem creepy, given that Retraction Watch writes that the story was “making us wonder if the animals weren’t essentially tortured (if the experiments truly took place) as part of someone’s misconduct.”
OK, and here are a few more for you:
Southeast University in Dhaka, Bangladesh (neuropharmacology researcher who published more than 160 papers since 2016; lots of problems)
Katholieke Universiteit Leuven (Ph.D. student who added someone’s name as author on a paper without his knowledge or permission)
Massachusetts Institute of Technology (computer science researcher who published a controversial paper on vaccines, and the journal refused to publish a critical letter on it)
The MIT researcher, who does not seem to be an expert in biology, had this wonderful quote: “I mean, I’m a computer scientist, so I can go and look at . . . what’s the statistical distributions, and what not.”
And let me stop with this blockbuster:
White House Office of Science and Technology Policy (deputy director for climate and environment, used family connections to publish a paper that contained an error)
In addition to the horrifying bit that this person was involved in national policy, we get the wonderful line:
Notably, catch benefits from MPA expansion are still positive and substantial once the erroneous stock is removed from the analysis, and the qualitative conclusions of our work remain the same.
The qualitative conclusions of our work remain the same! Has it ever been otherwise?
Major institutions
So, just in a few weeks, we have the University of Washington, Stanford, NIH, Columbia, Johns Hopkins, MIT, the University of Pennsylvania, and the White House, in addition to various lesser-known entities.
The problems are happening at all levels of the scientific establishment.
I’m not saying that things are worse at the top. There’s some selection going on—“Columbia cheats on U.S. News ranking” or “Harvard fencing coach indicted” are bigger news than some scandal in a less well-known school, indeed I never would’ve heard of the University of Southern Mississippi volleyball scandal had it not involved Ben Stiller co-star Brett Favre—; but, yeah, when bad behavior happens at renowned universities, I’m bothered. Not surprised, but bothered.
The number of McEnroe moments [“”] (need a list?) I’ve had in the course of the last four decades makes me admire your patience. Keep up the good work. And please don’t think of retiring any time soon. (NB lucky owner of a copy of Woodworth [1938], published long before ANOVA-mania set in. Replication problems? Nope.)
“https://www.youtube.com/watch?v=KdDH7b51yzw”.
(How do you paste a link here? Angled brackets don’t seem to work, oddly enough.)
Link to “University of Southern Mississippi volleyball scandal” does not work.
Link fixed; thanks.
First, they are not animals. They’re “research organisms”. Just like “livestock” and “poultry” are not animals when on a licensed ranch, so animal cruelty laws do not apply.
Second, yea. Now think about 99.99% of studies being designed around the assumption NHST tells you what is “real”. Tens of billions of dollars are spent each year on torturing “research organisms” for no reason beyond a confused person needs to publish a paper.
Gonna start adding an “and what not” when I want to sound really scientific.
If your data doesn’t matter, why bother? Of course many do take this to its natural conclusion and they just make their data up. Unethical? Certainly. Internally consistent? You betcha!
What’s your basis from the claim that “many do take this to its natural conclusion and they just make their data up”?
“What’s your basis from the claim that “many do take this to its natural conclusion and they just make their data up”?”
The basis for that claim is that numerous papers over the last decade (surely in the hundreds if not thousands) have been discovered to have falsified data – in some cases for many different papers by the same researcher. There’s let’s see: Spider Guy; Refilling Soup Bowl Guy (Wansink); The Highly Regarded Insurance Startup CBO – Corporate Behavioral Officer – who’s data, by the way, wasn’t even internally consistent; ESP dude; yada yada yada. All these people are academics – members of the professorial class – not new age religion wackos (although I’m sure we could get a pretty long list of made-up junk from them too).
Oh, here’s a recent one from retraction watch:
https://retractionwatch.com/2022/12/13/cancer-researcher-banned-from-federal-funding-for-faking-data-in-nearly-400-images-in-16-grant-applications/
Here’s another recent one. The discovery of image falsification in one paper cost his lab five other papers:
https://retractionwatch.com/2022/12/29/researchers-lost-five-papers-soon-after-scientists-critiqued-another-of-their-papers-in-retraction-watch/
Image falsification is gaining traction these days apparently.
I have to admit that there is some poetic justice in taking the Mickey out of a mouse researcher.
“Columbia University (mouse researcher whose published results can’t be replicated by outside groups)”
This case seems like an outlier here, and is IMO unfairly included in this list. There is no guarantee that research will replicate (whatever that means); and if it doesn’t, it’s not the personal fault of the researcher who published the original finding. They are just reporting what they found; YMMV.
Of course, if the mouse researcher faked data or something like that, it’s a different story. But simply reporting results that can’t be replicated should not put the poor guy on this list of bad actors.
BTW, I can’t replicate a lot of stuff that I published previously. But I did my best in each case.
I forgot to add that despite the failures to replicate some of my key findings in my career, the qualitative conclusions of my work magically remain the same. I was right all along and have been for at least the last 20 years, if not longer.
Shravan:
You might be right about the Columbia University researcher. I haven’t looked at the case in detail. From the Retraction Watch story from Matt Warman, it looks kinda bad:
Looks like the “stone soup” link points to the post ‘A whole fleet of gremlins: Looking more carefully at Richard Tol’s twice-corrected paper, “The Economic Effects of Climate Change”’, which doesn’t even mentions “stone soup”. Are you sure you didn’t mean this link?: https://statmodeling.stat.columbia.edu/2013/10/11/gladwell-vs-chabris-david-vs-goliath/
The most recent post on Retraction Watch is titled “Influential paper linking recessions and left-wing voting patterns retracted.” The paper reported that people growing up in worse macroeconomic conditions were more likely to vote for left-wing parties and support more redistribution. The authors seem to have agreed to retract the paper, which was originally published in a top economics journal in 2014.
The retraction notice says:
> The authors and editorial team are retracting this article because the original findings cannot be replicated, likely as a result of an inadvertent coding error. While the original codes and data sets are no longer available, new analysis with a markedly similar data set does not support the original results.
I would like to see the results of the attempted replication. Hopefully they will be published eventually (even if just as a permanent preprint). Andrew, you might find this interesting given the connection to your work on “generational voting.”
Will:
It’s plausible to me that there’s no clear link in the data between growing up in recessions and left-wing voting. What data could they possibly have? There aren’t so many past recessions during the period where survey data are available, so maybe the effective sample size is about 5? From that point of view, the real mystery is how it is that anyone thought this could ever work as a research project.
What am I missing here?
Reading the original paper, there is some slippage between the title (“Growing Up in a Recession”) and the actual analysis. The authors use “regional variation in macroeconomic conditions in the U.S. to identify the impact of economic shocks on the formation of preferences for redistribution.” A secondary analysis uses cross-country data “linking preferences for redistribution to experiences of economic disasters during youth in a sample of 37 countries.”
It’s plausible to me that local economic conditions experienced during one’s youth influence lifelong voting patterns (even if that’s not really the same as “growing up in a recession”). There’s evidence that economic shocks can influence policy preferences, though the effects seem to be transient and the effect on vote choice is less clear. Margalit has a nice review of the literature here (which cites the retracted paper but also covers much more): https://www.annualreviews.org/doi/full/10.1146/annurev-polisci-050517-110713
I agree with you that there’s a very small effective sample size for national elections, but I don’t think the actual research conducted in the retracted paper was dead on arrival. That’s why I say it would be nice to see what the replication attempt turned up.
Will,
It’s very plausible to me that local economic conditions experienced during one’s youth influence lifelong voting patterns. The part that’s not so plausible is that there’s enough evidence in the aggregate data to identify such an effect. But I can ask Yotam what he thinks.