Science Fictions links for March 2024
An Open Science embarrassment; a slew of profiles of dodgy researchers; AI taking over scientific writing; and much more
One year! That’s how long I’ve now been doing these link-post newsletters on the Science Fictions Substack. I do hope you’re finding them interesting and/or useful.
If you know someone who might appreciate this newsletter—a monthly list of the best links to do with bad science, with my opinions left in—please do share it with them! Post the link in a group chat, tweet it out - whatever works to spread the word. I’d be very grateful indeed. Okay, here we go…
The links
RIP Daniel Kahneman. Do I feel bad that over the years I trashed both of his pop-science books, the first (Thinking, Fast and Slow) for being packed full of uncritical citations of bad research, and the second (Noise) for being interminably boring? No. But do I think that overall he still had a positive effect on the public understanding of science? Also no.
His actual scientific work in the 1970s and 80s with Amos Tversky was genuinely great, though!
A disappointing update on a study I shared here a few months back. It was authored by some of the leading lights in the Open Science movement, and apparently showed that techniques like preregistration improved the replicability of research. The authors claimed the study was itself preregistered… but now it looks like it wasn’t. This is extremely bad, and the way the authors have now said that they’re “looking for the preregistration document” is embarrassing for everyone concerned.
The lesson: trust no one. Not me, not your co-authors, not your Open Science heroes. No one.
The Francesca Gino debacle continues. This month Harvard unsealed their 1,300-page report on their fraud investigation into her work. Here’s a brief summary, including mention of Gino trying to scapegoat another professor for falsifying the data.
On a similar topic, the story of the “slow-motion downfall” of Didier Raoult (who you might remember as the hydroxychloroquine-for-COVID guy, but there’s so much more to it than that) is a must-read.
…as is this article, also on a similar topic, about the superconductivity researcher Ranga Dias and the swirl of misconduct allegations around his lab and retractions of his research. After the last year, I feel really bad for all the scruplous, high-integrity superconductivity researchers out there - your field is getting a very bad reputation!
Here’s Derek Lowe’s additional commentary on the story, noting that too many senior researchers browbeat their students into doing bad science, and too many universities aren’t interested in getting to the bottom of it.
Informative YouTube video on David Sinclair, the anti-ageing researcher with a history of massive claims… and unreplicable research.
Yes, the style of the channel is a bit too “I am hyper-optimising for YouTube” for my taste, but I promise the content is good.
Paper comes out claiming that banning prositution in Sweden led to an increase in rape rates. Almost immediately, a re-analysis shows the result is not real, and just due to an error in the statistical analysis code. Oops!
“A Columbia Surgeon’s Study Was Pulled. He Kept Publishing Flawed Data.” One of the top cancer surgeons in NYC has been accused of perpetrating a particularly lazy scientific fraud where he used the same images across multiple papers, claiming they were all from separate experiments.
I’ve rarely read any pop-science article that makes as many dubious (or just plain ridiculous) statements as this one on “how the climate crisis affects our brains”.
The worst thing is finding out at the end that the author has apparently written an entire book on this stuff. Groan.
Cool to see that the Bill & Melinda Gates Foundation is no longer putting up with the extortionate “article processing charges” required by Open Access journals, and is now mandating that research it funds be posted online as a free preprint by default.
And more good news: the BMJ (formerly the British Medical Journal but it now just goes by the initials - what’s that all about?) is now mandating data and code sharing for all articles published there. Should’ve been the case a long time ago, but credit where due.
A sudden and vast increase in the times the word “delve”—which happens to be a favourite of ChatGPT for whatever reason—is used in medical papers, starting in 2023. Pretty good prima facie evidence that scientists are using ChatGPT to help write their papers… and I really do hope it’s “help write” rather than “churn out filler text which they then thoughtlessly copy and paste”.
“Given that it was arguably Beethoven’s skills as a musician and composer that made him an iconic figure in Western culture…”. The only place you’d see a mind-numbing sentence like that is, of course, in a scientific paper.
P.S. The Studies Show
The podcasts continue apace. I was particularly pleased with our recent episode on the trouble with meta-analysis - discussing the pros and cons of a technique we perhaps rely on a little too much.
And we’re now doing short episodes in addition to the main, hour-long ones! The first one is on emotional intelligence. Please do give it a go, and subscribe on The Studies Show page if you like it!
Image credit: DALL-E
> Do I feel bad that over the years I trashed both of his pop-science books, the first (Thinking, Fast and Slow) for being packed full of uncritical citations of bad research, and the second (Noise) for being interminably boring? No. But do I think that overall he still had a positive effect on the public understanding of science? Also no.
Do you have a link to where I can read about this? My understanding was that the priming results were bad, but most of the other stuff in Thinking Fast and Slow was good. Overall he seems to have had a very positive effect on pop science, shattering the delusion that humans always act rationally, and encouraging people to consider their own biases. I think that's done far more good than the harm done by a few specific effects not being real.
"RIP Daniel Kahneman. Do I feel bad that over the years I trashed both of his pop-science books, the first (Thinking, Fast and Slow) for being packed full of uncritical citations of bad research, and the second (Noise) for being interminably boring? No. But do I think that overall he still had a positive effect on the public understanding of science? Also no.
His actual scientific work in the 1970s and 80s with Amos Tversky was genuinely great, though!"
"Kahneman wrote of his experience in Nazi-occupied France, explaining in part why he entered the field of psychology:
It must have been late 1941 or early 1942. Jews were required to wear the Star of David and to obey a 6 p.m. curfew. I had gone to play with a Christian friend and had stayed too late. I turned my brown sweater inside out to walk the few blocks home. As I was walking down an empty street, I saw a German soldier approaching. He was wearing the black uniform that I had been told to fear more than others – the one worn by specially recruited SS soldiers. As I came closer to him, trying to walk fast, I noticed that he was looking at me intently. Then he beckoned me over, picked me up, and hugged me. I was terrified that he would notice the star inside my sweater. He was speaking to me with great emotion, in German. When he put me down, he opened his wallet, showed me a picture of a boy, and gave me some money. I went home more certain than ever that my mother was right: people were endlessly complicated and interesting."
The formative years of behavioral economics were made in the 40s, because the SS soldier did not see the star inside his sweater most likely. Or the SS soldier was absolutely sure that no other SS near him could see the seams from Kahneman's inside-out sweater. https://en.wikipedia.org/wiki/Daniel_Kahneman#Early_life