Science Fictions links for December 2023

Ruin your Christmas by reading depressing stories about bad science!

Here’s an early Christmas present for you: a selection of the most interesting “bad science” stories from the past month! If you’re looking for good (though perhaps quite dispiriting) stuff to read over the festive period, you’ve come to the right place.

(By the way, are you looking for something to listen to during that same festive period? Do check out my podcast The Studies Show, where we now have more than 20 episodes, each trying to get to the bottom of a controversial scientific issue!).

And so, to the links. If you haven’t already subscribed and you want more like this, do add your email address below. It’s free!

December’s best bad science links

• Presidents of major US universities are, er, not exactly looking great recently. But although much of the action has been on the East Coast, we shouldn’t forget that the ex-President of Stanford has been busily retracting his papers from Cell, Science, and (now) Nature due to “integrity issues”.

• One of the most unpleasant people I’ve ever interacted with on social media is Sander van der Linden, a psychologist and “misinformation expert” at Cambridge. Have any mild disagreement with him and you’ll be subject to a barrage of idiotic smears and sophistry. This is not a guy who appears to care about what’s true! So it’s been good to see some serious pushback against his absurd research on “inoculating” people against misinformation, firstly from philosopher Dan Williams, and secondly from biologist Ruxandra Tesloianu.

  • And if you were in any doubt that research on “misinformation” has become irredeemably political and should all be defunded immediately, take a look at this risible tweet.

• Remember the very embarrassing episode in the early 2010s (which has a Wikipedia page for some reason) where scientists claimed that conservatives were higher in “psychoticism” than liberals, but then realised they’d got the variables the wrong way around and actually liberals were higher? Well, something similar seems to have happened in a Lancet Public Health paper that claimed that hearing aids were linked to higher risk of dementia. It was retracted earlier this month. There but for the grace of god, etc.

• Psychedelics researchers are well aware of the big problems with blinding in their studies. Do they care? Well… no.

• In general, post-“replication crisis”, psychology journals won’t explicitly tell you that they’re not interested in replication studies. But it seems that some still will!

• Here’s a fun little exchange on exactly how many people died in the Black Death - and tracking down some false claims about the exact number.

  • And, relatedly, a great new article from Saloni Dattani on Our World In Data about pandemic death tolls.

• More than ten thousand scientific papers were retracted in 2023 - a new record. Given how many fake “paper mill” articles are out there, why don’t we try to beat the record again in 2024?

  • A huge majority of those retracted articles are from journals published by Hindawi. Looks like that “beleaguered” brand is now being sent to the glue factory.

• An attempt using a machine-learning model to assess the replicability of studies in psychology. Experiments are less replicable than observational/correlational studies, as you might expect. Surely that’s in part because experiments are usually more risky than the sort of research that asks: “does ‘extraversion’ correlate with ‘going to parties’? Wow, yes it does!”. Not to diss personality research at all, but the riskier parts of that research (linking personality to aspects of the brain, say), have been a heck of a lot less successful when it comes to replication.

• The US Food & Drug Administration is now naming and shaming pharma companies and researchers who break its rules about registering clinical trials or reporting accurate information about them. Long overdue!

• Also this month, French pharma company Servier had to pay a massive fine for marketing a weight-loss drug they knew for decades was harmful (and covering up evidence of the harm).

• If you think large areas of science are train-wrecks, you oughta see what’s been happening in history. Ian Leslie gives the latest update on a long-running story where an obviously false history paper was debunked, but was then defended (extremely unconvincingly) by the editors of the journal that published it, simply because it fit with their political preferences. In the process, they gave up on any commitment to truth and empiricism. Great stuff!

• “In 2022 alone, 1,266 non-physics authors published the equivalent of one paper every 5 days, including weekends, compared with 387 in 2016”. Are more scientists becoming massively more productive? Or are more scientists treating publication like a stupid game, sitting there chuckling every day as they realise they’re getting away with it? (It’s the latter BTW).

• And finally, a story to warm the cockles of your heart: fraudulent trachea surgeon Paolo Macchiarini, whose botched operations led to multiple deaths, is on his way to prison. Let’s hope he gets there just in time to spend Christmas behind bars.

Image credit: DALL-E