Good news everyone 🥳 Our (w @vincentab.bsky.social) primer on models as prediction machines (with the marginaleffects package) is finally officially published!>
journals.sagepub.com/doi/10.1177/...
Nice clear presentation of commonplace meta-analysis failure mode: Pooling coefficients that mean different things, because original models had different adjustment sets. A coefficient gets its meaning from the whole model, not just from the predictor variable it multiplies.
The XKCD Dependency meme
Computationally formalising a verbal theory is an eye-opening experience because it forces you to openly state all the assumptions that the theory rest on
Cool paper!
This is one of the best surprises I remember after having a baby
1. The paper with the implausibly large effects of Omega-3 fatty acids on mental health was now retracted. A little thread on the process where @ianhussey.mmmdata.io and I was involved.
www.sciencedirect.com/science/arti...
It’s inaccurate to say Mario is brave and Luigi is cowardly
Luigi is afraid of death, so he runs away from danger. Mario is afraid of living, so he runs towards death. Both brothers are cowards in their own way
A stack of chinos in different colors.
Someone asked if I could tell them where to buy a pair of good chinos. In this thread, I will tell you, but my answer is not simple. On the upside, I think this is a better approach when shopping for clothes and you can apply it to any kind of item. 🧵
I do not like this headline, Sam I Am.
Interestingly, the improvements in the cited study compare groups where only 25% complied with the detox. So just *being in a study* like this (and maybe the associated mindfulness around one’s own screen time?) explains much of the effect.
Finally someone dared to say it: effective psychodynamic therapy for anxiety disorders is plausibly based on exposure.
osf.io/preprints/ps...
Zero economic value. Enourmous social costs.
bsky.app/profile/more...
What makes behavioral interventions work beyond the psychological theory they implement? Their format, level of engagement, delivery modality?
In a new paper analyzing 274 interventions from 15 megastudies (4.1M+ participants), we tested 19 features: www.sciencedirect.com/science/arti...
1/2
The UK’s proposed study of social media reduction in kids doubles down on a failed experimental protocol & is likely to produce misleading results. It’s easy to tear apart a study post-hoc so I’m putting my marker down before the study begins: This is a bad study.
open.substack.com/pub/grimoire...
Does the body really keep the score? Steven Kotler & Karl Friston argues trauma isn’t stored in tissues. It’s a prediction error—the brain gets stuck in rigid threat expectations. www.frontiersin.org/journals/sys...
#Neuroscience #Trauma #PredictiveCoding #PTSD #FlowState #ActiveInference
Our new article on socioeconomic disadvantage, mental distress & functional impairment is out #OpenAccess in American Psychologist.
Led by the brilliant @emkbridger.bsky.social, with J. Maltby, @eikofried.bsky.social, co-supervised by me & @ludvigdb.bsky.social.
🧵
psycnet.apa.org/fulltext/202...
Great book!
Ricky Gervais shares a meme of him swearing into the camera that reads "You have every right to be offended. Just don't cry when no one cares!" Gervais comments on this meme "No idea who made this but I like it"
What kind of midlife crisis would I have to have to become this and how can I do everything in my power to avoid it?
Just out: A new RCT in The Lancet Psychiatry led by @cathycreswell.bsky.social and her fab team at Oxford shows a “screening-to-intervention” model cut child anxiety rates (61% vs 38% at 12 months) screening & through early detection + digital CBT + parent support
www.thelancet.com/journals/lan...
"Osmanovic Thunström says this suggests that some researchers are relying on AI-generated references without reading the underlying papers."
Could this really be the case?? /s
Seriously discomforting stuff.
www.nature.com/articles/d41...? A researcher at Gothenburg U made up a skin problem caused by too much screentime, and uploaded a popular science piece and "preprints" online. Soon, LLMs would treat the condition as real, and the fake study cited in real medical papers.
Many folk are surprised to discover thay Risk of Bias assessment tools tend not to interrogate the question “Did this study actually happen? And are its results trustworthy enough to believe?”
Jack’s Cochrane endorsed INSPECT-SR checks have done a lot to mainstream such Trustworthiness Assessment.
Photo of three men standing on top of rocky terrain. One is wearing a blue zip-up jacket with a blue chambray shirt and tan chinos. One is wearing a green-black checkered jacket with red chamois shirt, white pants, and cowboy hat. The other is wearing a blue parka with with blue jeans.
Two photos. In the first, three men are wearing bomber jackets with blue or white jeans. In the second image, one man is wearing an off white (possibly chamois?) shirt with blue jeans. Another is wearing a blue denim trucker with blue jeans. A third is wearing a blue quilted parka.
Two photos of groups of men. They are wearing Nordic sweaters, olive bomber jackets, flight suits, black-green checkered jackets, cowboy hats, beanies, and jeans.
Two photos. In one, there's a man wearing a brown tweed with a checked shirt. In the other photo, men are wearing bomber jackets and high-waisted cotton pants.
The Artemis II mission reminds me of these 1960s photos of NASA astronauts training in Iceland because they felt the terrain would be similar to what they'd encounter on the moon. The outfits are so great, even some 60 years later.
🚨New preprint and our results are rather concerning..
We find the "boiling frog" equivalent of AI use. Using large-scale RCTs, we provide *casual* evidence that AI assistance reduces persistence and hurts independent performance.
And these effects emerge after just 10–15 minutes of AI use!
1/
cover of the book "Bayesian Workflow" by Gelman, Vehtari, et al. Coming out later this year, in the summer probably.
I would have preferred to have the "draw the rest of the owl" meme on the cover, but this will do. Seems like it is on schedule, and we'll leave some typos so you know we didn't write it with AI.
This is such a great and weird picture. I don't think you could ask for a bigger depth-compression effect! #visionscience
Oncology has made impressive progress since the 1970’s. Et tu, clinical psychology?
New paper, out this week in PLOS One, suggests that most close relationship self-report measures are primarily capturing relationship quality 🧵
journals.plos.org/plosone/arti...
Wow.
Few people are aware of how much progress has been made against cancers.
Leukemia is probably the most striking example.
Before the 1970s, most children affected by leukemia would soon die from it.
Now, most children in rich countries survive.
ourworldindata.org/childhood-le...