"Here, we present two real-world examples demonstrating that sample overlap and relatedness is not a minor or theoretical concern but an important potential source of bias in PRS studies." Ellis et al. AJHG cell.com/ajhg/abstract/…

Alternatively, give them inaccurate data.

Every time there's one of these panic-inducing microplastic studies--every single time!--you look into it and find the evidence is very weak. Even aside from that there are all sorts of measurement issues in this field. We discussed them here: thestudiesshowpod.com/p/episode-32-m…

It’s not that much of a “blow” to patients because, to a close approximation, this drug has no effect (and yet it has serious side effects). Glad the NHS agrees. thetimes.co.uk/article/new-al…

This a nice psychological demo of how people store beliefs: just a set of conclusions that were accompanied by satisfaction or excitement at input time. Often works great, but very vulnerable to inception hacks that skip any derivation and cross check phase.

The “you might have had 5 years before going into residential care but with this treatment it would be 7 years” argument quoted here and repeatedly by John Hardy just isn’t supported by data. Patients and families depend on us for their understanding of what the drugs do.

Paret följde varandra genom eld och vatten, över land och hav.

Congratulations to our Head of Dept Belinda Lennox & the @uniofoxford Futuremakers: Brain & Mental Health podcast team. The series won the CASE Circle of Excellence Bronze award. Didn’t catch the podcast the first time ? Listen again here: tinyurl.com/3rfkdh3r

Our new Great Separation paper is out in AJS! Lead by Olivier Godechot, the headline finding is that top earners are increasingly working more together in the same workplaces and this growing economic segregation is unfolding in a broad set of countries. Please take a look!

To be fair, they have ranked Bret Weinstein's podcast as having more inaccurate info than InfoWars, which might not be wrong at this point x.com/zoecabina/stat…

Out today in Nature Human Behaviour: Genetic associations between non-cognitive skills and academic achievement over development nature.com/articles/s4156… Queen Mary University of London QM_SBBS UCL News A detailed thread👇

När du aldrig har tagit i för kung och fosterland: dn.se/kultur/catia-h…

(1/6) Because we didn't find racial discrimination, Reviewers thought our findings were wrong and our preregistered hiring experiments were flawed, leading to repeat rejections. Some lessons we learned, as outlined in our Discussion. onlinelibrary.wiley.com/doi/10.1111/17…

Hello GMC. Your public consultation on your intended regulation of physician associates closed on 9 May. Why haven’t the results been shared with the public? Please do so. [Day 1 of asking.] gmc-uk.org/news/news-arch…

At the 20th European Symposium on Suicide and Suicidal Behaviour #ESSSB24. Lots of sessions criticising risk assessment/prediction models. Some of this is justified, much of it not. 1/

On Nordic recidivism and rehabilitation It is widely believed that (a) the Nordic countries have very low recidivism rates, and (b) this is the result of a focus on rehabilitation rather than retribution. In this thread, I will argue that this is probably largely a myth.

Stockholm 💙💛

🎥 Pole Vaulter Mondo Duplantis 🇸🇪 defeats 400m hurdler Karsten Warholm 🇳🇴 over 100m, clocking 10.37s 🔥 📸 CITIUS MAG Weltklasse Zürich

Just left a sunny Stockholm for this.

Väldigt trist detta.