I have no idea why an engine which just guesses plausible sounding words without any real meaning behind them would be popular in management circles.

TIL that the NHS Health Research Authority approves **3500** research studies per year! www.hra.nhs.uk/about-us/wha... Thank you HRA

There are loads of well known assumptions, eg. difference in differences assumes parallel trends, instrumental variables assume there's no other causal paths. These assumptions aren't "violated", they're just more or less valid, or perhaps violated to greater or lesser degrees.

We looked at data on social media and mental health among 12-15-year-olds using data from Born In Bradford. I wrote a blog reflecting on the correlation between social media use and mental health, and our plans for The IRL Trial later this year. www.linkedin.com/pulse/from-o...

Imagine a world where you had more time than ideas

Both per capita and per unit area-- though I believe the standard per area illustration is that it has 2 popes per square kilometre.

Do you ever come across research studies that develop or somehow validate a survey questionnaire, but don't include the actual questions? What's that about?

If you'd like to hear about our planned trial of social media limitations among young people ("The IRL Trial"), please feel free to sign up for one of our webinars - we have one specifically for researchers, and one for stakeholders working in policy / services, etc: www.linkedin.com/feed/update/...

Nothing strengthens your argument like deleting half of it to meet the word limit

@newsfromgoogle.bsky.social Any chance you could make table borders a bit fiddlier in Google Docs? I like a challenge

2026 A.D. All the internet is occupied by AI. All? No! … One small indomitable footnote still holds out against the bots: "Message for Clarion WE KNOW YOUR LISTINING. The station has IP detection software"

If two people make a yes/no decision (eg. should a student pass their viva), they'll sometimes agree even if their decisions are random. If the event is rare or very common, they'll usually agree. Cohen's kappa accounts for random agreement. 0 = random; +1/-1 = complete agreement.

We're recruiting two research assistants to help us deliver "The IRL Trial" - a world-first randomised trial of social media limitations among 12-15 year-olds, and the effect on mental health. You'd be part of the Born In Bradford family. For more info and to apply: www.jobs.nhs.uk/candidate/jo...

Are you like me - do you love changing the fonts on your #rstats plots, but tire of fiddling around with extrafonts? Just load "ragg" and use the agg device - you can set the font in par (eg. `par(family = 'Verdana')` and you have direct access to system fonts! Thank me later.

Iain M Banks on avoiding self-destruction

Pockets of real deprivation

The hype around LLMs seems to be "yes the outputs aren't good and the costs and harms outweigh the benefits, but imagine if this or a different technology was actually useful"

New preprint 🥳 The city of Leipzig conducts large-scale surveys of adolescents. In 2023, the gender gap in life satisfaction has significantly widened, with girls declining more steeply than boys. What's up with that?> (work with @rmcelreath.bsky.social and @gregork.bsky.social)

Back on social media for a day and I already feel inadequate.

Journals - I'm totally up for doing peer reviews - sometimes I enjoy it! But it really is up to you whether you publish something

If you are not sure if method A or B is better, "sensitivity analysis" is probably not going to help you ...

I'm so tired of reading AI slop in essays / grants / articles etc ... I don't really have anything against AI aside from the pointless destruction of the planet ... it's just SO DULL

When reading authors' responses to your peer reviews, does anyone appreciate "we thank the reviewer for this valuable comment"?

Analyses of open health datasets can offer valuable contributions to health & medicine, but recent years have seen a proliferation of papers lacking robust or novel findings. In this Editorial, we provide guidance for conducting & reporting high-quality analyses using open datasets🧪 plos.io/4poBmZo

I love journals with retro titles

Adolescent mental health prediction models in the 1960s: At age 5-6 these researchers believed you could determine if a child would become delinquent based five factors in the environment of the child. www.nytimes.com/1965/01/31/a...

?untidy (after loading emmeans) "Dare to be un-tidy! ... Statistical analysis is not just a workflow; it is a discipline" #rstats

The Goldilocks research methods - www.the100.ci/2020/07/31/o... - from @dingdingpeng.the100.ci

Also, good studies addressing the same important research question as previous studies are deemed lower impact than studies addressing brand new poor research questions

I love this ... but @elsevierconnect.bsky.social why do you say the "peer review process is deemed to have been compromised", rather than the editing / publishing process? www.sciencedirect.com/science/arti...

Moral: Be careful what you ask for.

So absolutely pumped when I do a peer review and the publisher offers me a 25% discount on books for a limited duration and maximum total value

At what point in your career do people stop telling you that your definition of a p-value is wrong?

Really enjoyed this clear and concise presentation of the problem with studies that throw 10 different covariates into a multiple regression and interpret them causally. So obvious in hindsight that this is problematic, but I have definetly been guilty of this. bmjmedicine.bmj.com/content/4/1/...

Sunk costs and spite are basically the vodka red bull of the examined life

An LLM-controlled robot experiences a mental breakdown during a test to pass the butter from the kitchen. arxiv.org/abs/2510.21860 Some snapshots of its inner monologue from Appendix D:

📝 New Research Methods & Reporting article: Covariate adjustment in cluster randomised trials 📝 A practical guide for when and how to adjust for covariates. Read more ⬇️ www.bmj.com/content/391/...

Idea for a Christmas BMJ paper: submit loads of Christmas BMJ papers, then do a survival analysis of time to rejection. Then submit the survival analysis as a Christmas BMJ paper the following year

Also, sometimes people just either don't care, or more commonly are so in love with a hypothesis that they ignore other possibilities. Personally I don't think methods resolve this, it just needs hard case by case work. academic.oup.com/ije/article/...

It's possible to prove that under certain conditions, shrinkage plus data perturbation will lead to asymptotically zero false positive rate - but, it makes interpretation of the regression coefficients even harder.

We wrote an article explaining why you shouldn't put several variables into a regression model and report which are statistically significant - even as exploratory research. bmjmedicine.bmj.com/content/4/1/.... How did we do?

Sports stats

A really good talk on the stepped wedge design! Including some fascinating history of the first ever stepped wedge trial

Are cloud based data analysis platforms a massive practical joke

Annual PSA: if the interaction term in your regression model is significant, this suggests a deviation from additive or multiplicative effects depending on if you choose an additive or multiplicative model. Interactions can actually reverse depending on this decision!

When a corporate department gives you a template to complete, why is it always in Word 97-2003 format with floating text boxes and tiny margins

There are parallels with the "industry playbook" of citing insufficient evidence to support public health interventions

Thank you!!

The more I use R, the more I realise they did a pretty good job of deciding what you probably need in the basic package #RStats