20th to Christmas done…. Dodging hail and gale force winds at the North Coast or Ireland. I’m running with @colincperry.bsky.social @michaelaodhan.bsky.social and two others for Action Deaf Youth: please donate if you like at gofund.me/63c50e5a

And I’ve done my day 13 run! 25 runs to Christmas. I am not a natural runner, but just about managing to hang in there. If you want to donate to action deaf youth, link in Aodhán’s post below. 🏃🏻‍♀️

So… day 9 of going from almost never running, to running every day. My knee is sore and I am not cut out for this.

Still going with 25 runs to Christmas with the crew in aid of Action Deaf Youth… Northern Irish charity for Deaf young people. Donate if you can! gofund.me/63c50e5a

I got my Bluesky reciept (bskypt.vercel.app) Favourite words are “Tb, Extremely, People, Amazing, Work, Stats”. Frankly, that is extremely on brand. 🤣

But while I’m being a fan-girl of people who create educational stats materials… you wouldn’t believe how often I look at, and how extremely helpful my old Kirkwood and Sterne Medical Statistics textbook is. I ❤️ it. Just read their (short) section on multi-level modelling and all makes sense again…

Hahaha… reminds me of this chart of UK statistical units / government / health / authority units. theidpblog.wordpress.com/wp-content/u...

Also…. The amazing worldpop solves at least SOME of these population size issues… .

Bluesky now has over 10 million users, and I was #9,028,546! Clearly late among the “early adopters”.

Nice to chat today Tom! V initial thought… is this a causal problem or a latent class / conditional probability problem? Berkson’s paradox (as you said, more likely to observe bacteraemia if either severe or MDR… less likely both). Photo here. @petermacp.bsky.social v good at latent class analysis!