Once you realize p-values are probabilities relating to observing data rather than probabilities of an hypothesis being true, you are already doing significantly better than most people

Why even use LLMs to write your methods section? This is all it takes

I'm sure this satisfies the criteria set out by expert methodologists for clinical prediction tools...!! 🤔 [ref: jclinepi.com/article/S0895-…]

Just posted my first article on Medium! 🚀 LLMs have been game-changers, but are they hitting a ceiling in reasoning capabilities? 🤔 I dive into why this happens and explore a potential 'easy' fix! Check it out and share your thoughts! medium.com/p/5611bcbbbd4c #ML #AI #LLM

New blogpost from my group on large language models: why the methods behind popular LLMs such as ChatGPT will soon be (or are already?) outdated 👇

👋 In today‘s #JournalClub📚, we read & discussed the article ‘There is no such thing as a validated prediction model’ by #BenVanCalster, Ewout Steyerberg, Laure Wynants @laurewynants.bsky.social, & Maarten van Smeden. 👉For more, read doi.org/10.1186/s12916… #PredictionModel #valdation #diagnosis #prognosis

Reminds me of this children’s game where you start with a sentence and you whisper that from one to the next and you end up with a completely different sentence

We're hiring! Want to work on developing better methods for causal inference with large-scale databases? All while helping answer research questions with direct societal relevance? Come join our fast growing Real-World Evidence team at UMC Utrecht! werkenbijumcutrecht.nl/vacatures/seni…

#statstab #152 Common Pitfalls in Research Design, Data Analysis, and Reporting Maarten van Smeden Thoughts: "Without such errors, research would, at the very least, be more informative". Good list for ECRs. ncbi.nlm.nih.gov/pmc/articles/P… #research #design #methodology #causalinference

Happy to present this NEW PAPER of the The STRATOS initiative on 'Regression without regrets: Initial data analysis is a prerequisite for multivariable regression' bmcmedresmethodol.biomedcentral.com/articles/10.11… Frank Harrell Marianne Huebner #stratosida

Have you really lived if you haven’t made a disastrous stats flow diagram yourself?

The article I posted on Medium (ai.gopubby.com/think-before-y…) about improved LLM reasoning with diffusion has just been featured in the TLDR AI newsletter! (a.tldrnewsletter.com/web-version?ep…) Thankful and happy its getting some attention. Curious to hear your thoughts!

Nice to see TRIPOD+AI reporting recommendations (link to guidance👇) mentioned in this NICE position statement on the "Use of #AI in evidence generation" 😀 nice.org.uk/about/what-we-…

Looking forward to talking about the blessings and curses of AI for clinical prediction modeling later today in Kopenhagen. And finally meeting Leon Eyrich Jessen in real life :) Slides ICYI: slideshare.net/slideshow/clin…

Mandatory periodic reminder. ➡️ Statistics: lines through data point cloud ➡️ Data science: lines through data point cloud ➡️ Machine learning: lines through data point cloud ➡️ AI: lines through data point cloud

This is wrong. All of them: sheet of paper through bee swarm

ok this is the best one I've seen yet