'The mean side of the force : How regression to the mean can fool us', a new blog by Stephen John Senn author of DICING WITH DEATH ☑️ ow.ly/IJuK50LVWGX

'Lo and behold, the convolution has become a product in the generating function world. Life is full of surprises.' amzn.to/2nT4bAd

🎉 If you haven't checked it yet, highly recommend this very unique textbook "Information theory, inference and learning algorithms" by Sir MacKay combines Information Theory with Machine Learning 🆓 PDF available at: inference.org.uk/itila/book.html

Great preprint on justifying alpha levels using a Bayesian-frequentist hybrid perspective by Wulff and Taylor psyarxiv.com/3cbh7/ Given the recent paper discussing how few researchers were justifying alpha in practice: The reason is papers like these need to be written first.

'The Epistemic and Pragmatic Function of Dichotomous Claims Based on Statistical Hypothesis Tests' is now in press in Theory & Psychology. psyarxiv.com/af9by/ It provides, I think, the best defence of the use of dichotomous claims (e.g. p<.05) in the literature. 1/2

quando te falarem que inteligência artificial não tem ideologia, mostra isso aqui

Esse site é muito bom. Você digita uma música e ele usa IA para apontar músicas similares. É a primeira vez que vejo um que funciona de vdd. maroofy.com

More: Causation does not imply correlation: arxiv.org/abs/1505.03118 See also: The most counterintuitive facts in all of mathematics, computer science, and physics axisofordinary.substack.com/p/the-most-cou…

Isso aqui deveria estar exposto no MASP

🇰🇷📽️| Documentário: Cotidiano na Coreia do Norte: Meus irmãos e irmãs do Norte A documentarista Cho Sung-hyung foi a primeira cineasta sul-coreana a gravar um documentário na Coreia do Norte. Viajando pelo país, ela acompanha pessoas comuns em sua vida cotidiana e rotinas de

Rebecca Highly recommend this for bridging the gap between how statisticians and social scientists think about stats global.oup.com/academic/produ…

Every biologist should be familiar with the thirteen ways to look at the correlation coefficient. In particular #9. jstor.org/stable/pdf/268…

The central limit theorem achieved its final form around 1935. amzn.to/2Huehj5

Kording Lab 🦖 I used several rounds of "Critique this blog post. Identify points where things are unclear, where there are leaps in logic, where there are non sequiturs and new concepts and abbreviations not explained that appear. Identify awkward phrases. Suggest actionable improvements, like

se isso não é arte, não sei mais o que é...

New blog up! solomonkurz.netlify.app/blog/2023-05-0… In the 5th installment of my series on how to make causal inferences with the GLM and randomized experiments, we analyze count data with Poisson and negative-binomial models, and discuss robust SE's, too. #brms #marginaleffects #tidybayes

The Long History of Nobody Wants to Work Anymore 🧵

Toc toc: serei presuuuu

absolutely disgusting