Would like to see the details of this study, but this does track with my experience -- AI is likely not helping you as much as you think it is even for basic things like coding garymarcus.substack.com/p/breaking-n...

We see similar effects of grant terminations (as of June 30th). Over $8B in economic losses are spread across the country. Although affected institutions are only located in 191 counties, economic losses from terminated grants exceed $100K in 1,185 counties.

The #SCIMaP team announces a major update: an analysis of the economic impacts of the White House's NIH FY26 budget. Bottom-line, we estimate $46B in total economic loss, 202K lost jobs, and impacts in communities nationwide. Interactive map and shareable report: scienceimpacts.org/fy26 a 🧵

We are pleased to announce an experiment intended to stimulate academic discussion and exchange, centered on papers published in Econometrica 1/5

NIH spending more than pays for itself. It generates about $2.5 for every $1 spent. In terms of research productivity, NIH funding dwarfs the rest of the world. The only long-term potential rival is China. Politicizing the NIH's scientific processes will hand China a huge strategic advantage.

Please apply, this project involves quite a bit of fun work on combining data from observational studies with randomized trials to estimate heterogeneous treatment effects!

This view never gets old -- sad that Bridge 9 will be closed all summer and fall

Celebrating Black History Month: Initially drawn to combining math and finance, Sandra Safo's path shifted after she discovered data analysis during her studies. Sandra now develops innovative statistical and machine learning methods to advance health care research. magazine.amstat.org/...

Nothing like a commute home on fresh snow

MINNESOTA NIH Awards Funding: $718 M Jobs Supported : 7,992 Economic activity supported: $ 1.74 B www.unitedformedicalresearch.org/nih-in-your-...

Academia vs industry? I think this piece correctly describes most of the (dis)advantages of both in an almost fair way. Still, I opt for the intellectual freedom, although that's being constantly attacked by cuts to fundamental research. 🧪 www.harvardmagazine.com/2025/01/harv...

𝚝𝚒𝚗𝚢𝚝𝚊𝚋𝚕𝚎 0.6.1 for #Rstats is out! It's an ultra simple, super flexible, and 0-dependency package to draw beautiful tables in HTML, LaTeX, Typst, Word, PDF, and PNG. And for those who ❤️ documentation, 𝚝𝚒𝚗𝚢𝚝𝚊𝚋𝚕𝚎 ships with a billion pages of tutorials: vincentarelbundock.github.io/tinytable/

Trying again with a working link: We’ve just posted a new Contract Assistant Professor position at UMN Biostat & Health Data Science! Happy to discuss with anyone interested. #biostatistics #healthdatascience #academicjobs hr.myu.umn.edu/jobs/ext/365...

Reposting since people are actually on bsky:

I made a starter pack for health policy statistics! 📈🔢

When generalizing effects from RCTs to new populations, nonadherence can complicate interpretation. UMN Biostat PhD student Justin Clark's dissertation develops a principal stratification approach to deal with nonadherence when generalizing effects: arxiv.org/abs/2405.04419

For sensitivity analysis with PSM->matched pairs, there are rosenbaum bounds. For IPW there are a few options, all relatively new. Does anyone working in this area have a sense of which is the most commonly applied?

Thankful for the central limit theorem