Judea Pearl Tim Feeney Not quite. Blyth failed to grasp the causal structure of the paradox in Simpson's example and misrepresented it. Because "most people" read Blyth’s paper but not Simpson’s paper, the misunderstanding has been perpetuated—from Wikipedia to the #BookOfWhy. x.com/_MiguelHernan/…

Had they left the OR & walked to any of the provider workrooms or clinic offices, they may have fainted at the sight of their software profoundly interfering with healthcare delivery for patients & providers alike

The coolest part of twitter is seeing threads like this. Where else could you watch a journal debate epistemology with Judea Pearl & Alfredo Morabia....

#reproducibilitycrisis affects more than just scientific studies!

How to properly give a code review by tomerpacific medium.com/p/how-to-prope…

How to write code you will love in the future by Lusan medium.com/p/how-to-write…

A chaotic mind leads to chaotic code by Dler Ari medium.com/p/a-chaotic-mi…

The Math Required for Machine Learning  ow.ly/nodv30oaFIK

Very nice to see NeurIPS adopting a "reproducibility checklist" and a code sharing policy. This helps bring much needed transparency and rigor to our field (also helpful for getting reliable stats about framework usage... 😌) medium.com/@NeurIPSConf/c…

1/2 This discussion seems to be pitting PO vs. DAGs as two competitors. They are in fact complementary. DAGs are used to encode what we know and PO what we wish to know. SCM is a bi-lingual framework, ucla.in/2umzd65, tying the two through structural equation semantics

2/2 The reason some researchers still attempt to encode what they know in PO is purely cultural -- they were deprived of DAG education. Glancing at how a simple chain is encoded in the two languages makes you wonder why any bi-lingual would choose PO over x-->y-->z #Bookofwhy

Automated Machine Learning for Professionals buff.ly/2HTFsUw

Tutorials for IPTW and some diagnostics github.com/pzivich/Python…

Will IT expert doubts delay the VA’s community care rollout? federaltimes.com/management/201…

Limits for the Magnitude of M-bias and Certain Other Types of Structural Selection Bias Flanders, W. Danaa; Ye, Dongnib Epidemiology: July 2019 - Volume 30 - Issue 4 - p 501–508 journals.lww.com/epidem/pages/a…

Our team is so cool, we celebrated my visa renewal! Visa approvals are getting tougher, but my team stood by me. Thank you @ahalvani Brian Sauer Zach Burningham Kelli Rasmussen @chunyanglijones Sophia Chien @DMorreall Liz Christina. 🙂👏🙏❤️ It's a great joy to be part of this team.

A tool for empirical equipoise assessment in multigroup comparative effectiveness research - Yoshida - 2019 - Pharmacoepidemiology and Drug Safety - Wiley Online Library onlinelibrary.wiley.com/doi/10.1002/pd…

Debugging in R: How to Easily and Efficiently Conquer Errors in Your Code | R-bloggers "If we’ve run our code and it has already crashed, we can use traceback() to try to locate where this happened." buff.ly/38W4XlP How did I not know this one?

"Before the pandemic, we had 30 million who were still uninsured. And that number has grown considerably during the pandemic." Former health insurance executive turned whistleblower Wendell Potter talks to Joanne Silberner about the US insurance industry ow.ly/8yVP50C0e7K

Another one for fellows who happen to follow me :-)