Upgrade your #causalinference arsenal. A revision of our book "Causal Inference: What If" is available at https://t.co/3rrh0l8nFu Thanks to everyone who suggested improvements, reported typos, and proposed new citations and material. Enjoy the #WhatIfBook. Also, it's free.

This is getting increasingly harder to understand, @RobertKennedyJr . In the America First era, when a U.S. federal agency wants to do a randomized trial, it sends $1.6 million to Denmark? I don't run trials, but so many excellent U.S. researchers do. Aren't they good enough?

You're invited! 20th Kolokotrones Symposium: “Acetaminophen During Pregnancy and Autism: What Does Causal Inference Take?" Info in comments. In-person limited to Harvard ID holders due to space restrictions. Online attendance free & public. Register: https://t.co/cd1fWIhJ8h

I have a question for @RobertKennedyJr. You want to Make America Healthy Again. Exactly which year you’d like to go back to? 2007, 1992, 1983, 1976…? It’s hard to evaluate your policies without this piece of information. (Ok if a member of your staff replies)

Anyone interested in science in the U.S. should read this. https://t.co/WOx1lsQw2o

All CDC datasets have been anonymously uploaded here; https://t.co/EwH68Op366 Thanks to whoever made it possible that taxpayers can access the resources that were paid for by their taxes.

The CDC's page for HIV data has vanished. Who benefits from this? In what universe does this help the country? https://t.co/ulkvStjBbl

In our demo, we tested @Google's new Gemini 3 Flash model on complex government budget documents and saw more accurate extraction across documents compared to previous models.

2/ Immortal time may occur when individuals are 1) assigned to treatment strategies based on post-eligibility information or 2) determined to be eligible based on post-assignment information #TargetTrial emulation prevents it by synchronizing eligibility and assignment at time 0

1/ That "immortal time" is so frequent in survival analyses for #causalinference is fascinating. Because "immortal time" doesn't exist in the data, *we* create it when misanalyzing data. Our paper https://t.co/2TYg6hHCZj summarizes why immortal time arises & how to prevent it.

You're invited to the 18th Kolokotrones Symposium! “Causal Inference for Population Mental Health" 📆Nov 15, 2024 ⏰10:30 AM – 3:00 PM ET 📍@HarvardChanSPH / online This event launches our collaborator, the Population Mental Health Lab. Register now: https://t.co/PjblBXjVaD

Interested in working w/ @CAUSALab as a PhD student? We are hosting two Fall 2024 info sessions for prospective PhD students within @HarvardEpi. Session 1: Oct 3, 2024, 9:00 – 10:00 AM ET Session 2: Nov 25, 2024, 9:00 – 10:00 AM ET Register to attend: https://t.co/7cCGYvkcbE

We're partnering with @doodles to host an art challenge ( • -) • Doopies can smush together into objects as a way of avoiding danger or predators - make something made out of doopies for a chance to win! details ↓

In this Friday's PCT Grand Rounds, @_MiguelHernan of @CAUSALab at @HarvardChanSPH discusses causal estimands and asks whether we should be posing different causal questions in randomized trials. Join us! #biostatistics #pctGR ➡️ https://t.co/JFonVw5a4k

We aim to build opportunities for accessible #causalinference research. @CAUSALab awarded 42 tuition waivers for students to join us for #2024CAUSALabcourses! Recipients represented: 🎓 25 institutions 🌎 13 countries 📍 7 US states Congrats to this year's cohort!

See Dr. Miguel Hernan present at the Conference on Big Data in Biomedical and Population Health Sciences. Miguel Hernan, MD, DrPH, (@_MiguelHernan) is the Director of the CAUSALab (@CAUSALab) at Harvard T.H. Chan School of Public Health (@HarvardChanSPH). https://t.co/Eh2Xfffxvb

That👇should be the most comprehensive lesson-learned from the #covid19 pandemic. And people have to accept that politics makes also wrong decisions under these conditions. But somtimes it is even worse to do nothing. And retrospectively one is always smarter. Thx @_MiguelHernan

I recommend checking out the 'Tariffs Feed (full coverage, AI filtered)' stream on Synoptic. It has super comprehensive coverage of sources, both primary and secondary and its AI filtered for most important stuff - including potential resignations of Trump's cabinet (if you want

This week I discussed methods for health technology assessment @HTAiOrg: "Observational data (#RWD) can often be used to emulate a #TargetTrial, but we need more research to characterize questions that can only be answered by randomized trials." Let's learn the limits of #RWE

I was afraid my origin story would be revealed eventually…

The @TARGETGuideline team needs your help. If you work on #causalinference from observational data, please get involved in piloting the reporting guidelines for #TargetTrial emulation. Diverse viewpoints will make the guidelines more useful for everyone in #RWE. More info👇

Did you know that the LATE estimator was independently described in 1994 by Imbens & Angrist in Econometrica (@ecmaEditors) and Baker & Lindeman in Statistics in Medicine? https://t.co/OGML3c3iNg A delightful historical overview of LATE is now available https://t.co/CnJ0rOZbna

From our @PlayAcclaim family to yours, THANK YOU for riding with us, the new era of Acclaim The energy, support, and the belief means everything Here is a small gesture to show our appreciation More games, more personality and a lot more heart coming in '26 Happy Holidays🎄

The 2024 CAUSALab Summer Courses have officially begun 🎉 We are thrilled to welcome attendees from around the world and across various organizations. Here’s to an incredible two weeks of learning! #2024CAUSALabcourses