Excited to share that @PCORI has awarded funding for our proposed work developing Bayesian machine learning methods for causal inference! Thanks to Roee Gutman (co-PI) & others at @Brown_SPH https://t.co/dtOXdg0mvw @Brown_Epi @BrownUResearch @BrownMedicine @BrownUniversity

I’m hiring a Student Researcher at @GoogleDeepMind. This research role centers on topics of open-ended self-improvement and discovery with LLM agents. 📍 Location: London 🗓️ Duration: 6 months, 100% 🚀 Start date: June or July 2026 Apply now using the links below👇

Bayesian nonparametric models allow flexibility in regions w/ lots of data, while allowing priors about sensitivity parameters drive inference in regions w/o data (see bottom-right plot). Uncertainty about *all* unknowns flow into a single posterior for the causal quantity!

Are some patients missing outcome info? Condition on data, make inferences about unknown {regression lines & missing values}. Think the missingness is not at-random? Condition on data, making inferences about unknown {regression lines, missing values, & sensitivity parameters}

Why I find Bayesian nonparametric causal inference compelling in one figure. The key distinction is btwn (1) "known" vs (2) "unknown" quantities: Make inferences about (2) conditional on (1). Want cond. avg trt effects? Condition on data, make inferences about regression lines

Congrats to @amalsargsyan & @GevTamamyan! When I first got involved, my mind went straight to the stats - but from black markets to the complexities of an out-of-pocket system, this was so much more. One step towards better cancer policy in Armenia🤞 https://t.co/wUOJByOG32

We also discuss differences and similarities with methods for irregular visit processes that inverse-weight by the visit process intensity.

... and how identifiability conditions may be read off a Single World Intervention Graph (SWIG) template for the implicit DTR.

We discuss formalize connections between g-methods that use discrete-time versus continuous-time models adjustment models and the realtive pros/cons of each...

Progress can be made by (1) casting waiting times between decisions as potential outcomes of previous treatments and (2) framing subsequent decisions as outputs of an implicit dynamic Treatment Rule (DTR).

In causal inference problems w/ sequential treatments, long stretches of time may elapse between treatment decisions This paper, in press at Epidemiology, was really fun to write: it discusses biases that may arise & corresponding adjustment via g-methods https://t.co/RkzMhdxuP3

I hope it’s a helpful resource for people adopting Bayesian methods in their causal work. Associated demo code here: https://t.co/0Doi3xhNsH

I originally wrote this to share with trainees but was recently encouraged arXiv it. I touch on several Qs: Why does unit-level inference need stronger assumptions? When do/n’t we impute unit-level counterfactuals? How does this differ from g-computation? https://t.co/tFI4U6MvdD

Streaming live now on YouTube: https://t.co/l38IO4wolz Andrew Gelman is talking about @StatModeling is talking about "Taking Our Models Seriously"

StanBio Connect 2025: Advancing Biomedical Research with Stan This event is streaming live tomorrow and I’ll be speaking about some work on Bayesian causal inference with survival outcomes. https://t.co/4mVNFIRp91

I’m teaching a 3-hour session on Bayesian causal inference at Penn Causal Inference Summer Institute, 5/27-5/30. Virtual attendance options are available. There are sessions on many other really cool topics as well -check out the agenda: https://t.co/bGods91YQ6

Congratulations to the students, staff & faculty honored Wednesday at the #BrownSPH 2025 Dean's Awards Ceremony!🏅 “You remind us that in moments of uncertainty,” Dean @ashishkjha said, “the mission of our school becomes not just relevant, but essential.” https://t.co/pZ52bmUtVo

New paper here w/ Tony Linero on Bayesian causal inference: Independent priors on propensity score & outcome models often imply a strong prior on no *measured* confounding - a prior belief that 1) we rarely hold and 2) leads to bad frequentist performance https://t.co/GzN16Ze52w

Hot off the press from our @NIHAging osteoporosis R01: we created a multinational database of older adults in the US and Canada and found extreme shifts in which osteoporosis medications were used over the past decades. Notice anything interesting? https://t.co/er6V60knDg

Slides for this talk are posted on my site! https://t.co/UDxlagT03B

I’ll be talking about this paper on Bayesian causal inference w/ recurrent events @ #ENAR2025. Come check out our session! Session 50. It’s All Aabout the Estimand: Asking the Right Questions to Best Inform Health Policy Decisions March 24: 1:45-3:30pm; at Oakley, floor 4