Statistician | Assistant professor at @BrownBiostats | Nonparametric Bayesian methods for causal inference. https://xitter.cf/t.co/tdV1V2KpHg
Providence, RI
Joined March 2011
- Tweets1.7K
- Following863
- Followers1.8K
- Likes3.1K
Pinned Tweet
I'm looking forward to teaching a 4-hour short course on Bayesian sensitivity analysis methods at the American Causal Inference Conference (ACIC) 2026!
It will have a distinct focus on practical implementation in @mcmc_stan w/ examples.
Register here: sci-info.org/annual-meeting/…
What makes for productive LLM use in research?
Imo: an ability to think in an organized & modular way about getting to a working solution.
One-shotting tasks usually gives me a complete mess. Modular usage keeps things tight and understandable + keeps my brain engaged.
I’m pretty good now at intuiting where I need to go and breaking it down into sub-tasks. So LLMs are great for me!
But I got there by doing things manually, messing up, thinking + reading, then trying again.
New paper in press at Biometrics by PhD Candidate Esteban Fernández-Morales
1) Develops Bayesian spike & slab and horseshoe models for causal inference under spatial spillover
2) Analyzes Philly's 2017 beverage tax accounting for cross-border shopping
arxiv.org/pdf/2501.08231
With rapid AI advancement, I've been thinking more about this conversation 13 yrs ago, btwn Jack White & Conan O'Brien, and how it's as relevant for statistics/data science education now as it is for music, comedy, film, etc.
youtu.be/AJgY9FtDLbs?si=kaR2…
Excerpts below start @ 19:13
Jack: I'm a believer that new technology needs new responsibility. If you're a photographer you have photoshop now. And if you're going to call yourself a photographer and dedicate your life to it...
New paper up on arXiv.
“Stress-Testing Assumptions: A Guide to Bayesian Sensitivity Analyses in Causal Inference”
Examples include: exposure misclassification, unmeasured confounding, snd MNAR outcomes.
W/ @mcmc_stan code on GitHub.
arxiv.org/abs/2602.23640
Just got this ad while scrolling BBC 😂
I suppose this is partially a result of using ML models w.o. good causal/statistical reasoning.
Anyway I guess apply to the online @BrownBiostats ScM program to learn how to not do this!
New paper on causal inference with irregularly (and informatively) timed treatment decisions.
The paper has publicly available code and via generosity of @PCORI is open-access:
doi.org/10.1097/EDE.00000000…
Arman Oganisian retweeted
Our new paper is out: Bayesian nonparametric causal inference for high-dimensional nutritional data using factor-based exposure mapping. arxiv.org/abs/2601.16595
Joint work with my postdoc Dafne Zorzetto and collaborators @StableMarkets @BrownBiostats
Check out this paper on Bayesian nonparametric causal inference with high-dimensional exposure mappings!
Our new paper is out: Bayesian nonparametric causal inference for high-dimensional nutritional data using factor-based exposure mapping. arxiv.org/abs/2601.16595
Joint work with my postdoc Dafne Zorzetto and collaborators @StableMarkets @BrownBiostats
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
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}
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🤞
ascopubs.org/doi/10.1200/GO-…
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
arxiv.org/abs/2508.21804
... and how identifiability conditions may be read off a Single World Intervention Graph (SWIG) template for the implicit DTR.
