Pan Zhao

“All Our Default Models Are Wrong: Causal inference for varying treatment effects”: my talk this Saturday morning in Ottawa
statmodeling.stat.columbia.edu/2025/10/16/a...

For the 3rd day of my #MEstimatorMonday catch-up week (39/52), let's talk proximal causal inference

We can think about proximal causal inference as an extension of the standard identification assumptions to allow for more rich data structures. Specifically, we can account for unmeasured confounding

Jane Goodall, world-renowned primatologist, dies aged 91 Jane Goodall Institute says ‘tireless advocate’ for natural world died in California during US speaking tour The world-renowned primatologist Jane Goodall has died at the age of 91, her institute has said. The Jane Goodall Institute announced that she had passed away of natural causes while in California as part of a US speaking tour. Continue reading...

Jane Goodall, world-renowned primatologist, dies aged 91

link 📈🤖
A note on the relation between one--step, outcome regression and IPW--type estimators of parameters with the mixed bias property (Rotnitzky, Smucler, Robins) Bruns-Smith et al. (2025) established an algebraic identity between the one-step estimator and a specific outcome regression-type e

“Veridical (truthful) Data Science”: Another way of looking at statistical workflow
statmodeling.stat.columbia.edu/2025/09/28/v...

Does AI help humans make better decisions? A statistical evaluation framework for experimental and observational studies | PNAS The use of AI, or more generally data-driven algorithms, has become ubiquitous in today’s society. Yet, in many cases and especially when stakes ar...

Excited to see our paper on evaluating whether AI can help humans make better decisions is now out in @pnas.org!
www.pnas.org/doi/10.1073/...

Herbert P. Susmann, Alec McClean, Iv\'an D\'iaz
Non-overlap Average Treatment Effect Bounds
https://arxiv.org/abs/2509.20206

Causal Machine Learning Methods and Use of Cross‐Fitting in Settings With High‐Dimensional Confounding Observational epidemiological studies commonly seek to estimate the causal effect of an exposure on an outcome. Adjustment for potential confounding bias in modern studies is challenging due to the p...

NEW PAPER!!! "Causal Machine Learning Methods and Use of Cross‐Fitting in Settings With High‐Dimensional Confounding"

led by Susie Ellul, with Stijn Vansteelandt & John Carlin
Published in Stats in Med

Check it out 👇
onlinelibrary.wiley.com/doi/10.1002/...

#EpiSky #CausalSky

Title page for paper: DoubleGen: Debiased Generative Modeling of Counterfactuals arXiv:2509.16842 (stat) Alex Luedtke, Kenji Fukumizu

Selected attributes that are more common in smiling (n = 78 080) than in non-smiling (n = 84 690) CelebA faces. If a model is trained only on the smiling subset, it tends to over-produce these attributes instead of showing how the full population would look if everyone smiled. Table: Lipstick Makeup Female* Earrings No-beard Blonde Smiling 56 % 47 % 65 % 26 % 88 % 18 % Not smiling 38 % 30 % 52 % 12 % 79 % 12 % Overall 47 % 38 % 58 % 19 % 83 % 15 %

Counterfactual smiling celebrities generated by a traditional diffusion model trained on only smiling faces (top) and a DoubleGen diffusion model (bottom). Columns contain coupled samples, with the random seed set to the same value before generation. The stars mark the most qualitatively different pairs. What’s visible: two horizontal rows, each showing twelve AI-generated smiling portraits. Starred columns highlight the biggest shifts: in those pairs, DoubleGen produces faces with traits under-represented among smiling faces in the original data. Non-starred columns look nearly identical between the two rows.

New paper on generative modeling of counterfactual distributions! We give a way to answer "what if" questions with generative models.

For example: what would faces look like if they were all smiling?

arxiv.org/abs/2509.16842

link 📈🤖
DoubleGen: Debiased Generative Modeling of Counterfactuals (Luedtke, Fukumizu) Generative models for counterfactual outcomes face two key sources of bias. Confounding bias arises when approaches fail to account for systematic differences between those who receive the intervention and thos

link 📈🤖
Variable Selection for Additive Global Fr\'echet Regression (Yang, Bhattacharjee, Xue et al) We present a novel framework for variable selection in Fr\'echet regression with responses in general metric spaces, a setting increasingly relevant for analyzing non-Euclidean data such as probab

Han Cui, Xinran Li
Robust Sensitivity Analysis via Augmented Percentile Bootstrap under Simultaneous Violations of Unconfoundedness and Overlap
https://arxiv.org/abs/2509.13169

Georgi Baklicharov, Kelly Van Lancker, Stijn Vansteelandt
Weakening assumptions in the evaluation of treatment effects in longitudinal randomized trials with truncation by death or other intercurrent events
https://arxiv.org/abs/2509.10067

Join #ENAR_ibs Friday, October 3 at 12 PM, when we host the #WebENAR "Sequential Causal Inference in Experimental or Observational Settings." More details at www.enar.org/education/in...
#causalinference #Biostatistics #statistics

https://hsph.harvard.edu/research/causalab/seminars/

The CAUSALab Methods Series @ki.se is back!

Fall 2025 lineup kicks off Sep 23 with Vanessa Didelez (BIPS), “Causal mediation and separable treatments in time-to-event analyses.”

All talks are virtual, except for Nov. 4, 2025 hybrid session.

Learn more & register:
hsph.harvard.edu/research/cau...

link 📈🤖
Comment on "Deep Regression Learning with Optimal Loss Function" (Li) OpenReview benefits the peer-review system by promoting transparency, openness, and collaboration. By making reviews, comments, and author responses publicly accessible, the platform encourages constructive feedback, re

link 📈🤖
Confounder selection via iterative graph expansion (Guo, Zhao) Confounder selection, namely choosing a set of covariates to control for confounding between a treatment and an outcome, is arguably the most important step in the design of an observational study. Previous methods, such as Pe

link 📈🤖
Evaluation of Surrogate Endpoints Based on Meta-Analysis with Surrogate Indices (Stijven, Gilbert) We introduce in this paper an extension of the meta-analytic (MA) framework for evaluating surrogate endpoints. While the MA framework is regarded as the gold standard for surrogate endpoint

An announcement, which might be of some interest:

In the period 2022-2024, myself and a number of other postdocs on the "CoSInES" and "Bayes4Health" EPSRC grants were involved in organising a number of internal tutorial workshops, on topics relevant to researchers in computational statistics.

Susan Athey, Guido Imbens, Zhaonan Qu, Davide Viviano
Triply Robust Panel Estimators
https://arxiv.org/abs/2508.21536

虎兕出于柙，龟玉毁于椟中，是谁之过与？

Carlos Cinelli, Avi Feller, Guido Imbens, Edward Kennedy, Sara Magliacane, Jose Zubizarreta
Challenges in Statistics: A Dozen Challenges in Causality and Causal Inference
https://arxiv.org/abs/2508.17099

link 📈🤖
Post-selection inference with a single realization of a network (Ancell, Witten, Kessler) Given a dataset consisting of a single realization of a network, we consider conducting inference on a parameter selected from the data. In particular, we focus on the setting where the parameter of

An abbreviation (ABB) in a journal article (JA) or Grant Application (GA) is rarely worth the words it saves. Every ABB requires cognitive resources (CR) and at my age by the time I'm halfway through a JA or GA I no longer have the CR to remember what your ABB stood for.

link 📈🤖
Efficient Statistical Estimation for Sequential Adaptive Experiments with Implications for Adaptive Designs (Zhang, Laan) Adaptive experimental designs have gained popularity in clinical trials and online experiments. Unlike traditional, fixed experimental designs, adaptive designs can dy

?

Physicist & Nobel Laureate Paul Dirac FRS was born #OnThisDay in 1902. He is known for his exploration of quantum mechanics & predicting antimatter. He was also known for his quiet nature, with his colleagues jokingly defining a unit called a 'dirac' which was one word per hour.

#ScienceHistory

Recently accepted to #REStud, "Behavioral Causal Inference," from Ran Spiegler:

www.restud.com/behavioral-c...

#econsky

Matias D. Cattaneo, Rocio Titiunik
The Regression Discontinuity Design in Medical Science
https://arxiv.org/abs/2508.03878

Linying Yang, Robin J. Evans, Xinwei Shen
Frugal, Flexible, Faithful: Causal Data Simulation via Frengression
https://arxiv.org/abs/2508.01018