The bias-corrected matching estimator has the same form as the doubly robust estimator; see proposition 15.2 of arxiv.org/pdf/2305.18793… Zhexiao and Fang made the argument rigorous! Zhexiao Lin Fang Han

Open rank teaching professor position in EECS/Data Science at UC Berkeley CDSS! Our teaching professors are in the Academic Senate, can get tenure equivalent (Security of Employment in our HRese), and get to shape a world-leading data science program. aprecruit.berkeley.edu/JPF04169

Congratulations to Denis Kojevnikov who received the 2023 Journal of Econometrics Zellner Award, together with @vadimmarmer & Kyungchul Song, for the best theory paper in the journal: Limit theorems for network dependent random variables 👏🏽👏🏽 journals.elsevier.com/journal-of-eco… Aureo de Paula

I just posted my notes for Stat 230 ``Linear Models'' to ArXiv: arxiv.org/pdf/2401.00649… It covers the linear model and many extensions. I will teach it again in the spring and continue polishing the notes. Comments are welcome.

How should we analyze experiments that randomly form groups of people? A new paper by Basse, @pengdingpku, Avi Feller, and Panos Toulis proposes a simple-to-implement, randomization-based test that is exact in finite samples.📊 econometricsociety.org/publications/e…

My co-author Rebecca Barter and I are thrilled to announce the online release of our MIT Press book "Veridical Data Science: The Practice of Responsible Data Analysis and Decision Making" (vdsbook.com), an essential source for producing trustworthy data-driven results.

``control'' means many different things in statistics, e.g. treatment-control experiment; control for confounding; case-control study; negative control; controlled direct effect; control function. ``control'' can even mean covariates (good or bad controls).

Hope we provide some new insights into the old problem of missing data in RCTs. Fan Li

Fan Li's slides for causal inference

Just gave a guest lecture on Bayesian Causal Inference at Williams College, with slides and R code at doi.org/10.7910/DVN/JO… which is an introduction to our review paper royalsocietypublishing.org/doi/10.1098/rs… Fabrizia Mealli Fan Li (I never taught any Bayesian Statistics at Berkeley.)

In ASA JASA, Anqi Zhao & Peng Ding consider alternative strategies to address covariate missingness in randomized experiments and recommend including missingness indicators when estimating average treatment effects. tandfonline.com/doi/abs/10.108…

This is an interesting and useful trick. However, centering factors has some special restrictions on the estimated factorial effects when there are more than 3 factors (3 is the magic number there!). This motivates us to write this paper: academic.oup.com/biomet/article…

IPW with the estimated propensity score is another example. The first-stage estimation reduces the asymptotic variance, which surprises many people. A recent paper is arxiv.org/pdf/2303.17102 Also, Newey&McFadden chapter 6 is about "two-step estimation" sciencedirect.com/science/articl…

🧑‍🏫 UC Berkeley seeks applicants for 4 tenure-track and 1 tenured professor position in "AI, Inequality, and Society." Explore areas like employment, algorithmic bias, and more. Salary: $78,200-$310,300. aprecruit.berkeley.edu/JPF04498

We view DID as a factorial design with the panel data structure. It is a fun paper. Comments are welcome.

101 years ago, Neyman introduced potential outcomes and design-based inference. For a special issue of Journal of Causal Inference, Sandrine Dudoit (Sandrine Dudoit), Deb Nolan (Deborah Nolan), Terry Speed, and I wrote about how 4 books from Berkeley Statistics & Thad Dunning explain 1/

excited to see the published version of our paper on dealing with missing covariates and outcomes in randomized trials: academic.oup.com/biomet/article… It is a simple paper with some intriguing results. Slides are here: dropbox.com/scl/fi/jrkfnj0… Fan Li

very excited to see the paper "Nonparametric identification is not enough, but randomized controlled trials are" with comments from Ben Recht ("A Bureaucratic Theory of Statistics") and myself ("What randomization can and cannot guarantee"): muse.jhu.edu/issue/54591

I just sent the new version of the textbook to CRC: arxiv.org/pdf/2401.00649… with R code and data at: dataverse.harvard.edu/dataset.xhtml?… I can still make minor changes. Comments are welcome.