Unmeasured Spatial Confounding

on YouTube ↗27m

" summary

Proposes that unmeasured confounders with spatial structure are identifiable and correctable by exploiting spatial correlation, using a distance-adjusted propensity score that combines propensity distance and geographic proximity (tuned by weight W) and an affine bias correction that models (U,Z) as multivariate normal with a cross-Markov property on the precision matrix. Implements DAPS for ATT and a weighted‑least‑squares plus bias‑correction estimator, proves identifiability on ring graphs, and shows both methods flip implausible naive results (ozone and grocery–CVD analyses) toward subject‑matter plausible null or weaker effects.

AI-written summary · about

" chapters (16)

1. 0:00Intro
2. 1:56Spatial confounding in spatial statistics
3. 3:44Unmezoured spatial confounding in causal inference
4. 5:47Spatial confounding: What does it mean?
5. 7:01Question 1: Spatial causal inference in air pollution research
6. 8:31Question 2: Supermarket access and cardiovascular health
7. 9:22Distance Adjusted Propensity Score Matching
8. 11:05Choosing
9. 11:36Balance and distance
10. 14:29Potential outcomes
11. 15:50Blas of common estimators, and the affine estimator
12. 17:58A Gaussian Markov random field construction of the joint distribution
13. 19:31Calculating the affine estimator
14. 21:08Learning the spatial parameters
15. 24:10Effect of poor supermarket access on CVD deaths
16. 25:22Conclusions

" related talks

• 01h 02mUnmeasured Confounding and More Recent Developments/Challenges in Causal DiscoveryDaniel Malinsky · Causality Boot Camp
• 31mSome Attention for Attenuation BiasRuben Mak · PyData Eindhoven 2021
• 52mCausal Matrix EstimationDevavrat Shah · Algorithmic Advances for Statistical Inference with Combinatorial Structure
• 27mDreadful Frailties in Propensity Score Matching and How to Fix ThemAlexandre Abraham · PyData Paris 2024
• 45mMulticalibration, Universal Adaptability and CausalityOmer Reingold · Algorithmic Aspects of Causal Inference 2022
• 09mTowards Credible and Effective Data-Driven Decision-MakingAngela Zhou · Meet the Fellows Welcome Event
• 11mCausal Inference With Corrupted Data: Measurement Error, Missing Values, Discretization, And...Rahul Singh · Meet the Fellows Welcome Event
• 49mCausal Inference in Complex Systems: Network Interference, Strategic Agents, and BeyondPanos Toulis · Quantifying Uncertainty: Stochastic, Adversarial, and Beyond