Panel Discussion
summary
Panel compares concrete problems and pragmatic answers: spatial confounding needs assumptions about smoothness or tail behavior (non‑Gaussianity might help), mediation is usefully framed as separable treatment components (a four‑arm target trial is a clear estimand), and multi‑site experiments often merit the finite-sample person-weighted estimate rather than optimistic superpopulation generalization. Also discussed: whether to adjust for cluster size, when site covariates can help, and checking model fit and estimator stability across alternative estimands.
AI-written summary · about
chapters (8)
- 0:00Intro
- 1:25Can you hear me
- 8:30More specific
- 13:50Space
- 16:10Spatial structure
- 17:10Multisite experiment
- 19:30Reality check
- 20:10Discussion
related talks
01h 06mPanel on Societal Dimensions of Explanation
01h 11mStatistical Limits of Causal Inference
49mCausal Inference in Complex Systems: Network Interference, Strategic Agents, and Beyond
30mTalk by Suhas Diggavi (UCLA)
41mQ&A with
18mConstantine
57mPanel on Interpretability in the Physical Sciences
13mAlgorithmic Advances For The Design And Analysis Of Randomized Control Trials