A Weighting Framework for Clusters as Confounders in Observational Studies
Eli Ben-Michael, Avi Feller, Luke Keele · 2026
Read A Weighting Framework for Clusters as Confounders in Observational Studies by Eli Ben-Michael, Avi Feller, Luke Keele (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-05041.txt.
Description
Studies that watch people often group them: pupils inside schools, patients inside hospitals. Eli Ben-Michael and 2 co-authors built a single weighting scheme for that setup in this 2026 statistics paper. Grouping creates two kinds of skew to fix, gaps across groups and gaps inside them, and common methods quietly fix only one.
The usual approach, weighting by a random-effects propensity score, aims only at the across-group balance. The authors offer two fixes. Hierarchical balancing weights control both kinds of balance through one optimization task. A new Mundlak-style estimator adjusts for group summaries and copes with tiny groups where all share the same treatment state. The approaches rest on varied assumptions, spelled out in the paper. Simulations plus uses in education and health services research compare them.
Catalog record
| CREATOR | Eli Ben-Michael, Avi Feller, Luke Keele |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 86 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-05041