A Weighting Framework for Clusters as Confounders in Observational Studies

2026-02-04

AI Summary

This study addresses confounding bias in observational studies with clustered units—such as students nested within schools or patients within hospitals—where existing methods struggle to simultaneously balance covariates across clusters (global imbalance) and within clusters (local imbalance). The authors propose a unified weighting framework that explicitly distinguishes and jointly optimizes these two types of balance: first, hierarchical balancing weights derived via constrained optimization to achieve dual-level balance; second, Mundlak-type balancing weights that replace cluster indicators with cluster-level sufficient statistics, making them suitable for small clusters with all-treated or all-control units. Theoretical analysis clarifies the identification assumptions required for each approach. Simulations and empirical applications in education and healthcare demonstrate that the proposed method substantially outperforms conventional inverse probability weighting across diverse clustering structures, yielding markedly less biased causal effect estimates.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Web Mining and Content Analysis: Normalization, clustering, classification, and summarization of Web text
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

When units in observational studies are clustered in groups, such as students in schools or patients in hospitals, researchers often address confounding by adjusting for cluster-level covariates or cluster membership. In this paper, we develop a unified weighting framework that clarifies how different estimation methods control two distinct sources of imbalance: global balance (differences between treated and control units across clusters) and local balance (differences within clusters). We show that inverse propensity score weighting (IPW) with a random effects propensity score model -- the current standard in the literature -- targets only global balance and constant level shifts across clusters, but imposes no constraints on local balance. We then present two approaches that target both forms of balance. First, hierarchical balancing weights directly control global and local balance through a constrained optimization problem. Second, building on the recently proposed Generalized Mundlak approach, we develop a novel Mundlak balancing weights estimator that adjusts for cluster-level sufficient statistics rather than cluster indicators; this approach can accommodate small clusters where all units are treated or untreated. Critically, these approaches rest on different assumptions: hierarchical balancing weights require only that treatment is ignorable given covariates and cluster membership, while Mundlak methods additionally require an exponential family structure. We then compare these methods in a simulation study and in two applications in education and health services research that exhibit very different cluster structures.

Problem

• confounding
• clustering
• observational studies
• balance
• weighting

Innovation

• hierarchical balancing weights
• Mundlak balancing weights
• clustered observational studies
• global and local balance
• propensity score weighting