A More Efficient, Doubly Robust, Nonparametric Estimator of Treatment Effects in Multilevel Studies

TL;DR

This paper introduces a new nonparametric estimator for treatment effects in multilevel studies that is doubly robust, more efficient, and does not rely on parametric assumptions, leveraging machine learning for nuisance function estimation.

Contribution

The paper presents a novel, doubly robust, nonparametric estimator for treatment effects in multilevel data that improves efficiency without parametric assumptions.

Findings

Estimator is asymptotically Normal and more efficient than existing methods.

Performs well in simulated and real data, especially for cluster-specific effects.

Uses machine learning to estimate nuisance functions without parametric models.

Abstract

When studying treatment effects in multilevel studies, investigators commonly use (semi-)parametric estimators, which make strong parametric assumptions about the outcome, the treatment, and/or the correlation structure between study units in a cluster. We propose a novel estimator of treatment effects that does not make such assumptions. Specifically, the new estimator is shown to be doubly robust, asymptotically Normal, and often more efficient than existing estimators, all without having to make any parametric modeling assumptions about the outcome, the treatment, and the correlation structure. We achieve this by estimating two non-standard nuisance functions in causal inference, the conditional propensity score and the outcome covariance model, using existing existing machine learning methods designed for independent and identically distributed (i.i.d) data. The new estimator is also demonstrated in simulated and real data where the new estimator is drastically more efficient than existing estimators, especially when studying cluster-specific treatment effects.