A Bayesian machine learning approach to direct and spillover effects in cluster-randomized trials with noncompliance
Abstract
We develop a Bayesian machine learning approach for estimating direct and spillover effects in cluster-randomized trials with treatment noncompliance. To flexibly model outcomes and treatment receipt while accounting for within-cluster dependence, we combine Bayesian additive regression trees with cluster-level random effects. We augment the outcome regression with a peer-treatment clever covariate, defined as the predicted mean proportion of treated peers from a preliminary receipt model, to mitigate regularization-induced distortion of the effect decomposition. Within a Bayesian g-computation framework, we estimate cluster-average intent-to-treat effects and cluster-average local average treatment effects and examine sensitivity to the cross-world restrictions underlying their decomposition. Simulation studies show reduced bias under model misspecification, while the correctly specified parametric models show expected frequentist properties. We find that the clever covariate improves the decomposition of the total effect into direct and spillover components under moderate or strong spillover, with similar gains in smaller clusters. Its benefits vary with spillover strength, and the 95% credible intervals remain conservative. An application to the Primary School Deworming Project in Kenya suggests reduced infection risk, with a larger estimated contribution from individual treatment receipt than from peer receipt. Overall, our approach provides a flexible Bayesian estimation of direct and spillover effect decomposition in the presence of treatment noncompliance.
Related
- Bayesian Neural Networks via Dirichlet Process Priors on the Data-Generating Distribution for Randomized Value Functions in Deep Reinforcement Learning
- A PAC-Bayesian View of Generalisation for Physics-Informed Machine Learning
- Dynamic Spatial Bayesian Machine Learning Model: Applications to Intergenerational Economic Mobility and Geographic Income Inequality in the United States
- Causal sensitivity analyses for indirect and spillover mediation effects in cluster-randomized trials: Application to a pragmatic trial on chronic pain management
- Calibrated Bayesian inference for overall and subgroup-specific causal effects in stepped-wedge cluster-randomized trials
BibTeX
@misc{bhandari2026bayesian,
title = {{A Bayesian machine learning approach to direct and spillover effects in cluster-randomized trials with noncompliance}},
author = {Saurabh Bhandari and Fan Li and Michael J. Daniels},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.27425},
eprint = {2610.27425},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.27425},
}
RIS
TY - UNPB
TI - A Bayesian machine learning approach to direct and spillover effects in cluster-randomized trials with noncompliance
AU - Saurabh Bhandari
AU - Fan Li
AU - Michael J. Daniels
PY - 2026
DA - 2026/10/20
PB - arXiv
DO - 10.48550/arxiv.2610.27425
UR - https://arxiv.org/abs/2610.27425
AB - We develop a Bayesian machine learning approach for estimating direct and spillover effects in cluster-randomized trials with treatment noncompliance. To flexibly model outcomes and treatment receipt while accounting for within-cluster dependence, we combine Bayesian additive regression trees with cluster-level random effects. We augment the outcome regression with a peer-treatment clever covariate, defined as the predicted mean proportion of treated peers from a preliminary receipt model, to mitigate regularization-induced distortion of the effect decomposition. Within a Bayesian g-computation framework, we estimate cluster-average intent-to-treat effects and cluster-average local average treatment effects and examine sensitivity to the cross-world restrictions underlying their decomposition. Simulation studies show reduced bias under model misspecification, while the correctly specified parametric models show expected frequentist properties. We find that the clever covariate improves the decomposition of the total effect into direct and spillover components under moderate or strong spillover, with similar gains in smaller clusters. Its benefits vary with spillover strength, and the 95% credible intervals remain conservative. An application to the Primary School Deworming Project in Kenya suggests reduced infection risk, with a larger estimated contribution from individual treatment receipt than from peer receipt. Overall, our approach provides a flexible Bayesian estimation of direct and spillover effect decomposition in the presence of treatment noncompliance.
ER -