A Matching Procedure for Sequential Experiments that Iteratively Learns\n which Covariates Improve Power

2020/10/12 by Adam Kapelner, Abba Μ. Krieger, Kapelner, Adam +1 · 1 citation
Mathematics · #Statistical Methods in Clinical Trials #Advanced Causal Inference Techniques #Statistical Methods and Inference

paper · pdf · doi:10.48550/arxiv.2010.05980

Abstract

We propose a dynamic allocation procedure that increases power and efficiency\nwhen measuring an average treatment effect in sequential randomized trials\nexploiting some subjects' previous assessed responses. Subjects arrive\nsequentially and are either randomized or paired to a previously randomized\nsubject and administered the alternate treatment. The pairing is made via a\ndynamic matching criterion that iteratively learns which specific covariates\nare important to the response. We develop estimators for the average treatment\neffect as well as an exact test. We illustrate our method's increase in\nefficiency and power over other allocation procedures in both simulated\nscenarios and a clinical trial dataset. An R package "SeqExpMatch" for use by\npractitioners is available.\n

Cited by

• Reservoir Designs for Online Paired Experiments

Related

• Matching on-the-fly in Sequential Experiments for Higher Power and Efficiency
• Pairwise Sequential Randomization and Its Properties
• Rematching on-the-fly: sequential matched randomization and a case for covariate-adjusted randomization
• Adjusting SPRT for an Efficient Procedure with Finite Number of Applications of Less Effective Treatment
• Covariate Adjustment in Experiments with Matched Pairs