Online Program Home
My Program

Abstract Details

Activity Number: 579 - Sampling, Variance Estimation, and Advancements with Auxiliary Data
Type: Contributed
Date/Time: Wednesday, July 31, 2019 : 2:00 PM to 3:50 PM
Sponsor: Survey Research Methods Section
Abstract #305299 Presentation
Title: Benefit of Probability-Proportional-To-Size Sampling in Cluster Randomized Experiments
Author(s): Yeng Xiong* and Michael Higgins
Companies: and Kansas State University
Keywords: Probability proportional to size sampling; cluster randomized experiment; Horvitz-Thompson; Neyman Rubin ; causal inference

Cluster randomized experiments (CREs) have three defining features: (i) treatments are randomized to clusters of units rather than units themselves, (ii) clusters are formed a priori to experimentation and without researcher intervention, and (iii) the research objective and analysis are still centered on units. Conventionally, clusters in CREs are sampled using simple random sampling, but the inherent disparity between the experimental and observational units in CREs can lead to some analytical and design challenges. Stratifying or blocking on clusters based on important covariates can improve precision on estimation. However, we propose utilizing a different sampling scheme: sampling with probability proportional to size without replacement. Under the Neyman-Rubin potential outcomes framework, this modification leads to an unbiased Horvitz-Thompson estimator of the population average treatment effect, which we called HT-PPS, that can accommodate the clustering structure in CREs without having to compromise on desirable statistical properties. We then discuss how our estimator and sampling scheme can be incorporated into best practices for designing CREs.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2019 program