Multiple Randomization Designs: Estimation and Inference with Interference

2021/12/27 by Masoero, Lorenzo, Vijaykumar, Suhas, Richardson, Thomas +5 · 9 citations
#62B15 (Primary) 91B82 #91B26 #91B80 #91C20 #91C20 (Secondary) #Econometrics (econ.EM) #FOS: Computer and information sciences #FOS: Economics and business #FOS: Mathematics #G.3 #I.2.6 #J.4 #Methodology (stat.ME) #Social and Information Networks (cs.SI) #Statistics Theory (math.ST)

paper · doi:10.48550/arxiv.2112.13495

Abstract

Completely randomized experiments, originally developed by Fisher and Neyman in the 1930s, are still widely used in practice, even in online experimentation. However, such designs are of limited value for answering standard questions in marketplaces, where multiple populations of agents interact strategically, leading to complex patterns of spillover effects. In this paper, we derive the finite-sample properties of tractable estimators for "Simple Multiple Randomization Designs" (SMRDs), a new class of experimental designs which account for complex spillover effects in randomized experiments. Our derivations are obtained under a natural and general form of cross-unit interference, which we call "local interference". We discuss the estimation of main effects, direct effects, and spillovers, and present associated central limit theorems.

Cited by

• Design-based theory for causal inference
• Unbiased Platform-Level Causal Estimation for Search Systems: A Competitive Isolation PSM-DID Framework
• A Two-armed Bandit Framework for A/B Testing
• Choosing the Better Bandit Algorithm under Data Sharing: When Do A/B Experiments Work?
• SERP Interference Network and Its Applications in Search Advertising
• Robust and efficient multiple-unit switchback experimentation
• Balancing Interference and Correlation in Spatial Experimental Designs: A Causal Graph Cut Approach
• The Price of Isolation: Estimating the Ecosystem Cost of Symmetric Two-Sided A/B Testing
• Randomization Inference in Two-Sided Market Experiments

Related

• Multiple Randomization Designs: Estimation and Inference with Interference
• Policy design in experiments with unknown interference
• Randomization Test for the Specification of Interference Structure
• Agnostic Characterization of Interference in Randomized Experiments
• Experimental Design under Network Interference