Multiple Randomization Designs: Estimation and Inference with Interference
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2021
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914175035375616 |
|---|---|
| author | Masoero, Lorenzo Vijaykumar, Suhas Richardson, Thomas McQueen, James Rosen, Ido Burdick, Brian Bajari, Pat Imbens, Guido |
| author_facet | Masoero, Lorenzo Vijaykumar, Suhas Richardson, Thomas McQueen, James Rosen, Ido Burdick, Brian Bajari, Pat Imbens, Guido |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2112_13495 |
| institution | arXiv |
| publishDate | 2021 |
| record_format | arxiv |
| spellingShingle | Multiple Randomization Designs: Estimation and Inference with Interference Masoero, Lorenzo Vijaykumar, Suhas Richardson, Thomas McQueen, James Rosen, Ido Burdick, Brian Bajari, Pat Imbens, Guido Methodology Social and Information Networks Econometrics Statistics Theory 62B15 (Primary) 91B82, 91B26, 91C20, 91B80, 91C20 (Secondary) J.4; G.3; I.2.6 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. |
| title | Multiple Randomization Designs: Estimation and Inference with Interference |
| topic | Methodology Social and Information Networks Econometrics Statistics Theory 62B15 (Primary) 91B82, 91B26, 91C20, 91B80, 91C20 (Secondary) J.4; G.3; I.2.6 |
| url | https://arxiv.org/abs/2112.13495 |