Genuinely Robust Inference for Clustered Data

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Chiang, Harold D., Sasaki, Yuya, Wang, Yulong
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914073233326080
author Chiang, Harold D.
Sasaki, Yuya
Wang, Yulong
author_facet Chiang, Harold D.
Sasaki, Yuya
Wang, Yulong
contents Conventional cluster-robust inference can be invalid when data contain clusters of unignorably large size. We formalize this issue by deriving a necessary and sufficient condition for its validity, and show that this condition is frequently violated in practice: specifications from 77% of empirical research articles in American Economic Review and Econometrica during 2020-2021 appear not to meet it. To address this limitation, we propose a genuinely robust inference procedure based on a new cluster score bootstrap. We establish its validity and size control across broad classes of data-generating processes where conventional methods break down. Simulation studies corroborate our theoretical findings, and empirical applications illustrate that employing the proposed method can substantially alter conventional statistical conclusions.
format Preprint
id arxiv_https___arxiv_org_abs_2308_10138
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Genuinely Robust Inference for Clustered Data
Chiang, Harold D.
Sasaki, Yuya
Wang, Yulong
Econometrics
Conventional cluster-robust inference can be invalid when data contain clusters of unignorably large size. We formalize this issue by deriving a necessary and sufficient condition for its validity, and show that this condition is frequently violated in practice: specifications from 77% of empirical research articles in American Economic Review and Econometrica during 2020-2021 appear not to meet it. To address this limitation, we propose a genuinely robust inference procedure based on a new cluster score bootstrap. We establish its validity and size control across broad classes of data-generating processes where conventional methods break down. Simulation studies corroborate our theoretical findings, and empirical applications illustrate that employing the proposed method can substantially alter conventional statistical conclusions.
title Genuinely Robust Inference for Clustered Data
topic Econometrics
url https://arxiv.org/abs/2308.10138