A bias-adjusted estimator in quantile regression for clustered data

Research output: Contribution to journal › Journal article › Research › peer-review

Standard

A bias-adjusted estimator in quantile regression for clustered data. / Battagliola, Maria Laura; Sørensen, Helle; Tolver, Anders; Staicu, Ana Maria.

In: Econometrics and Statistics, Vol. 23, 2022, p. 165-186.

Research output: Contribution to journal › Journal article › Research › peer-review

Harvard

Battagliola, ML, Sørensen, H, Tolver, A & Staicu, AM 2022, 'A bias-adjusted estimator in quantile regression for clustered data', Econometrics and Statistics, vol. 23, pp. 165-186. https://doi.org/10.1016/j.ecosta.2021.07.003

APA

Battagliola, M. L., Sørensen, H., Tolver, A., & Staicu, A. M. (2022). A bias-adjusted estimator in quantile regression for clustered data. Econometrics and Statistics, 23, 165-186. https://doi.org/10.1016/j.ecosta.2021.07.003

Vancouver

Battagliola ML, Sørensen H, Tolver A, Staicu AM. A bias-adjusted estimator in quantile regression for clustered data. Econometrics and Statistics. 2022;23:165-186. https://doi.org/10.1016/j.ecosta.2021.07.003

Author

Battagliola, Maria Laura ; Sørensen, Helle ; Tolver, Anders ; Staicu, Ana Maria. / A bias-adjusted estimator in quantile regression for clustered data. In: Econometrics and Statistics. 2022 ; Vol. 23. pp. 165-186.

Bibtex

@article{5c7586c2eafd42edbd417860d8d96335,

title = "A bias-adjusted estimator in quantile regression for clustered data",

abstract = "Quantile regression models with random effects are useful for studying associations between covariates and quantiles of the response distribution for clustered data. Parameter estimation is examined for a class of mixed-effects quantile regression models, with focus on settings with many but small clusters. The main contributions are the following: (i) documenting that existing methods may lead to severely biased estimators for fixed effects parameters; (ii) proposing a new two-step estimation methodology where predictions of the random effects are first computed by a pseudo likelihood approach (the LQMM method) and then used as offsets in standard quantile regression; (iii) proposing a novel bootstrap sampling procedure in order to reduce bias of the two-step estimator and compute confidence intervals. The proposed estimation and associated inference is assessed numerically through rigorous simulation studies and applied to an AIDS Clinical Trial Group (ACTG) study.",

keywords = "AIDS clinical trial group study, Bias-adjustment, Clustered data, Linear quantile regression, Random effects, Wild bootstrap",

author = "Battagliola, {Maria Laura} and Helle S{\o}rensen and Anders Tolver and Staicu, {Ana Maria}",

note = "Publisher Copyright: {\textcopyright} 2021 EcoSta Econometrics and Statistics",

year = "2022",

doi = "10.1016/j.ecosta.2021.07.003",

language = "English",

volume = "23",

pages = "165--186",

journal = "Econometrics and Statistics",

issn = "2452-3062",

publisher = "Elsevier",

}

RIS

TY - JOUR

T1 - A bias-adjusted estimator in quantile regression for clustered data

AU - Battagliola, Maria Laura

AU - Sørensen, Helle

AU - Tolver, Anders

AU - Staicu, Ana Maria

PY - 2022

Y1 - 2022

N2 - Quantile regression models with random effects are useful for studying associations between covariates and quantiles of the response distribution for clustered data. Parameter estimation is examined for a class of mixed-effects quantile regression models, with focus on settings with many but small clusters. The main contributions are the following: (i) documenting that existing methods may lead to severely biased estimators for fixed effects parameters; (ii) proposing a new two-step estimation methodology where predictions of the random effects are first computed by a pseudo likelihood approach (the LQMM method) and then used as offsets in standard quantile regression; (iii) proposing a novel bootstrap sampling procedure in order to reduce bias of the two-step estimator and compute confidence intervals. The proposed estimation and associated inference is assessed numerically through rigorous simulation studies and applied to an AIDS Clinical Trial Group (ACTG) study.

AB - Quantile regression models with random effects are useful for studying associations between covariates and quantiles of the response distribution for clustered data. Parameter estimation is examined for a class of mixed-effects quantile regression models, with focus on settings with many but small clusters. The main contributions are the following: (i) documenting that existing methods may lead to severely biased estimators for fixed effects parameters; (ii) proposing a new two-step estimation methodology where predictions of the random effects are first computed by a pseudo likelihood approach (the LQMM method) and then used as offsets in standard quantile regression; (iii) proposing a novel bootstrap sampling procedure in order to reduce bias of the two-step estimator and compute confidence intervals. The proposed estimation and associated inference is assessed numerically through rigorous simulation studies and applied to an AIDS Clinical Trial Group (ACTG) study.

KW - AIDS clinical trial group study

KW - Bias-adjustment

KW - Clustered data

KW - Linear quantile regression

KW - Random effects

KW - Wild bootstrap

UR - http://www.scopus.com/inward/record.url?scp=85120695107&partnerID=8YFLogxK

U2 - 10.1016/j.ecosta.2021.07.003

DO - 10.1016/j.ecosta.2021.07.003

M3 - Journal article

AN - SCOPUS:85120695107

VL - 23

SP - 165

EP - 186

JO - Econometrics and Statistics

JF - Econometrics and Statistics

SN - 2452-3062

ER -

ID: 291755082

Department of Mathematical Sciences