Empirical likelihood for estimating equations with missing values

Dong Wang, Song Xi Chen

Research output: Contribution to journalArticle

74 Citations (Scopus)

Abstract

We consider an empirical likelihood inference for parameters defined by general estimating equations when some components of the random observations are subject to missingness. As the nature of the estimating equations is wide-ranging, we propose a nonparametric imputation of the missing values from a kernel estimator of the conditional distribution of the missing variable given the always observable variable. The empirical likelihood is used to construct a profile likelihood for the parameter of interest. We demonstrate that the proposed nonparametric imputation can remove the selection bias in the missingness and the empirical likelihood leads to more efficient parameter estimation. The proposed method is further evaluated by simulation and an empirical study on a genetic dataset on recombinant inbred mice.

Original languageEnglish (US)
Pages (from-to)490-517
Number of pages28
JournalAnnals of Statistics
Volume37
Issue number1
DOIs
StatePublished - Feb 1 2009

Fingerprint

Empirical Likelihood
Estimating Equation
Missing Values
Imputation
Selection Bias
Profile Likelihood
Likelihood Inference
Efficient Estimation
Kernel Estimator
Conditional Distribution
Empirical Study
Parameter Estimation
Mouse
Demonstrate
Missing values
Empirical likelihood
Simulation

Keywords

  • Empirical likelihood
  • Estimating equations
  • Kernel estimation
  • Missing values
  • Nonparametric imputation

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Empirical likelihood for estimating equations with missing values. / Wang, Dong; Chen, Song Xi.

In: Annals of Statistics, Vol. 37, No. 1, 01.02.2009, p. 490-517.

Research output: Contribution to journalArticle

Wang, Dong ; Chen, Song Xi. / Empirical likelihood for estimating equations with missing values. In: Annals of Statistics. 2009 ; Vol. 37, No. 1. pp. 490-517.
@article{133a76809bc94600ae5057c08570897c,
title = "Empirical likelihood for estimating equations with missing values",
abstract = "We consider an empirical likelihood inference for parameters defined by general estimating equations when some components of the random observations are subject to missingness. As the nature of the estimating equations is wide-ranging, we propose a nonparametric imputation of the missing values from a kernel estimator of the conditional distribution of the missing variable given the always observable variable. The empirical likelihood is used to construct a profile likelihood for the parameter of interest. We demonstrate that the proposed nonparametric imputation can remove the selection bias in the missingness and the empirical likelihood leads to more efficient parameter estimation. The proposed method is further evaluated by simulation and an empirical study on a genetic dataset on recombinant inbred mice.",
keywords = "Empirical likelihood, Estimating equations, Kernel estimation, Missing values, Nonparametric imputation",
author = "Dong Wang and Chen, {Song Xi}",
year = "2009",
month = "2",
day = "1",
doi = "10.1214/07-AOS585",
language = "English (US)",
volume = "37",
pages = "490--517",
journal = "Annals of Statistics",
issn = "0090-5364",
publisher = "Institute of Mathematical Statistics",
number = "1",

}

TY - JOUR

T1 - Empirical likelihood for estimating equations with missing values

AU - Wang, Dong

AU - Chen, Song Xi

PY - 2009/2/1

Y1 - 2009/2/1

N2 - We consider an empirical likelihood inference for parameters defined by general estimating equations when some components of the random observations are subject to missingness. As the nature of the estimating equations is wide-ranging, we propose a nonparametric imputation of the missing values from a kernel estimator of the conditional distribution of the missing variable given the always observable variable. The empirical likelihood is used to construct a profile likelihood for the parameter of interest. We demonstrate that the proposed nonparametric imputation can remove the selection bias in the missingness and the empirical likelihood leads to more efficient parameter estimation. The proposed method is further evaluated by simulation and an empirical study on a genetic dataset on recombinant inbred mice.

AB - We consider an empirical likelihood inference for parameters defined by general estimating equations when some components of the random observations are subject to missingness. As the nature of the estimating equations is wide-ranging, we propose a nonparametric imputation of the missing values from a kernel estimator of the conditional distribution of the missing variable given the always observable variable. The empirical likelihood is used to construct a profile likelihood for the parameter of interest. We demonstrate that the proposed nonparametric imputation can remove the selection bias in the missingness and the empirical likelihood leads to more efficient parameter estimation. The proposed method is further evaluated by simulation and an empirical study on a genetic dataset on recombinant inbred mice.

KW - Empirical likelihood

KW - Estimating equations

KW - Kernel estimation

KW - Missing values

KW - Nonparametric imputation

UR - http://www.scopus.com/inward/record.url?scp=65349172950&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=65349172950&partnerID=8YFLogxK

U2 - 10.1214/07-AOS585

DO - 10.1214/07-AOS585

M3 - Article

VL - 37

SP - 490

EP - 517

JO - Annals of Statistics

JF - Annals of Statistics

SN - 0090-5364

IS - 1

ER -