Test the reliability of doubly robust estimation with missing response data

Baojiang Chen, Jing Qin

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

In statistical inference, one has to make sure that the underlying regression model is correctly specified otherwise the resulting estimation may be biased. Model checking is an important method to detect any departure of the regression model from the true one. Missing data are a ubiquitous problem in social and medical studies. If the underlying regression model is correctly specified, recent researches show great popularity of the doubly robust (DR) estimates method for handling missing data because of its robustness to the misspecification of either the missing data model or the conditional mean model, that is, the model for the conditional expectation of true regression model conditioning on the observed quantities. However, little work has been devoted to the goodness of fit test for DR estimates method. In this article, we propose a testing method to assess the reliability of the estimator derived from the DR estimating equation with possibly missing response and always observed auxiliary variables. Numerical studies demonstrate that the proposed test can control type I errors well. Furthermore the proposed method can detect departures from model assumptions in the marginal mean model of interest powerfully. A real dementia data set is used to illustrate the method for the diagnosis of model misspecification in the problem of missing response with an always observed auxiliary variable for cross-sectional data.

Original languageEnglish (US)
Pages (from-to)289-298
Number of pages10
JournalBiometrics
Volume70
Issue number2
DOIs
StatePublished - Jun 2014

Fingerprint

Robust Estimation
Regression Model
Missing Data
Robust Estimate
Auxiliary Variables
testing
Dementia
Model Misspecification
Misspecification
Estimating Equation
Type I error
Social Problems
Conditional Expectation
Goodness of Fit Test
Statistical Inference
Conditioning
Model
Data Model
Model Checking
Biased

Keywords

  • Auxiliary
  • Doubly robust
  • Estimating equation
  • Goodness of fit
  • Missing data

ASJC Scopus subject areas

  • Statistics and Probability
  • Biochemistry, Genetics and Molecular Biology(all)
  • Immunology and Microbiology(all)
  • Agricultural and Biological Sciences(all)
  • Applied Mathematics

Cite this

Test the reliability of doubly robust estimation with missing response data. / Chen, Baojiang; Qin, Jing.

In: Biometrics, Vol. 70, No. 2, 06.2014, p. 289-298.

Research output: Contribution to journalArticle

Chen, Baojiang ; Qin, Jing. / Test the reliability of doubly robust estimation with missing response data. In: Biometrics. 2014 ; Vol. 70, No. 2. pp. 289-298.
@article{c51d1346ac904fa5874d942922156a66,
title = "Test the reliability of doubly robust estimation with missing response data",
abstract = "In statistical inference, one has to make sure that the underlying regression model is correctly specified otherwise the resulting estimation may be biased. Model checking is an important method to detect any departure of the regression model from the true one. Missing data are a ubiquitous problem in social and medical studies. If the underlying regression model is correctly specified, recent researches show great popularity of the doubly robust (DR) estimates method for handling missing data because of its robustness to the misspecification of either the missing data model or the conditional mean model, that is, the model for the conditional expectation of true regression model conditioning on the observed quantities. However, little work has been devoted to the goodness of fit test for DR estimates method. In this article, we propose a testing method to assess the reliability of the estimator derived from the DR estimating equation with possibly missing response and always observed auxiliary variables. Numerical studies demonstrate that the proposed test can control type I errors well. Furthermore the proposed method can detect departures from model assumptions in the marginal mean model of interest powerfully. A real dementia data set is used to illustrate the method for the diagnosis of model misspecification in the problem of missing response with an always observed auxiliary variable for cross-sectional data.",
keywords = "Auxiliary, Doubly robust, Estimating equation, Goodness of fit, Missing data",
author = "Baojiang Chen and Jing Qin",
year = "2014",
month = "6",
doi = "10.1111/biom.12150",
language = "English (US)",
volume = "70",
pages = "289--298",
journal = "Biometrics",
issn = "0006-341X",
publisher = "Wiley-Blackwell",
number = "2",

}

TY - JOUR

T1 - Test the reliability of doubly robust estimation with missing response data

AU - Chen, Baojiang

AU - Qin, Jing

PY - 2014/6

Y1 - 2014/6

N2 - In statistical inference, one has to make sure that the underlying regression model is correctly specified otherwise the resulting estimation may be biased. Model checking is an important method to detect any departure of the regression model from the true one. Missing data are a ubiquitous problem in social and medical studies. If the underlying regression model is correctly specified, recent researches show great popularity of the doubly robust (DR) estimates method for handling missing data because of its robustness to the misspecification of either the missing data model or the conditional mean model, that is, the model for the conditional expectation of true regression model conditioning on the observed quantities. However, little work has been devoted to the goodness of fit test for DR estimates method. In this article, we propose a testing method to assess the reliability of the estimator derived from the DR estimating equation with possibly missing response and always observed auxiliary variables. Numerical studies demonstrate that the proposed test can control type I errors well. Furthermore the proposed method can detect departures from model assumptions in the marginal mean model of interest powerfully. A real dementia data set is used to illustrate the method for the diagnosis of model misspecification in the problem of missing response with an always observed auxiliary variable for cross-sectional data.

AB - In statistical inference, one has to make sure that the underlying regression model is correctly specified otherwise the resulting estimation may be biased. Model checking is an important method to detect any departure of the regression model from the true one. Missing data are a ubiquitous problem in social and medical studies. If the underlying regression model is correctly specified, recent researches show great popularity of the doubly robust (DR) estimates method for handling missing data because of its robustness to the misspecification of either the missing data model or the conditional mean model, that is, the model for the conditional expectation of true regression model conditioning on the observed quantities. However, little work has been devoted to the goodness of fit test for DR estimates method. In this article, we propose a testing method to assess the reliability of the estimator derived from the DR estimating equation with possibly missing response and always observed auxiliary variables. Numerical studies demonstrate that the proposed test can control type I errors well. Furthermore the proposed method can detect departures from model assumptions in the marginal mean model of interest powerfully. A real dementia data set is used to illustrate the method for the diagnosis of model misspecification in the problem of missing response with an always observed auxiliary variable for cross-sectional data.

KW - Auxiliary

KW - Doubly robust

KW - Estimating equation

KW - Goodness of fit

KW - Missing data

UR - http://www.scopus.com/inward/record.url?scp=84902329021&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84902329021&partnerID=8YFLogxK

U2 - 10.1111/biom.12150

DO - 10.1111/biom.12150

M3 - Article

C2 - 24571677

AN - SCOPUS:84902329021

VL - 70

SP - 289

EP - 298

JO - Biometrics

JF - Biometrics

SN - 0006-341X

IS - 2

ER -