A weighting approach for GEE analysis with missing data

Cuiling Wang, Myunghee Cho Paik

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

We propose a new weighting (WT) method to handle missing categorical outcomes in longitudinal data analysis using generalized estimating equations (GEE). The proposed WT provides a valid GEE estimator when the data are missing at random (MAR), and has more stable weights and shows advantage in efficiency compared to the inverse probability weighing method in the presence of small observation probabilities. The WT estimator is similar to the stabilized weighting (SWT) estimator under mild conditions, but it is more stable and efficient than SWT when the associations of the outcome with the observation probabilities and the covariate are strong.

Original languageEnglish (US)
Pages (from-to)2397-2411
Number of pages15
JournalCommunications in Statistics - Theory and Methods
Volume40
Issue number13
DOIs
StatePublished - Jan 2011

Fingerprint

Generalized Estimating Equations
Missing Data
Weighting
Estimator
Longitudinal Data Analysis
Missing at Random
Categorical
Covariates
Valid

Keywords

  • Generalized estimating equation (GEE)
  • Imputation
  • Inverse probability weighting (IPW)
  • Longitudinal studies
  • Missing at random (MAR)
  • Stabilized weighting

ASJC Scopus subject areas

  • Statistics and Probability

Cite this

A weighting approach for GEE analysis with missing data. / Wang, Cuiling; Paik, Myunghee Cho.

In: Communications in Statistics - Theory and Methods, Vol. 40, No. 13, 01.2011, p. 2397-2411.

Research output: Contribution to journalArticle

@article{41e8e0de84594660af15590e062b28a4,
title = "A weighting approach for GEE analysis with missing data",
abstract = "We propose a new weighting (WT) method to handle missing categorical outcomes in longitudinal data analysis using generalized estimating equations (GEE). The proposed WT provides a valid GEE estimator when the data are missing at random (MAR), and has more stable weights and shows advantage in efficiency compared to the inverse probability weighing method in the presence of small observation probabilities. The WT estimator is similar to the stabilized weighting (SWT) estimator under mild conditions, but it is more stable and efficient than SWT when the associations of the outcome with the observation probabilities and the covariate are strong.",
keywords = "Generalized estimating equation (GEE), Imputation, Inverse probability weighting (IPW), Longitudinal studies, Missing at random (MAR), Stabilized weighting",
author = "Cuiling Wang and Paik, {Myunghee Cho}",
year = "2011",
month = "1",
doi = "10.1080/03610921003764282",
language = "English (US)",
volume = "40",
pages = "2397--2411",
journal = "Communications in Statistics - Theory and Methods",
issn = "0361-0926",
publisher = "Taylor and Francis Ltd.",
number = "13",

}

TY - JOUR

T1 - A weighting approach for GEE analysis with missing data

AU - Wang, Cuiling

AU - Paik, Myunghee Cho

PY - 2011/1

Y1 - 2011/1

N2 - We propose a new weighting (WT) method to handle missing categorical outcomes in longitudinal data analysis using generalized estimating equations (GEE). The proposed WT provides a valid GEE estimator when the data are missing at random (MAR), and has more stable weights and shows advantage in efficiency compared to the inverse probability weighing method in the presence of small observation probabilities. The WT estimator is similar to the stabilized weighting (SWT) estimator under mild conditions, but it is more stable and efficient than SWT when the associations of the outcome with the observation probabilities and the covariate are strong.

AB - We propose a new weighting (WT) method to handle missing categorical outcomes in longitudinal data analysis using generalized estimating equations (GEE). The proposed WT provides a valid GEE estimator when the data are missing at random (MAR), and has more stable weights and shows advantage in efficiency compared to the inverse probability weighing method in the presence of small observation probabilities. The WT estimator is similar to the stabilized weighting (SWT) estimator under mild conditions, but it is more stable and efficient than SWT when the associations of the outcome with the observation probabilities and the covariate are strong.

KW - Generalized estimating equation (GEE)

KW - Imputation

KW - Inverse probability weighting (IPW)

KW - Longitudinal studies

KW - Missing at random (MAR)

KW - Stabilized weighting

UR - http://www.scopus.com/inward/record.url?scp=79956099447&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79956099447&partnerID=8YFLogxK

U2 - 10.1080/03610921003764282

DO - 10.1080/03610921003764282

M3 - Article

AN - SCOPUS:79956099447

VL - 40

SP - 2397

EP - 2411

JO - Communications in Statistics - Theory and Methods

JF - Communications in Statistics - Theory and Methods

SN - 0361-0926

IS - 13

ER -