A nonparametric multiple imputation approach for data with missing covariate values with application to colorectal adenoma data

Chiu-Hsieh Hsu, Qi Long, Yisheng Li, Elizabeth T Jacobs

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

A nearest neighbor-based multiple imputation approach is proposed to recover missing covariate information using the predictive covariates while estimating the association between the outcome and the covariates. To conduct the imputation, two working models are fitted to define an imputing set. This approach is expected to be robust to the underlying distribution of the data. We show in simulation and demonstrate on a colorectal data set that the proposed approach can improve efficiency and reduce bias in a situation with missing at random compared to the complete case analysis and the modified inverse probability weighted method.

Original languageEnglish (US)
Pages (from-to)634-648
Number of pages15
JournalJournal of Biopharmaceutical Statistics
Volume24
Issue number3
DOIs
StatePublished - May 4 2014

Fingerprint

Missing Covariates
Multiple Imputation
Adenoma
Covariates
Missing at Random
Imputation
Nearest Neighbor
Demonstrate
Simulation
Datasets
Model

Keywords

  • Missing at random
  • Multiple imputation
  • Nearest neighbor
  • Nonparametric imputation.

ASJC Scopus subject areas

  • Pharmacology (medical)
  • Pharmacology
  • Statistics and Probability
  • Medicine(all)

Cite this

@article{51d4887f13fd4b6aa6eddcb9151fdb3e,
title = "A nonparametric multiple imputation approach for data with missing covariate values with application to colorectal adenoma data",
abstract = "A nearest neighbor-based multiple imputation approach is proposed to recover missing covariate information using the predictive covariates while estimating the association between the outcome and the covariates. To conduct the imputation, two working models are fitted to define an imputing set. This approach is expected to be robust to the underlying distribution of the data. We show in simulation and demonstrate on a colorectal data set that the proposed approach can improve efficiency and reduce bias in a situation with missing at random compared to the complete case analysis and the modified inverse probability weighted method.",
keywords = "Missing at random, Multiple imputation, Nearest neighbor, Nonparametric imputation.",
author = "Chiu-Hsieh Hsu and Qi Long and Yisheng Li and Jacobs, {Elizabeth T}",
year = "2014",
month = "5",
day = "4",
doi = "10.1080/10543406.2014.888444",
language = "English (US)",
volume = "24",
pages = "634--648",
journal = "Journal of Biopharmaceutical Statistics",
issn = "1054-3406",
publisher = "Taylor and Francis Ltd.",
number = "3",

}

TY - JOUR

T1 - A nonparametric multiple imputation approach for data with missing covariate values with application to colorectal adenoma data

AU - Hsu, Chiu-Hsieh

AU - Long, Qi

AU - Li, Yisheng

AU - Jacobs, Elizabeth T

PY - 2014/5/4

Y1 - 2014/5/4

N2 - A nearest neighbor-based multiple imputation approach is proposed to recover missing covariate information using the predictive covariates while estimating the association between the outcome and the covariates. To conduct the imputation, two working models are fitted to define an imputing set. This approach is expected to be robust to the underlying distribution of the data. We show in simulation and demonstrate on a colorectal data set that the proposed approach can improve efficiency and reduce bias in a situation with missing at random compared to the complete case analysis and the modified inverse probability weighted method.

AB - A nearest neighbor-based multiple imputation approach is proposed to recover missing covariate information using the predictive covariates while estimating the association between the outcome and the covariates. To conduct the imputation, two working models are fitted to define an imputing set. This approach is expected to be robust to the underlying distribution of the data. We show in simulation and demonstrate on a colorectal data set that the proposed approach can improve efficiency and reduce bias in a situation with missing at random compared to the complete case analysis and the modified inverse probability weighted method.

KW - Missing at random

KW - Multiple imputation

KW - Nearest neighbor

KW - Nonparametric imputation.

UR - http://www.scopus.com/inward/record.url?scp=84898962290&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84898962290&partnerID=8YFLogxK

U2 - 10.1080/10543406.2014.888444

DO - 10.1080/10543406.2014.888444

M3 - Article

C2 - 24697618

AN - SCOPUS:84898962290

VL - 24

SP - 634

EP - 648

JO - Journal of Biopharmaceutical Statistics

JF - Journal of Biopharmaceutical Statistics

SN - 1054-3406

IS - 3

ER -