Imputation for semiparametric transformation models with biased-sampling data

Hao Liu, Jing Qin, Yu Shen

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Widely recognized in many fields including economics, engineering, epidemiology, health sciences, technology and wildlife management, length-biased sampling generates biased and right-censored data but often provide the best information available for statistical inference. Different from traditional right-censored data, length-biased data have unique aspects resulting from their sampling procedures. We exploit these unique aspects and propose a general imputation-based estimation method for analyzing length-biased data under a class of flexible semiparametric transformation models. We present new computational algorithms that can jointly estimate the regression coefficients and the baseline function semiparametrically. The imputation-based method under the transformation model provides an unbiased estimator regardless whether the censoring is independent or not on the covariates. We establish large-sample properties using the empirical processes method. Simulation studies show that under small to moderate sample sizes, the proposed procedure has smaller mean square errors than two existing estimation procedures. Finally, we demonstrate the estimation procedure by a real data example.

Original languageEnglish (US)
Pages (from-to)470-503
Number of pages34
JournalLifetime Data Analysis
Volume18
Issue number4
DOIs
StatePublished - Aug 20 2012
Externally publishedYes

Fingerprint

Biased Sampling
Transformation Model
Imputation
Semiparametric Model
Sampling
Biased
Right-censored Data
Epidemiology
Mean square error
Empirical Process
Unbiased estimator
Computational Algorithm
Censoring
Regression Coefficient
Statistical Inference
Health
Covariates
Baseline
Sample Size
Economics

Keywords

  • Biased sampling
  • Estimating equation
  • Imputation
  • Transformation models

ASJC Scopus subject areas

  • Applied Mathematics

Cite this

Imputation for semiparametric transformation models with biased-sampling data. / Liu, Hao; Qin, Jing; Shen, Yu.

In: Lifetime Data Analysis, Vol. 18, No. 4, 20.08.2012, p. 470-503.

Research output: Contribution to journalArticle

@article{7bebc61692474193a1b8fdea45e33daa,
title = "Imputation for semiparametric transformation models with biased-sampling data",
abstract = "Widely recognized in many fields including economics, engineering, epidemiology, health sciences, technology and wildlife management, length-biased sampling generates biased and right-censored data but often provide the best information available for statistical inference. Different from traditional right-censored data, length-biased data have unique aspects resulting from their sampling procedures. We exploit these unique aspects and propose a general imputation-based estimation method for analyzing length-biased data under a class of flexible semiparametric transformation models. We present new computational algorithms that can jointly estimate the regression coefficients and the baseline function semiparametrically. The imputation-based method under the transformation model provides an unbiased estimator regardless whether the censoring is independent or not on the covariates. We establish large-sample properties using the empirical processes method. Simulation studies show that under small to moderate sample sizes, the proposed procedure has smaller mean square errors than two existing estimation procedures. Finally, we demonstrate the estimation procedure by a real data example.",
keywords = "Biased sampling, Estimating equation, Imputation, Transformation models",
author = "Hao Liu and Jing Qin and Yu Shen",
year = "2012",
month = "8",
day = "20",
doi = "10.1007/s10985-012-9225-5",
language = "English (US)",
volume = "18",
pages = "470--503",
journal = "Lifetime Data Analysis",
issn = "1380-7870",
publisher = "Springer Netherlands",
number = "4",

}

TY - JOUR

T1 - Imputation for semiparametric transformation models with biased-sampling data

AU - Liu, Hao

AU - Qin, Jing

AU - Shen, Yu

PY - 2012/8/20

Y1 - 2012/8/20

N2 - Widely recognized in many fields including economics, engineering, epidemiology, health sciences, technology and wildlife management, length-biased sampling generates biased and right-censored data but often provide the best information available for statistical inference. Different from traditional right-censored data, length-biased data have unique aspects resulting from their sampling procedures. We exploit these unique aspects and propose a general imputation-based estimation method for analyzing length-biased data under a class of flexible semiparametric transformation models. We present new computational algorithms that can jointly estimate the regression coefficients and the baseline function semiparametrically. The imputation-based method under the transformation model provides an unbiased estimator regardless whether the censoring is independent or not on the covariates. We establish large-sample properties using the empirical processes method. Simulation studies show that under small to moderate sample sizes, the proposed procedure has smaller mean square errors than two existing estimation procedures. Finally, we demonstrate the estimation procedure by a real data example.

AB - Widely recognized in many fields including economics, engineering, epidemiology, health sciences, technology and wildlife management, length-biased sampling generates biased and right-censored data but often provide the best information available for statistical inference. Different from traditional right-censored data, length-biased data have unique aspects resulting from their sampling procedures. We exploit these unique aspects and propose a general imputation-based estimation method for analyzing length-biased data under a class of flexible semiparametric transformation models. We present new computational algorithms that can jointly estimate the regression coefficients and the baseline function semiparametrically. The imputation-based method under the transformation model provides an unbiased estimator regardless whether the censoring is independent or not on the covariates. We establish large-sample properties using the empirical processes method. Simulation studies show that under small to moderate sample sizes, the proposed procedure has smaller mean square errors than two existing estimation procedures. Finally, we demonstrate the estimation procedure by a real data example.

KW - Biased sampling

KW - Estimating equation

KW - Imputation

KW - Transformation models

UR - http://www.scopus.com/inward/record.url?scp=84866424625&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84866424625&partnerID=8YFLogxK

U2 - 10.1007/s10985-012-9225-5

DO - 10.1007/s10985-012-9225-5

M3 - Article

VL - 18

SP - 470

EP - 503

JO - Lifetime Data Analysis

JF - Lifetime Data Analysis

SN - 1380-7870

IS - 4

ER -