Model averaging estimation of generalized linear models with imputed covariates

Risultato della ricerca: Article

2 Citazioni (Scopus)

Abstract

We address the problem of estimating generalized linear models when some covariate values are missingbut imputations are available to fill-in the missing values. This situation generates a bias-precision tradeoffin the estimation of the model parameters. Extending the generalized missing-indicator methodproposed by Dardanoni et al. (2011) for linear regression, we handle this trade-off as a problem ofmodel uncertainty using Bayesian averaging of classical maximum likelihood estimators (BAML). We alsopropose a block model averaging strategy that incorporates information on the missing-data patterns andis computationally simple. An empirical application illustrates our approach.
Lingua originaleEnglish
pagine (da-a)452-463
Numero di pagine12
RivistaJournal of Econometrics
Volume184
Stato di pubblicazionePublished - 2015

Fingerprint

Trade-offs
Linear regression
Missing data
Maximum likelihood estimator
Imputation
Covariates
Generalized linear model
Uncertainty
Model averaging
Missing values

All Science Journal Classification (ASJC) codes

  • Economics and Econometrics

Cita questo

@article{52b2581ae6304364aa86b10431ed411d,
title = "Model averaging estimation of generalized linear models with imputed covariates",
abstract = "We address the problem of estimating generalized linear models when some covariate values are missingbut imputations are available to fill-in the missing values. This situation generates a bias-precision tradeoffin the estimation of the model parameters. Extending the generalized missing-indicator methodproposed by Dardanoni et al. (2011) for linear regression, we handle this trade-off as a problem ofmodel uncertainty using Bayesian averaging of classical maximum likelihood estimators (BAML). We alsopropose a block model averaging strategy that incorporates information on the missing-data patterns andis computationally simple. An empirical application illustrates our approach.",
author = "Salvatore Modica and {De Luca}, Giuseppe and Valentino Dardanoni and Franco Peracchi",
year = "2015",
language = "English",
volume = "184",
pages = "452--463",
journal = "Journal of Econometrics",
issn = "0304-4076",
publisher = "Elsevier BV",

}

TY - JOUR

T1 - Model averaging estimation of generalized linear models with imputed covariates

AU - Modica, Salvatore

AU - De Luca, Giuseppe

AU - Dardanoni, Valentino

AU - Peracchi, Franco

PY - 2015

Y1 - 2015

N2 - We address the problem of estimating generalized linear models when some covariate values are missingbut imputations are available to fill-in the missing values. This situation generates a bias-precision tradeoffin the estimation of the model parameters. Extending the generalized missing-indicator methodproposed by Dardanoni et al. (2011) for linear regression, we handle this trade-off as a problem ofmodel uncertainty using Bayesian averaging of classical maximum likelihood estimators (BAML). We alsopropose a block model averaging strategy that incorporates information on the missing-data patterns andis computationally simple. An empirical application illustrates our approach.

AB - We address the problem of estimating generalized linear models when some covariate values are missingbut imputations are available to fill-in the missing values. This situation generates a bias-precision tradeoffin the estimation of the model parameters. Extending the generalized missing-indicator methodproposed by Dardanoni et al. (2011) for linear regression, we handle this trade-off as a problem ofmodel uncertainty using Bayesian averaging of classical maximum likelihood estimators (BAML). We alsopropose a block model averaging strategy that incorporates information on the missing-data patterns andis computationally simple. An empirical application illustrates our approach.

UR - http://hdl.handle.net/10447/97660

UR - http://www.sciencedirect.com/science/article/pii/S0304407614001420

M3 - Article

VL - 184

SP - 452

EP - 463

JO - Journal of Econometrics

JF - Journal of Econometrics

SN - 0304-4076

ER -