Flexible parametric approach to classical measurement error variance estimation without auxiliary data : Classical Measurement Error Variance Estimation

Bertrand, Aurélie [UCL] Van Keilegom, Ingrid [UCL] Legrand, Catherine [UCL]

Measurement error in the continuous covariates of a model generally yields bias in the estimators. It is a frequent problem in practice, and many correction procedures have been developed for different classes of models. However, in most cases, some information about the measurement error distribution is required. When neither validation nor auxiliary data (e.g., replicated measurements) are available, this specification turns out to be tricky. In this article, we develop a flexible likelihood-based procedure to estimate the variance of classical additive error of Gaussian distribution, without additional information, when the covariate has compact support. The performance of this estimator is investigated both in an asymptotic way and through finite sample simulations. The usefulness of the obtained estimator when using the simulation extrapolation (SIMEX) algorithm, a widely used correction method, is then analyzed in the Cox proportional hazards model through other simulations. Finally, the whole procedure is illustrated on real data.

• metadata
• References
• relations

Document type	Article de périodique (Journal article) – Article de recherche
Access type	Accès libre
Publication date	2019
Language	Anglais
Journal information	"Biometrics" - Vol. 75, no. 1, p. 297-307 (2019)
Peer reviewed	yes
Publisher	Wiley
issn	0006-341X
Publication status	Publié
Affiliation	UCL - SSH/LIDAM/ISBA - Institut de Statistique, Biostatistique et Sciences Actuarielles
Keywords	Error variance ; Measurement error ; Proportional hazards model ; Survival analysis
Links	https://doi.org/10.1111/biom.12960[DOI] http://hdl.handle.net/2078.1/214798[Handle]

1. Bernstein, Communications de la Société mathématique de Kharkow, 13, 1 (1912)
2. Bertrand A., Legrand C., Léonard D., Van Keilegom I., Robustness of estimation methods in a survival cure model with mismeasured covariates, 10.1016/j.csda.2016.11.013
3. Butucea Cristina, Matias Catherine, Minimax estimation of the noise level and of the deconvolution density in a semiparametric convolution model, 10.3150/bj/1116340297
4. Butucea Cristina, Matias Catherine, Pouet Christophe, Adaptivity in convolution models with partially known noise distribution, 10.1214/08-ejs225
5. Carroll Raymond J., Roeder Kathryn, Wasserman Larry, Flexible Parametric Measurement Error Models, 10.1111/j.0006-341x.1999.00044.x
6. Carroll, Measurement Error in Nonlinear Models: A Modern Perspective. (2006)
7. Cook J. R., Stefanski L. A., Simulation-Extrapolation Estimation in Parametric Measurement Error Models, 10.1080/01621459.1994.10476871
8. Cox, Journal of the Royal Statistical Society, Series B, 34, 187 (1972)
9. Crainiceanu, Johns Hopkins University, Dept. of Biostatistics Working Papers, 116, 1 (2006)
10. Delaigle Aurore, Hall Peter, Methodology for non-parametric deconvolution when the error distribution is unknown, 10.1111/rssb.12109
11. Gorfine M., Hsu L., Prentice R. L., Nonparametric correction for covariate measurement error in a stratified Cox model, 10.1093/biostatistics/5.1.75
12. Guan Zhong, Efficient and robust density estimation using Bernstein type polynomials, 10.1080/10485252.2016.1163349
13. Hu Ping, Tsiatis Anastasios A., Davidian Marie, Estimating the Parameters in the Cox Model When Covariate Variables are Measured with Error, 10.2307/2533667
14. Janssen Paul, Swanepoel Jan, Veraverbeke Noël, Large sample behavior of the Bernstein copula estimator, 10.1016/j.jspi.2011.11.020
15. Janssen Paul, Swanepoel Jan, Veraverbeke Noël, A note on the asymptotic behavior of the Bernstein estimator of the copula density, 10.1016/j.jmva.2013.10.009
16. Leblanc Alexandre, On estimating distribution functions using Bernstein polynomials, 10.1007/s10463-011-0339-4
17. Matias Catherine, Semiparametric deconvolution with unknown noise variance, 10.1051/ps:2002015
18. Meister, Statistica Sinica, 16, 195 (2006)
19. Meister A., Deconvolving compactly supported densities, 10.3103/s106653070701005x
20. PRENTICE R. L., Covariate measurement errors and parameter estimation in a failure time regression model, 10.1093/biomet/69.2.331
21. 2017 R: A Language and Environment for Statistical Computing Vienna, Austria: R Foundation for Statistical Computing
22. Reiersol Olav, Identifiability of a Linear Relation between Variables Which Are Subject to Error, 10.2307/1907835
23. Schennach Susanne M., Recent Advances in the Measurement Error Literature, 10.1146/annurev-economics-080315-015058
24. Schennach S. M., Hu Yingyao, Nonparametric Identification and Semiparametric Estimation of Classical Measurement Error Models Without Side Information, 10.1080/01621459.2012.751872
25. Schwarz Maik, Van Bellegem Sébastien, Consistent density deconvolution under partially known error distribution, 10.1016/j.spl.2009.10.012
26. Wang J., Ghosh S.K., Shape restricted nonparametric regression with Bernstein polynomials, 10.1016/j.csda.2012.02.018
27. White Halbert, Maximum Likelihood Estimation of Misspecified Models, 10.2307/1912526
28. Yan, Statistical Methods on Survival Data with Measurement Error (2014)
29. Zhang Daowen, Davidian Marie, Linear Mixed Models with Flexible Distributions of Random Effects for Longitudinal Data, 10.1111/j.0006-341x.2001.00795.x

See also
	[boreal:187163]	Flexible parametric approach to classical measurement error variance estimation without auxiliary data