Abstract
We consider two consistent estimators for the parameters of the linear predictor in the Poisson regression model, where the covariate is measured with errors. The measurement errors are assumed to be normally distributed with known error variance σ 2 u . The SQS estimator, based on a conditional mean-variance model, takes the distribution of the latent covariate into account, and this is here assumed to be a normal distribution. The CS estimator, based on a corrected score function, does not use the distribution of the latent covariate. Nevertheless, for small σ 2 u , both estimators have identical asymptotic covariance matrices up to the order of σ 2 u . We also compare the consistent estimators to the naive estimator, which is based on replacing the latent covariate with its (erroneously) measured counterpart. The naive estimator is biased, but has a smaller covariance matrix than the consistent estimators (at least up to the order of σ 2 u ).
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Cameron AC, Trivedi PK (1998).Regression Analysis of Count Data. Cambridge University Press.
Carroll RJ, Ruppert D, Stefanski LA (1995).Measurement Error in Nonlinear Models. Chapman and Hall, London.
Edlefsen LE, Jones SD (1995).GAUSS version 3.2.6, Kent, Washington: Aptech Systems, Inc.
Fahrmeir L, Tutz G (1994).Multivariate Statistical Modelling Based on Generalized Linear Models. Springer, Heidelberg.
Heyde CC (1997).Quasi-Likelihood And Its Application. Springer, New York
Kukush A, Schneeweiss H (2000).A Comparison of Asymptotic Covariance Matrices of Adjusted Least Squares and Structural Least Squares in Error Ridden Polynomial Regression. Discussion Paper 218, SFB 386, Ludwig-Maximilians-Universität München.
Kukush A, Schneeweiss H, Wolf R, (2001).Comparison of Three Estimators in a Polynomial Regression with Measurement Errors. Discussion Paper 233, SFB 386, Ludwig-Maximilians-Universität München.
Nakamura T (1992).Corrected score functions for errors-in-variables models: methodology and applications to generalized linear models. Biometrika 77, 127–137.
Stefanski LA (1989).Unbiased estimation of a nonlinear function of a normal mean with application to measurement error models. Communications in Statistics A, 18, 4335–4358.
Stefanski LA, Carroll RJ (1987).Conditional scores and optimal scores in generalized linear measurement error models. Biometrika 74, 703–716.
Thamerus M (1997).Modelling Count Data with Heteroscedastic Measurement Error in the Covariates. Discussion Paper 58, SFB 386, Ludwig-Maximilians-Universität München.
Thamerus M (1998).Different nonlinear regression models with incorrectly observed covariates. In: Galata, R. and H. Küchenhoff (eds.).Econometrics in Theory and Practice. Physica, Heidelberg.
Wedderburn RWM (1976).On the existence and uniqueness of the maximum likelihood estimates for certain generalized linear models. Biometrika 63, 27–32.
Winkelmann R (1997).Econometric Analysis of Count Data. (2nd ed.) Springer, Berlin.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Kukush, A., Schneeweis, H. & Wolf, R. Three estimators for the poisson regression model with measurement errors. Statistical Papers 45, 351–368 (2004). https://doi.org/10.1007/BF02777577
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/BF02777577