Biometrika Advance Access originally published online on June 13, 2008
Biometrika 2008 95(3):653-666; doi:10.1093/biomet/asn006
Articles |
Generalized varying coefficient models for longitudinal data
entürkDepartment of Statistics, Pennsylvania State University, University Park, Pennsylvania 16802, U.S.A. dsenturk{at}stat.psu.edu
Department of Statistics, University of California, Davis, California 95616, U.S.A. mueller{at}wald.ucdavis.edu
Received for publication 1 July 2006. Revision received 1 November 2007.
We propose a generalization of the varying coefficient model for longitudinal data to cases where not only current but also recent past values of the predictor process affect current response. More precisely, the targeted regression coefficient functions of the proposed model have sliding window supports around current time t. A variant of a recently proposed two-step estimation method for varying coefficient models is proposed for estimation in the context of these generalized varying coefficient models, and is found to lead to improvements, especially for the case of additive measurement errors in both response and predictors. The proposed methodology for estimation and inference is also applicable for the case of additive measurement error in the common versions of varying coefficient models that relate only current observations of predictor and response processes to each other. Asymptotic distributions of the proposed estimators are derived, and the model is applied to the problem of predicting protein concentrations in a longitudinal study. Simulation studies demonstrate the efficacy of the proposed estimation procedure.
Key Words: Linear regression • Measurement error model • Prediction • Smoothing • Two-step procedure
References
-
Cai Z., Fan J., Yao Q. Functional coefficient regression models for nonlinear time series. J. Am. Statist. Assoc. (2000) 95:941–56.[CrossRef][Web of Science]
Carroll R. J., Ruppert D., Tosteson T., Crainiceanu C., Karagas M. Instrumental variables and nonparametric regression. J. Am. Statist. Assoc. (2004) 99:736–50.[CrossRef][Web of Science]
Chen R., Tsay R. S. Functional coefficient autoregressive models. J. Am. Statist. Assoc. (1993) 88:298–308.[CrossRef][Web of Science]
Chiang C., Rice J. A., Wu C. O. Smoothing spline estimation for varying coefficient models with repeatedly measured dependent variables. J. Am. Statist. Assoc. (2001) 96:605–17.[CrossRef][Web of Science]
Fan J., Gijbels I. Local Polynomial Modeling and Its Applications (1996) London: Chapman and Hall.
Fan J., Zhang J. Two-step estimation of functional linear models with applications to longitudinal data. J. R. Statist. Soc. (2000) B 62:303–22.[CrossRef]
Fan J., Yao Q., Cai Z. Adaptive varying coefficient linear models. J. R. Statist. Soc. (2003) B 65:57–80.[CrossRef]
Hastie T. J., Tibshirani R. J. Varying-coefficient models (with Discussion). J. R. Statist Soc. (1993) B 55:757–96.
Hoover D. R., Rice J. A., Wu C. O., Yang L.-P. Nonparametric smoothing estimates of time-varying coefficient models with longitudinal data. Biometrika (1998) 85:809–22.
Huang J. Z., Wu C. O., Zhou L. Polynomial spline estimation and inference for varying coefficient models with longitudinal data. Statist. Sinica (2004) 14:763–88.
Kaysen G. A., Dubin J. A., Müller H. G., Mitch W. E., Levin N. W., HEMO Study Group. Levels of 
1 acid glycoprotein and ceruloplasmin predict future albumin levels in hemodialysis patients. Kidney Int. (2001) 60:2360–6.[CrossRef][Web of Science][Medline]
Wu C. O., Chiang C. T. Kernel smoothing on varying coefficient models with longitudinal dependent variable. Statist. Sinica (2000) 10:433–56.
Wu C. O., Yu K. F. Nonparametric varying coefficient models for the analysis of longitudinal data. Int. Statist. Rev. (2002) 70:373–93.
Wu C. O., Yu K. F., Chiang C. T. A two-step smoothing method for varying coefficient models with repeated measurements. Ann. Inst. Statist. Math. (2000) 52:519–43.[CrossRef]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||