Online Program

A Flexible Model for the Mean and Variance Functions with Application to Longitudinal/Clustered Medical Cost Data

*Jinsong Chen, University of Illinois at Chicago 
Lei Liu, Northwestern University 
Daowen Zhang, North Carolina State University 

Keywords: Generalized linear model; semiparametric regression; health econometrics.

We propose a generalized semiparametric models for longitudinal data, in which the mean function is partial linear. Moreover, the variance of response is modeled as an unknown but smooth function of mean. We use the estimating equations for the simultaneous estimation of regression and correlation parameters, and variance function. B-spline is used for estimate nonlinear functions. Asymptotic normality of the estimates are represented. Simulation study is conducted to assess the performance of our method. We also apply our method to the analyze the medical cost data from a multicenter COMBINE randomized controlled trial of alcohol study.