Department of Statistics Seminar
North Carolina StateUniversity

presents

Dr. Annie Qu

OregonStateUniversity

 

Model Diagnostic Tests for Selecting Informative Correlation

 

Abstract

 

In the generalized method of moments approach to longitudinal data analysis, unbiased estimating functions can be constructed to incorporate both the marginal mean and the correlation structure of the data. Increasing the number of parameters in the correlation structure corresponds to increasing the number of estimating functions. Thus, building a correlation model is equivalent to selecting estimating functions. This paper proposes a chi-squared test to choose informative unbiased estimating functions. We show that this methodology is useful to identify which source of correlation is important to incorporate when there are multiple possible sources of correlation. This method can also be applied to determine the optimal working correlation for the generalized estimating equation approach. Lung cancer prevention data from MD Anderson will be illustrated for the proposed method.

This is joint work with J. Jack Lee of MD Andersonand Bruce Lindsay of PennState.

Friday, November 30, 2007
3:35 - 4:35 pm
301 Riddick Hall
Refreshments will be served in the common area of 301 Riddick at 3:00 pm.