Assessing GEE Models with Longitudinal Ordinal Data by Global Odds Ratio
Kuo-Chin Lin1, Yi-Ju Chen2
1Graduate Institute of Business and Management, Tainan University of Technology, Tainan, Taiwan; 2Department of Statistics, Tamkang University, New Taipei, Taiwan

Longitudinal studies are commonly occurred in the social and biomedical sciences. The overall test for model adequacy is an essential issue in longitudinal categorical data analysis. Two goodness-of-fit tests are proposed for the generalized estimating equations (GEE) models with longitudinal ordinal data in terms of Pearson chi-squared and unweighted sum of square residual using the global odds ratio as a measure of association. Our method can be regarded as the extension of method proposed by Lin (Computational Statistics and Data Analysis 2010; 54:1872-1880) considering four major variants of working correlation structures. For large samples, the distributions of the two test statistics are approximated by the standard normal distribution based on the asymptotic means and variances. In simulation studies, the type I error rate and the power performance comparison of the proposed tests and current tests are presented for various sample sizes. The application of the proposed tests is illustrated by a numerical example.

Keywords: GEE; Global odds ratio; Goodness-of-fit; Longitudinal ordinal data

Biography: Professor, Graduate Institute of Business and Management, Dean, College of Management, Tainan University of Technology.

Education: Ph.D. in Statistics (1996), University of Missouri-Columbia, USA

Research Interests: Nonparametric smoothing, Longitudinal data analysis, Goodness-of-fit test