Quantile Marginal Regression for Longitudinal Data with Dropouts

• Activity Number: 344
• Type: Topic Contributed
• Date/Time: Tuesday, August 11, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Korean International Statistical Society
• Author(s): Hyokyoung Hong* and Mi-Ok Kim and Hyunkeun Cho
• Companies: Michigan State University and Cincinnati Children's Hospital Medical Center and Western Michigan University
• Keywords: Empirical likelihood ; Longitudinal data ; Missing at random ; Monotone missing ; Quadratic inference function

Abstract

We develop empirical likelihood inference procedures for longitudinal data with dropouts under the framework of quantile regression. We borrow matrix expansion idea of quadratic inference function and incorporate the within subject correlations under an informative working correlation structure. The proposed procedure does not assume the exact knowledge of the true correlation structure nor estimate the parameters involved in the informative working correlation structure. Theoretical results show that the resulting estimator is asymptotically normal and more efficient than one attained under a working independence correlation structure. We further expand the proposed approach to account for informative dropouts that are missing at random. The methodology is illustrated by empirical studies that include a real life example of HIV longitudinal data analysis.