Generalized Quasi-Likelihood Method in Quantile Regression for Longitudinal Data

• Activity Number: 317
• Type: Contributed
• Date/Time: Tuesday, August 6, 2013 : 8:30 AM to 10:20 AM
• Sponsor: SSC
• Author(s): Xiaoming Lu*+
• Companies: Memorial Univ of Newfoundland
• Keywords: Quantile regression ; Generalized quasi-likelihood ; Longitudinal data ; Induced smoothing

Abstract

Quantile regression is gradually becoming a powerful statistical methodology that complements the classical least-squares regression by examining how covariates influence the location, scale, and shape of the entire response distribution and offering a more complete view of the statistical landscape. This paper proposes efficient estimating functions for quantile regression parameters based on generalized quasi-likelihood (GQL) which is applied to analyse longitudinal data. The generalized quasi-likelihood estimate can be obtained by solving these estimating equations. In order to use the Newton-Raphson iteration method to get convergent estimates, the estimating functions are redefined as smoothed functions which is differentiable with respect to regression parameters. This GQL method for quantile regression gives consistent estimates that have asymptotically normal distributions. Finally, a simulation study is carried out to evaluate the performance of the proposed method and comparisons are also made with some other methods and within different types of data.