Testing for outliers in nonlinear longitudinal data models based on M-estimation

Abstract

In this paper we propose and analyze nonlinear mixed-effects models for longitudinal data, obtaining robust maximum likelihood estimates for the parameters by introducing Huber's function in the log-likelihood function. Furthermore, the test for outliers in the model based on robust estimation is investigated through generalized Cook's distance. The obtained results are illustrated by plasma concentrations data presented in Davidian and Giltiman, which was analyzed under the non-robust situation.