REPEATED MEASURE DATA MODELING ASSUMING A
 FINITE MIXTURE OF NORMAL DISTRIBUTIONS FOR THE
 ERROR: A BAYESIAN ANALYSIS

Jorge Alberto ACHCAR[1]

Edson Zangiacomi MARTINEZ1

Margaret de CASTRO[2]

§     ABSTRACT: In this paper, we introduce a Bayesian analysis of repeated measure data using MCMC (Markov Chain Monte Carlo) methods to obtain the posterior summaries of interest. We assume a finite mixture of normal distributions for the error, which gives better fit for the longitudinal data in the presence of covariates. We illustrate the proposed methodology considering a real data set introduced by Castro et al. (2003) related to a dose-response study with different dosages of dexamethasone (dex) to assess the cordicotropic resistance in Cushing disease (CD) using salivary Cortisol.

§     KEYWORDS: Bayesian analysis; MCMC methods; repeated measures; cushing disease.



[1] Departamento de Medicina Social, Faculdade de Medicina de Ribeirão Preto, Universidade de São Paulo -- USP, CEP: 14049-900, Ribeirão Preto, SP, Brazil. E-mail: achcar@fmrp.usp.br / edson@fmrp.usp.br

[2] Departamento de Clínica Médica, Faculdade de Medicina de Ribeirão Preto, Universidade de São Paulo -- USP, CEP: 14049-900, Ribeirão Preto, SP, Brazil. E-mail: castrom@fmrp.usp.br