Portal de Periódicos da CAPES

Modeling the Mean: Analyzing Response Profiles

2011; Wiley; Linguagem: Inglês

10.1002/9781119513469.ch5

1940-6347

Statistical Methods and Bayesian Inference

This chapter presents a method for analyzing longitudinal data that imposes minimal structure or restrictions on the mean response over time and on the covariance among the repeated measures. The method focuses on analyzing response profiles and can be applied to longitudinal data when the design is balanced, with the timing of the repeated measures common to all individuals in the study. The chapter considers how the analysis of response profiles can be implemented in the general linear model. It illustrates main ideas by conducting an analysis of response profiles of the blood lead data of 100 children from the succimer and placebo groups of the Treatment of Lead-Exposed Children (TLC) Trial. The MIXED procedure in SAS is a general procedure for fitting linear models to longitudinal and clustered data. The chapter presents illustrative source code for an analysis of response profiles in general terms and describes the most salient parts of the command syntax.