This short course will cover the most commonly used statistical methods for analysing longitudinal data, including general linear models with various correlated random errors, such as compound symmetry, AR(1), exponential correlation, ante-dependence structures and so on, and linear mixed models or random effects models. Estimation methods and goodness-of-fit will be addressed in details. Real data from medical sciences and public health will be analysed for illustration. Statistical software R will be used in the teaching.
Registration by 31.3.2010 with e-form: https://elomake3.uta.fi/lomakkeet/3006/lomake.html. More information from professor Tapio Nummi (tan@uta.fi).