For multilevel designs, standard regression techniques lead to biased model-parameter estimates and incorrect standard errors. Biased regression parameter estimates typically occur if the subjects were not measured at the same time points (unbalanced data) and if missing observations are involved. In this course, the main emphasis is on linear models with random effects. Non-linear models (like Logistic model) are only mentioned briefly. Special attention will be given on the analysis of longitudinal intervention studies, where the objective is to evaluate some treatment effect. More information about the course can be found in the attachment. |