This course covers statistical methods to be used when one or more variables are repeatedly measured in time on the same experimental unit. For instance, in a clinical trial, the outcome variable can be measured at baseline and at different times during the treatment period. In a meta-analysis, the study can be regarded as the experimental unit and the observations of patients within the same study as repeated measurements.
In the last 10 or 15 years much progress has been made in the development of new methods of analysis. In recent years several of these new methods have been implemented in a wide variety of computer packages.
The course starts with a short overview of simple methods for analyzing repeated measurements data, followed by a short recap of the most basic concepts of linear algebra needed for the presentation of the most advanced models.
Then the main focus turns on more advanced methods. For approximately normally distributed repeated measurements outcomes marginal and linear mixed models are introduced. For non-normal responses, first the generalized estimating equations (GEE) approach for marginal inferences is presented, followed by extensions of random effects models to categorical outcomes. All these methods are exemplified using data from of clinical and epidemiological studies.
Computer practicals in the statistical programming language R will be used to acquire hands on experience in applying these techniques to real data. All code used during the course will be live demonstrated using a web app, which will be made available to participants.
Participants need to bring their own laptop with certain pre-installed software (instructions will be given before start of the course). The assignment for this course needs to be handed in two weeks after the ending of the course.
Please note that this courses has a blended design, which means that it includes both online modules and live meetings. The first meeting lasts 1 hour, the following meetings last 2 hours.
