Longitudinal and multilevel analysis
Course infoSchedule
Course moduleMED-BMS84
Credits (ECTS)3
CategoryMA (Master)
Language of instructionEnglish
Offered byRadboud University; Faculty of Medical Sciences; Biomedische wetenschappen;
dr. F. Atsma
Other course modules lecturer
dr. F. Atsma
Other course modules lecturer
Contactperson for the course
dr. F. Atsma
Other course modules lecturer
dr. R.J.F. Melis
Other course modules lecturer
Academic year2022
W06-A  (06/02/2023 to 31/08/2023)
Starting block
Course mode
RemarksMonday and Tuesday for 4 weeks after the start of the period.
Registration using OSIRISYes
Course open to students from other facultiesYes
Pre-registration openfrom 01/08/2022 up to and including 30/09/2022
Waiting listYes
Placement procedureDone manually by Back Office
ExplanationDone manually by Back Office
Main Objectives:

Upon completing this course the student should be able to:

1.describe and understand the basic principles and the underlying assumptions of (regression) models for the analysis of multilevel, longitudinal and clustered and unclustered survival data.
2.choose appropriate statistical methods to analyze longitudinal, multilevel and survival data gathered in the context of health research
3.Be proficient at using the statistical software package R to fit models and perform computations for multilevel, longitudinal and survival data analyses, and to interpret and report results.

In this course we will discuss advanced statistical methods that are used to address biomedical and epidemiological research questions that require the collection and analysis of multilevel, longitudinal and survival data. The computer package R will be used for the analysis of  data from a range of research contexts in order to obtain practical experience.  

In multilevel data analysis the focus is on the clustering of data at multiple levels, for example, patients who are nested within the same doctors and doctors who are nested within the same hospitals. Even if we are interested in measuring outcomes at an individual level, we need to take into account the clustering in our data, because individuals within clusters are often more similar than individuals between clusters. 

In longitudinal data analysis the focus is on the repeated measurement of an outcome over time, rather than at a single timepoint. This enables the change over time to be observed, and the comparison of rate of change between groups, for example the impact of quitting smoking versus not quitting smoking on the change in lung function over time.

In survival  analysis the focus is on the time until an event of interest occurs, such as  death after the diagnosis of cancer, or age at time of onset of a disease. We will focus on the Cox proportional hazards model and random effects models for clustered and recurrent survival data.     

Required prior knowledge
This is a course on advanced statistics for complex statistical models to answer biomedical and epidemiological research questions. It is assumed that students have knowledge of statistical principles of analysis including (linear) regression modeling and are motivated to study more advanced statistical methodology. Students who passed the course BMS14 (design and analysis of experiments) or BMS61 (statistical modeling in medical research) should have sufficient pre-knowledge. For others who wish to follow the course, it is strongly recommended that students assess their own knowledge and understanding of regression modelling and take sufficient time to be sure that their level is sufficiently high.

The computer assignments are performed within the statistical software package R. Knowledge of this package is a must in order to participate in the computer assignments. 

Instructional modes
Working group

Written exam
Test weight100
Test typeWritten exam
OpportunitiesBlock W06-A, Block W06-A