Multilevel Modeling in R: First steps

Novice and advanced R users from all professional groups.

• CHF 400.- for members of UZH/ETH and associated institutes
• CHF 600.- for alumni of UZH/ETH, members of other universities, the public sector and non-profit organizations
• CHF 800.- for companies

Persons without current employment can register for the UZH/ETH fee upon request.

This 1-day course is an introduction to multilevel modeling (also called mixed modeling or hierarchical modeling) and its practical application in R. Multilevel models are used when data are hierarchically structured, for instance when the data contains information about students that are clustered in classes and schools, or when data contains multiple measurement occasions of the same person.

The goal of this course is to introduce the basics of multilevel modeling (Random Intercept and Random Slope models, models with cross-level interactions), to demonstrate how multilevel models can be estimated using R, and to practice how Multilelvel models can be interpreted and critically evaluated. We will also touch upon more advanced topics, such as model fit and estimation issues.

The workshop is aimed at participants with a basic understanding of R (e.g. have already visited an introductory course in the past), but little or no previous experience in multilevel modeling. The course can also be attended by participants who have used multilevel models in other statistical software, but please be prepared that multilevel theory will also be covered in the beginning.

During the course, participants will work on multilevel data examples and exercises provided by the course instructor. Participants are also invited to bring their own multilevel research questions and/or data to the course and can work on their own examples if time permits.

The learning goals are:

1) Participants understand the basics of multilevel modeling.

2) Participants can build their own multilevel models.

3) Participants can apply multilevel models in R and interpret the results.