This 2-part multilevel modeling (MLM) tutorial is designed for newbies as well as researchers who have been exposed to it through a prior class or workshop but still have lots of questions.

Topics in Part 1 include:

  1. Identifying if MLM is necessary – the first step is determing whether the data actually violates assumptions of independence.
  2. Figuring out the nested structure of your data (including cross-classified models) – Identifying the sources of non-independence in your data, including the possibility of cross-classification.
  3. Approaches to dealing with non-independence – when to deal with non-independence through random versus fixed factors.

Topics in Part 2 include: 

1. Fixed versus random effects – the difference between the two and what changes in the analysis process when random slopes are allowed in the model.

2. Grand-mean versus group centering – what they are and when to use them, unconfounding within and between person effects.

3. Covariance matrices – cover the basics of the residual and random effects covariance matrices.

In this series

About the Presenter

Amie M. Gordon headshot

Amie M. Gordon is an Assistant Professor of Psychology at the University of Michigan. Her research focuses on uncovering the social cognitive, affective, and biological factors that shape our closest relationships.