Multilevel analysis techniques are generally understood as multiple regression models for hierarchically ordered data, with subjects nested in groups. An early extension was the application of multilevel analysis to longitudinal data, where the nesting structure is measurement occasions within subjects. A more recent development is multilevel structural equation modeling, which allows more complex modeling than multiple regression does. At the same time, multilevel models have been extended to allow non-normal dependent variables, such as dichotomous or ordered categorical variables. This presentation aims to give an overview of current multilevel techniques available to applied researchers.
Presented by:
Joop Hox, Utrecht University
Date & time:
February 21, 2011 4:00 pm
External seminars home