An ever-increasing number of software packages are implementing multilevel approaches to SEMs. While computationally, analytically and structurally complex, these methods are effectively implemented in a user-friendly way in R and Mplus (also Stata).
In this course we will go though the most common applications of MSEMs with examples in R and Mplus.
Mplus has several advantages (user friendly, flexible) but clear disadvantages (expensive, closed). Developers of the R package for SEM: Lavaan also started to implement MSEM applications. Significantly less flexible than Mplus, most models in the book we use for the course can be estimated with R using the Lavaan package. All examples used in the course will run in the free demo version of Mplus.
Please bring your laptop with R + Lavaan and/or Mplus (Demo) installed. R and Mplus are compatible with Windows, Mac and Linux. Android and iOS are inadequate for the purposes of this course but they will run on Project Crostini compatible Chromebooks in the Linux sandbox.
Monday We spend some time getting to know each other and reviewing the most important components of MLM and SEM. We go into notation to define our MSEMs (may reach into Tuesday).
Tuesday We start with multilevel path analysis.
Wednesday We cover multilevel confirmatory factor analysis.
Thursday We discuss full structural models in a multilevel framework.
Friday We discuss advanced topics and the various conceptualisations for defining MSEMs.
The course, for the most part, is focused on theory and the understanding of the models at hand. We will review applications and arm you with the tools to run your own models, but the course will not be software focused for the most part. You are coming here to learn MSEM, not just how to run models in one software or another. (My teaching philosophy is, once you know what you are doing, software is easy. If you don’t know what you are doing and want to compensate by learning more software without proper foundations, you are doing it wrong anyway.)