The application deadline has passed, applying is no longer possible
Complicated issues such as estimating and testing are explained intuitively with clear instructions. Interpreting results in SEM receives extensive attention.
The course begins with an intuitive approach to structural equation modeling (SEM) where you learn what SEM entails. You will learn about estimation procedures & assumptions, model evaluation & testing, in an intuitive way, with as little statistics as possible.
Then you'll learn about measurement and factor analysis. Measurement is the basis of any empirical research, but how do you know if your measurement scales are good (enough)? You will learn how to evaluate (and develop) measurement scales using factor analysis. The main advantage of SEM is that relationships between latent variables can be analyzed, reducing the influence of a variety of measurement problems. Models that include latent variables are preferable, but not always feasible. You will learn how to simplify an "ideal" model and understand how simplification affects the conclusions you can draw. Next, you will learn about causal modeling. The topic of causality is often avoided in courses on SEM, we don't. You will learn to develop causal models and test them in a (non-)experimental context. Once you see how causal modeling works, you will understand the importance of study design. In this context, you'll learn how to test for mediation and moderation (including using multigroup analysis).
Finally, we briefly discuss some advanced techniques, such as measurement invariance, cross-lagged panel models, post-hoc power analysis, and dealing with non-normal data (e.g., ordinal data or dichotomous data).
In this course we will use R-Studio and the package lavane (both free).
