In various disciplines of quantitative empirical social research, models with latent variables (latent variable models [LVM]) play a central role in the analysis of social science constructs (e.g. abilities, characteristics or attitudes of individuals). In most cases, the measurement assumes that the observed behaviour in a specific situation (manifest variable, e.g. answers in a test) can be used to draw conclusions about generalisable, more broadly defined, not directly observable constructs (latent variable, e.g. underlying ability).
The course provides an overview of LMVs for analysing relationships between latent variables and latent or manifest variables: the various model classes (structural equation models, factor analyses, latent profile analyses, latent class analyses, item response theory) are introduced and discussed using empirical examples. Possible challenges in modelling and parameter estimation are addressed and the test theory concepts relevant to LMVs are discussed. Using sample data, the basics of modelling using LMVs are taught and practised.
The course is aimed at people with a basic knowledge of statistics and test theory (from their bachelor's degree). Participants should be familiar with the basic principles of regression analysis and (exploratory) factor analysis, as well as those of classical test theory. A basic knowledge of the data analysis software R is also required.
In the first session (see dates), the course schedule will be discussed and the participants' prior knowledge will be assessed. There is the option of setting specific focal points on the block days, which will also be discussed in the first session.

Course participants will learn about the model classes for LVMs and will be able to identify use cases for different LVMs and analyse data using LVMs.

Further information on teaching evaluation: https://www.leuphana.de/en/teaching/quality-management/evaluation/course-evaluation.html