Only open to Master’s and Research Master’s students from Psychology.
Students who have not attended the “Introduction in R and Statistical Computing” course should spend a few hours, before the course starts, by: 1) Studying and working through the examples and exercises of the first chapter of Beaujean’s book. 2) Going through the examples in the introduction to R on http://data.princeton.edu/R/default.html.
Psychological characteristics cannot be measured directly, like temperature or distances. They are latent variables, which can only be measured indirectly through, for example, items of tests or questionnaires. These item responses are used as a measure, or indicator, of the psychological characteristic (construct) of interest. Using latent variable models, we can assess how well these latent variables are measured, how they change over time, and/or how they are associated with other (latent or directly observed) variables.
Two widely known frameworks for latent variable modeling are Structural Equation Modeling (SEM) and Item Response Theory (IRT). Traditionally, IRT models were often used for analysis of dichotomous and (ordered) categorical item responses. SEM models have often been used for analyses such as path analysis, confirmatory factor analysis and latent growth modeling.
This course offers a theoretical and practical introduction to SEM and IRT models. Several latent variable modeling applications will be discussed, including path analysis, confirmatory factor analysis, IRT modeling, measurement invariance (‘differential item functioning’) and latent growth curve modeling.
This course focuses on translating substantial theories into latent variable models. Students acquire basic skills in using R and the lavaan package for SEM: they are taught how to perform latent variable analyses and how to interpret the results. Applications from psychology will be used to illustrate the methods.
Students will acquire basic knowledge of :
- The distinction between the structural and the measurement model
- Identification of the scale of latent variables in SEM and IRT
- Principles of IRT, and how they can be translated into a SEM model
- How IRT and CFA compares to classical test theory
- Estimation and calculation of path coefficients
Students will acquire basic skills in:
- Fitting and interpreting IRT models
- Fitting and interpreting basic and hierarchical CFA models
- Fitting and interpreting models with multiple groups or measurement occasions
- Testing interaction and indirect effects (moderation and mediation)
Mode of instruction
Lectures will be combined with computer assignments in seven four-hour meetings. The meetings will cover both latent variable modeling theory, as well as practical applications and computing in R.
In the course, we will be using the R-package ‘lavaan’ (short for LAtent VAriable ANalysis). Please bring your laptop to the course with R installed.
One structured assignment (50%), and one unstructured assignment (50%). The unstructured assignment consists of performing and reporting on an advanced analysis involving latent variable(s), on a dataset of a student’s own choosing, or supplied by the lecturer.
- Beaujean, A. A. (2014). Latent variable modeling using R: A step by step guide. New York, NY: Routledge/Taylor and Francis.
- Additional course material to be announced on Blackboard.
Dr. Marjolein Fokkema