Structural Equation Modeling in R using lavaan (E-Learning Course)

About

Structural equation modeling (SEM) is a powerful tool with which researchers can build intricate models of structural relationships and measurement processes. Traditionally, open-source software options for estimating structural equation models have been quite limited, but the lavaan project (https://lavaan.ugent.be/) has been steadily changing this narrative. The lavaan package is a completely free and open-source R package that implements a wide range of SEMs. The package is also being actively developed, so its capabilities continue to grow.

In this e-learning course, we will go over the basics of SEM using lavaan. We will first introduce lavaan syntax and explain the workflow of analyses using lavaan. We will then discuss how to implement common SEMs in lavaan. For example:

Path models
Latent regression models
Confirmatory factor analyses (CFAs)
Multiple-group CFAs
We will also consider issues that complicate the analysis and how these issues are addressed in lavaan. For example:

Missing data
Model selection
Categorical indicators
The course will be open for four weeks. During this time the participants can work through the material at their own pace. In addition to the scaffolded, self-paced instruction provided by the electronic course materials, participants will also have the opportunity to interact directly with lavaan experts via forum discussions and live, online Q&A sessions.

The Q&A sessions serve both as a time to discuss the course content and as consulting time. In these sessions, the instructors will answer participants’ questions about the course material and help participants apply the newly learned techniques to their own data.

Participants should install both R and RStudio (the free desktop version) on their computers before the beginning of the course.

R can be downloaded here.
RStudio can be downloaded here.

No prior experience with lavaan is required, but we will not cover basic R usage. We will focus primarily on how to implement various SEM analyses using lavaan. So, some prior experience with R is strongly encouraged. We will provide information on external resources that participants can use to brush-up on their R or SEM knowledge before following this course.

The substantive content of this course largely overlaps with S20: Introduction to Structural Equation Modeling using Mplus (with differential specializations in the advanced topics). These two courses can, therefore, be viewed as largely equivalent alternatives that target different audiences by employing different software (i.e., lavaan vs. Mplus) and instructional formats (i.e., online vs. on-location).

There is no direct follow-up to this course that covers more advanced analyses using lavaan, but participants who are interested in using SEM to analyze longitudinal data can follow S23: Advanced Longitudinal Modeling in Mplus. After completing this course, participants who wish to follow S23 will have sufficient statistical background knowledge, but they will need to review the corresponding Mplus usage/syntax.

In this course, we will adopt a psychometric perspective that approaches SEM as a way of combining path analysis and CFA to model complex, multivariate processes. We will not teach SEM as a method for estimating directed acyclic graphs (DAGs) to support causal inference. Therefore, this course is probably not applicable to anyone who is mainly interested in learning how to implement DAG-based causal analysis.

Course leader

Kyle M. Lang

Target group

Master and PhD students, PhDs, researchers, and other professionals who are interested in learning (more) about the lavaan project, how to use lavaan to estimate common structural equation models, and how they can apply lavaan in their own work. The course will be taught at a beginning master level.

We also offer tailor-made M&S courses and in-house M&S training. If you want to look at the possibilities, please contact Dr. Laurence Frank at pe.dsai@uu.nl.

Course aim

After completing this course, participants will be able to:

Compare and contrast the strengths/weaknesses of the lavaan software relative to other SEM packages
Translate their own analyses into lavaan models
Write R scripts to implement common SEM-based analyses using lavaan

Fee info