Latent Class Analysis
With Open-Source Software in R
General Information
| Course name | Latent Class Analysis with Open-Source Software in R |
| Instruction language | English |
| Type of instruction | Lectures and lab sessions |
| Level | Graduate School of Psychology, University of Basel |
| Course load | 8 hrs |
| Lectures | Dr. C. J. Van Lissa |
Are there meaningful subgroups in your data that conventional analyses miss? Latent class analysis (LCA) can identify subgroups in your data that differ qualitatively in their response patterns. You may know it by other names, such as mixture modeling, latent class growth analysis, or hidden Markov modeling.
Relevance
This workshop is designed for PhD students who want to use person-centered methods to answer substantive research questions about unobserved groups. For example, you can explore whether your population is homogeneous or heterogeneous; test theories about categorical latent variables, like identity status theory or developmental stage theory; estimate people’s unknown group membership - like whether they are at risk for a clinical disorder, are in a filter bubble, or have become radicalized. Group membership can be based on simple descriptive statistics (e.g., means on questionnaire items), or on more complex models (e.g., developmental trajectory over time).
Requirements
The workshop is suitable for a wide range of experience levels: complete beginners in R, researchers who already use R but are new to LCA, and advanced users who want to transition to open-source software or discuss current best practices.
Content
Thursday September 17th, 14:00 to 18:00
You will learn the core ideas behind LCA and get hands-on experience with user-friendly R functions for estimating latent class models with continuous, ordinal, and dichotomous indicators, and combinations of these. We will focus on the practical decisions researchers face: evaluating model fit, assessing classification accuracy, interpreting model parameters, and reporting results clearly with publication-ready tables and figures.
Friday September 18th, 9:00 to 13:00
Several advanced topics are introduced, and students choose two to focus on:
- Confirmatory LCA: testing hypotheses about the number of classes
- Auxiliary variables: examining differences between classes on external variables
- Latent class growth analysis
- Custom model specification for LCA
Practicalities
You will need to complete the setup tutorial before the workshop, so we spend our time together productively: Appendix A — Setup your Computer. All exercises are supported by tutorial vignettes and fully reproducible R code that you can adapt to your own research. You are encouraged to bring your own data, so you can directly apply what you learn to your dissertation or ongoing projects. Exercises are completed in pairs, creating space for discussion, feedback, and shared problem-solving.
Course materials
You do not need a book for this course. This GitBook provides all relevant course materials, along with primary literature linked in the GitBook.
Staff
Software
The official software for this course is R, which you should install on your laptop as explained in Appendix A — Setup your Computer.