GESIS Training Courses

Scientific Coordination

Alisa Remizova

Administrative Coordination

Claudia O'Donovan-Bellante
Tel: +49 621 1246-221

Introduction to Longitudinal Structural Equation Modeling

Mannheim B6, 4-5
General Topics:
Course Level:
Software used:
Students: 220 €
Academics: 330 €
Commercial: 660 €
Additional links
Lecturer(s): Daniel Seddig

About the lecturer - Daniel Seddig

Course description

The workshop provides an introduction and overview of analyzing longitudinal (panel) data using structural equation modeling (SEM). SEM is a flexible and versatile suite for data analysis, allowing researchers to simultaneously test complex relationships among variables, model latent constructs, assess measurement validity, and account for measurement error within a single statistical framework.
In the first part, the workshop covers the fundamental concepts of path analysis (PA) and explores how to specify, estimate, and interpret various longitudinal models. These models are, for example, autoregressive models, fixed- and random-effects models, cross-lagged models, and latent growth curve models.
In the second part, the workshop extends these models to scenarios where the variables of interest are latent (not directly observed) and measured by multiple (directly observed) indicators. Therefore, confirmatory factor analysis (CFA) procedures are discussed and adapted for longitudinal data, emphasizing the specification of error correlations across time and testing for measurement invariance across time.
Special topics covered in the workshop include, for example, handling non-normal and categorical data, conducting multiple group analysis, and employing Bayesian estimation methods. Theoretical concepts are introduced alongside practical demonstrations using social science example data in the R programming environment, particularly with the lavaan package. Participants engage in exercises utilizing the example data and have the option to apply these methods to their own data for practical application and learning.
Organizational structure of the course
The two days are divided into lecture sessions and exercise sessions. The instructor prepares the materials for both. It is possible to discuss questions of participants regarding the procedures presented in the lectures and exercises as well as the participants' own projects focusing on conceptual and analytical problems. Participants are expected to participate in discussions on the topics of the workshop, engage in the exercises, and read the workshop literature. Participants are also encouraged to work on their own projects and analyze their own data individually or in groups.

Target group

Participants will find the course useful if:
  • they are interested in assessing and explaining change over time;
  • they are interested in longitudinal relationships between different (observed and latent) variables over time;
  • they are interested in conducting comparisons of (latent) variables over time and across groups.

Learning objectives

By the end of the course, participants will:
  • know how to specify structural equation models for longitudinal data with observed and latent variables;
  • know how to test for measurement invariance of latent variables across time;
  • be able to specify, estimate, and interpret the models in the R programming environment.


  • Basic knowledge of and basic experience with regression analysis and factor analysis is mandatory.
  • Basic knowledge of and basic experience with the R programming environment is mandatory.
  • Basic knowledge of and basic experience with longitudinal data analysis, path analysis, confirmatory factor analysis, and structural equation modeling is an advantage but not mandatory.
Software and hardware requirements
To be able to follow the script and execute the exercises, workshop participants need a computer with the latest versions of R and RStudio installed. The installation files for R for Windows can be found here and for Mac here, while the current installation files for RStudio can be found here.
The main R package used in this workshop is lavaan. Other useful packages are semTools, semPlot, descr, psych, and ggplot2.


Recommended readings