Continuous Time Structural Equation Modeling (CTSEM)
Abstract
In the last 10 years, continuous time structural equation modeling (CTSEM) has made the possibilities for analyzing longitudinal data considerably more flexible. Among other things, CTSEM allows the analysis of data sets with intraindividually and interindividually varying time intervals, longitudinal moderator analyses, and provides the basis for meta-analysis of longitudinal primary studies. Diary data can serve as a basis as well as data from classic longitudinal designs with two or more measurement time points. Participants can bring their own data – it is recommended to get in touch beforehand (cdormann@uni-mainz.de). The workshop is primarily application-oriented. The theoretical introduction is limited to about 45min. For mathematical details, I can carry forward participants to Voelkle et al. (2012) or Driver and Voelkle (2021). It should be noted that the procedures described there (for using the R package ctsem) do not exactly correspond to the current state.
Prerequisites:
• Installation of R and R-Studio on your own notebook computer, but no R knowledge necessary.
• Basic (is sufficient) knowledge of SEM. Knowledge of multi-level modeling is beneficial, but not required.
Optional:
• Own diary data (ideally – but not necessarily – with varying survey times or time points) or longitudinal data (min. 2 waves with 2 variables, ideally > 2). Additional data sets are provided for exercises.
Workshop Programme:
Sep 26 10:00 – 10:45 Theoretical Foundations of CTSEM
Day 1 10:50 – 11:20 Conceptual considerations: when, how frequently, and what to measure.
11:30 – 12:00 Preparing data for ctsem (long format, wide format, missing values, time
points vs. time intervals).
14:00 – 15:00 Showcase: A simple but wildly applicable ctsem model
15.10 – 16.00 Model setup and interpretation of output.
Sep 27 10:00 – 10:30 Plotting effect trajectories across time.
Day 2 10:35 – 11:25 Moderator models Part 1 (continuous moderators).
11.30 – 12.00 Moderator models Part 2 (categorical moderators).
14:00 – 15:00 Advanced Models: Fixed and random intercept/slopes.
15.00 – 16.00 Exercises and Q&A.
Speaker Biography
Professor Christian Dormann is currently chair of business education & management at the faculty of economics & law of the Johannes Gutenberg-University Mainz, Germany, and adjunct research professor at the School of Psychology at the University of South Australia. Earlier, he was chair or business psychology at the University of Bochum and chair of work, organizational, and business psychology at the university of Mainz. He served as editor-in-chief of the European Journal of Work and Organizational Psychology and as associate and consulting editor of several other journals, including the Journal of Occupational and Organizational Psychology, Journal of Organizational Behavior, and Journal of Occupational Health Psychology. His research focus is on stress in organizations with special focus on methodological considerations including moderating effects, continuous time modeling, and meta-analysis. Most recently, he and his co-authors were awarded with the 2022 Schmidt-Hunter Meta-Analysis Award for a longitudinal meta-analysis of stressor-burnout effects.