This ATI is designed to highlight recent methodological advances in the analysis of longitudinal psychological data using structural equation modeling (SEM). The training is intended for faculty, postdocs and advanced graduate students who are familiar with SEM (e.g., took an introductory SEM course). The workshop covers a range of topics, including growth models, factorial invariance, dealing with incomplete data, growth mixture models, ordinal outcomes, and latent change score models (see a tentative course schedule).
Course materials include basic readings on the fundamental theoretical issues in contemporary longitudinal data analysis, lecture notes and computer scripts for commonly used SEM programs. The workshop alternates between lectures on theory and specification of models using Mplus, and lab sections, which review the specification of the same models in lavaan (available through R) and AMOS, and include time to fit models to your data. Participants are strongly encouraged to bring their own data and research problems, and a notebook computer equipped with SEM and general statistical (R, SAS, SPSS) software.
Course instructors will include Craig Enders of the Department of Psychology at The University of California, Los Angeles, Kevin J. Grimm of the Department of Psychology at Arizona State University, and Nilam Ram of the Departments of Human Development and Family Studies and Psychology at The Pennsylvania State University.
Applications are invited from investigators at the faculty/professional, postdoctoral and advanced graduate student levels. The ATI is open to investigators from both within and outside of the United States.
The language of the course is English; no translation services will be provided.
The Advanced Training Institute on Structural Equation Modeling in Longitudinal Research will be held remotely. To provide maximum flexibility for participants, lecture presentations will be pre-recorded and made available prior to the workshop. For each lecture, the instructors will host, via video conference, a half-hour online question and answer (Q&A) session and a one-hour lab session that covers code and implementation details.
We encourage participants to watch the lectures prior to the live sessions and be ready with their questions. During the lab sessions, we encourage participants to apply the longitudinal models reviewed in the lectures to analyze their own data or exemplar longitudinal data provided by instructors, and to ask questions about the programming and interpretation of the modeling results.
Day 1: Introduction to Structural Equation Modeling & Longitudinal Data
- A: Introduction to SEM with longitudinal data
- B: Q & A – Introduction to SEM with longitudinal data
- C: Lab – Computer programming with longitudinal data
- D: Alternative models for two-occasion longitudinal data
- E: Q & A – Alternative models for two-occasion longitudinal data
- F: Lab – Computer programming for two-occasion models
Day 2: Latent Growth Models in SEM
- G: Latent growth models and time-invariant covariates
- H: Q & A – Latent growth models and time-invariant covariates
- I: Lab – Computer programming for growth models
- J: Approaches to handling incomplete longitudinal data
- K: Q & A – Approaches to handling incomplete longitudinal data
- L: Lab – Computer programming for incomplete longitudinal data
Day 3: Modeling Change with Latent Variable Models
- M: Modeling change using latent variable factor models
- N: Q & A – Modeling change using latent variable factor models
- O: Lab – Computer programming for longitudinal factor models
- P: Modeling change based on binary and ordinal outcomes
- Q: Q & A – Modeling change based on binary and ordinal outcomes
- R: Lab – Computer programming for longitudinal binary/ordinal data
Day 4: Multiple Groups & Nonlinearity
- S: Multiple group growth models & growth mixture models
- T: Q & A – Multiple group growth models & growth mixture models
- U: Lab – Computer programming for multiple group growth models
- V: Nonlinearity in growth models
- W: Q & A – Nonlinearity in growth models
- X: Lab – Computer programming for nonlinear growth models
Day 5: Latent Change Scores & Multivariate Dynamics
- Y: Univariate dynamics based on latent change scores
- Z: Multivariate dynamics based on latent change scores
- AA: Q & A – Latent change score models