Longitudinal confirmatory factor analysis and measurement invariance testing
Latent change score, autoregressive, and growth curve models
LST models and trait-change models
Dr. Christian Geiser is a Professor of Quantitative Psychology at Utah State University, author of two books on Mplus, and a leader in the development of S.E.M. techniques for complex data. With his accessible books and sought-after workshops, he has helped thousands of researchers and students around the world to achieve their analytic goals. Longitudinal S.E.M. with Mplus is the first in a series of companion workshops to his book of the same name.
I don't have access to Mplus. Can I still benefit from this course?
Absolutely. You can engage in the full curriculum using only the demo version of Mplus, which is FREE. Dr. Geiser will provide instructions about getting the demo installed on your computer.
What level of statistical training is required for the course?
This course is designed for learners with a working understanding of structural equation modeling and basic experience with Mplus. Need a beginner-focused course? Mplus from Scratch gives you everything you need to be ready for Longitudinal S.E.M. with Mplus.
There is no way that I will be able to set aside the three days required to take this course. Is the timing flexible?
We understand completely! The timing is entirely up to you. Once you enroll and the course has launched, you will have unlimited access to the content. You are free to pause and return to your lessons as needed.
Will you give me feedback on my analyses?
Unfortunately, personal feedback is not possible given the volume of researchers we serve. We are certain that Dr. Geiser's thorough approach to guiding you through every step of using Mplus, along with the helpful handouts and guides included in the course, will enable you to successfully perform longitudinal SEM with your own data. Note that we are always happy to help you navigate any technical issues related to the course itself.
Longitudinal S.E.M. with Mplus uses Christian Geiser's video-based instruction in combination with associated datasets, syntax, and a workbook to form a solid foundation for performing LST-based longitudinal analyses. The course is broken into 12 sessions that can be completed in about 3 days, though the timing in which you work through the course is entirely up to you. Keep scrolling to learn more about what to expect:
Session 1: Introduction to Latent State-Trait Theory
- History, notation, and basic concepts of latent state-trait theory
- Psychometric definition of latent state, trait, state residual, and measurement error variables
- Variance decomposition
- R2-type coefficients of consistency, occasion specificity, and reliability
Sessions 2 & 3: Longitudinal Confirmatory Factor Analysis and Measurement Equivalence Testing
- A basic model for longitudinal multiple-indicator data: the latent state model
- Theory and levels of measurement equivalence (aka measurement invariance)
- Structural (latent variable) equivalence across time
- Testing different levels of measurement and structural equivalence in Mplus
- Assessing latent mean, variance, and correlation differences across time
Session 4: Latent State Model With Indicator-Specific Factors
- Modeling indicator-specific (method effects) in longitudinal studies
- Specifying indicator-specific residual factors in Mplus
- Quantifying convergent validity, indicator specificity, and reliability
Sessions 5 & 6: Latent Change Score Models
- Modeling true inter-individual differences in intra-individual change
- Analyzing change as a latent variable
- Specifying and analyzing latent state difference score variables in Mplus
Session 7: Latent Autoregressive/Cross-Lagged Models
- Analyzing autoregressive effects
- Simplex and multiple-indicator autoregressive models
- Analyzing cross-lagged variable effects in multi-construct models
Session 8: Latent State-Trait Models
- Modeling variability and stability with models that contain both trait and state residual factors
- Separating trait effects from situation effects
- Determining consistency, occasion specificity, and reliability
- Different approaches for testing indicator-specific (method) effects
- Modeling autoregressive (carry-over) effects in latent state-trait models
Session 9 & 10: Latent growth curve models
- Random intercept model
- Linear, quadratic, and free growth models
- Visualizing individual growth trajectories in Mplus
- Multiple-indicator latent growth curve models
Session 11: Latent trait-change models
- Separating trait change from situation-specific (state residual) and error variance
- Analyzing latent trait change variables in Mplus
Session 12: How to Choose and Appropriate Model and Guidelines for Reporting Results
- Longitudinal modeling strategies
- Model selection
- What to include in a report of a longitudinal structural equation modeling analysis
QuantFish is dedicated to providing courses that improve your analytic skills with accessible lessons from the world's leading methodologists. Longitudinal S.E.M. with Mplus is backed by a complete money-back guarantee for 7 days following the start of the course.