The main aim of this interavtive tutorial is to show how different univariate and bivariate latent change score models (lcsm) can be implemented in R using lavaan syntax (Rosseel, 2012).
At the moment it is possible to:
- Generate lavaan Syntax for different model specifications and varying time points
- Simulate Data to explore the effect of different parameters
- Fit Models using example datasets
- Create Longitudinal Plots and simplified Path Diagrams to visualise data and model specifications
I started working on this project to better understand how latent change score modeling works. The underlying functions of this shiny application come from the lcsm R package that I created to make it easier to write lavaan syntax for different models. The lcsm package combines the strengths of other R packages like lavaan, broom, and semPlot by generating lavaan syntax that helps these packages work together.
This is work in progress and feedback is very welcome. The code of the R package lcsm and this shiny application can be found on GitHub. Please send me your thoughts on Twitter, GitHub, or by email.
You can install the development version from GitHub with:
# install.packages("devtools")
devtools::install_github("milanwiedemann/lcsm")For details about this methods see for example McArdle (2009), Grimm et al. (2012), or Usami et al. (2019).
Examples illustrating how to implement different latent change score models in R can be found for example in: Ghisletta & McArdle (2012), Grimm, Ram & Estabrook (2017), Kievit et al., (2018) including a shiny interface, and Jacobucci et al. (2019).
Online tutorials from the Quantitative Developmental Systems Methodology Core by Xiao Yang and Miriam Brinberg show examples how to specify lavaan syntax for univariate and bivariate latent change score models described in Grimm, Ram & Estabrook (2017).