So I've been trying to build a second-order latent growth curve model in the R package lavaan, with four observed indicators each measured at six time points, six latent variables representing the repeated measures (of 'physical capability') and an underlying intercept and slope structure. I managed to get it working if I pretend that I have continuous observed variables, but I really don't. Trying to reformulate the model with WLSMV and ordinal variables just gives a bunch of errors, so I think I should start by fully understanding what I'm doing.
Does anyone know about any good literature/web pages/tutorial videos that explain how to get a working second-order latent growth curve model with discrete observed variables in lavaan? I can't find any source that shows how to do this specific thing.
Thanks in advance for any tips!
Check https://lavaan.ugent.be/tutorial/tutorial.pdf
https://cran.r-project.org/web/packages/lavaan/lavaan.pdf
Also check https://stats.idre.ucla.edu/r/seminars/lgm/
https://quantdev.ssri.psu.edu/sites/qdev/files/GMCh14_Common_Factor.html
Kindly check on this also
https://stats.idre.ucla.edu/r/seminars/lgm/
Thank you for your replies!
OK I think I should definitely take the seminar by Johnny Lin, as you both recommended it. From the looks of it, it will surely help me better understand some of the core modelling concepts.
It doesn't seem that he does something like a second-order model though, right (I may be wrong as I'm judging solely from the course outline)?
The chapter that Chinaza Godswill Awuchi recommended is from a book I also found and found interesting, though this chapter is on continuous outcome variables. Their chapter 15 is actually precisely on second-order models with ordinal outcomes, but it is modelled only in Mplus. For this reason I was wondering whether there is a similar source for lavaan, as that might help for understanding why I keep getting specific error messages.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Graphs