|
|
Latent-basis GCM with individually-va... |
|
Message/Author |
|
|
Hello, I came across a paper that shows a method to estimate a shape-factor growth curve model (or latent-basis or fully-latent GCM or whatever it's called, where the loadings are estimated from the data) with individually-varying times of observation. Here's the part where the author describes the matrix of factor loadings: Matrix of factor loadings with individually-varying times of observation And here's the corresponding Mplus syntax: Corresponding Mplus syntax I was surprised to see the time variables (with time values that vary from person to person) in the constraint section, and I was also surprised that this method to accommodate individually-varying times of observation does not use the TSCORES=... option in conjunction with "AT" in the MODEL command. So I wanted to check in: is this indeed doing what the paper says it's doing (i.e., using different factor-loading matrices for each person)? |
|
|
The best source for this is the new book Growth Modeling by Grimm, Ram, Estabrook. |
|
Back to top |
|
|