|
Message/Author |
|
|
I am hoping to run a growth mixture model with three classes in which (1) the intercept and slope factors, along with their variances, are all freely estimated and not constrained to equality across classes; and (2) one of the loadings on the latent growth factor is freely estimated in each class and not constrained across classes. With respect to issue (1), I believe the following MODEL specifications will take care of the first set of requirements: MODEL: %OVERALL% i s | W1RS@0 W2RS W3RS@1; %c#1% i s; %c#2% i s; %c#3% i s; With respect to issue (2), however, I am uncertain as to how to freely estimate the middle loading in the growth factor within and between classes (i.e., unconstrained to equality across classes). |
|
|
You can try s BY W2RS*; Check in Tech1 that you get what you want. |
|
|
Excellent. Thank you for the quick response. Your suggested approach provides exactly what I was looking for from the analysis. |
|
Back to top |
|
|