|
|
 |
| MIMIC vs. Multi-Group Model |
 
|
| Message/Author |
|
|
|
|
Hello,
in multigroup modeling, invariance of factor loadings and item intercepts often is recommended before latent factor means can be compared.
However, couldn't I also compare the latent means by using a mimic model, where the latent factor is regressed on the grouping variable using this simple syntax:
f by x1 x2 x3 x4;! measurement model
f ON Z; ! structural model, z is !dichotomous grouping variable
Question #1: Shouldn't the unstandardized effect of the grouping variable be identical to the difference of the latent means?
Question#2: And would'nt the MIMIC model shown above allow me to circumvent testing for item intercept differences, if I were to compare the relative differences in the means?
Many thanks for your reply,
Tino |
|
|
|
|
| You may want to ask this general analysis question on SEMNET. |
|
|
| Back to top |
|
|
