|
|
 |
| Alignment and cross loadings |
 
|
| Message/Author |
|
|
| Olev Must posted on Wednesday, September 27, 2017 - 12:19 am
|
|
|
I need to compare the test scores over 2 time points.I have binary items and models use cross loadings of items (WLSMV method). In some subtests I obtained the estimation of the measurement invariance via classical MGCFA. In some subtests I had problems with model convergence (scalar invariance) although the configural invariance was good. I tried to use alignment opimization, but I got the following feedback:
Cross loadings are not allowed for invariance analysis with ALIGNMENT=FREE or ALIGNMENT=FIXED.
Do you have suggestions how I should continue my modelling?
Thank you!
Olev |
|
|
|
|
Note that if you have 2 time points, the subjects need to be different at the 2 time points in order to use MGCFA and Alignment.
For analysis strategies, try SEMNET. |
|
|
| Olev Must posted on Wednesday, September 27, 2017 - 11:42 pm
|
|
|
Thank you for suggestion!
Yes, the subjects are different: different age cohorts.
I shall try SEMNET possibilities. |
|
|
| Back to top |
|
|
