| Message/Author |
|
|
| HALUS posted on Tuesday, June 16, 2009 - 6:29 am
|
|
|
Hi,
Not sure if Mplus set factor loadings of first- and second-order factor model equal as default. If not, how do I impose those equalities myself?
thanks! |
|
|
|
|
The second-order factor loadings are not held equal as the default. You use equality labels as follows:
f4 BY f1 f2 (1)
f3 (2); |
|
|
| HALUS posted on Tuesday, June 16, 2009 - 8:00 am
|
|
|
| So does this mean that I use the equality labels you suggested for my baseline group as well as the comparison groups? |
|
|
|
|
| If you put the equalities in the overall MODEL command, they apply to both groups. Please see the discussion of multiple group analysis in Chapter 13 of the Mplus User's Guide. |
|
|
|
|
Hello,
I am trying to fit a multi-group CFA on 14-item scale (with three dimensions- subscales). I am however confused about what is the best way to make the metric invariance converge? I want to fix the coefficients but in order to make the model converge I have to fix more than just the coefficients. I have found two methods:
1.fixing also the means of the subscales 2.fixing the first item’s (marker) intercept and error variance instead of the means
What is the correct approach, or what is the impact of these two different methods on my results? I can see the RMSEA of the model under method 2. is worse than method 1.
Thank you for help!
Annamaria |
|
|
|
|
| Please send the output where you had the original convergence problems and your license number to support@statmodel.com. Make sure the model fits in each group as a first step. |
|
|
| Back to top |
