|
|
Model identification in twolevel CFA/SEM |
|
Message/Author |
|
Student 09 posted on Friday, June 19, 2009 - 1:56 am
|
|
|
Hello I wonder how twolevel SEM's are identified in Mplus. a) Is it correct that in Mplus, the number of elements on the between-level equals (p(p+1)/2) + intercepts, where p are the observed indicators, just as in 'regular' covariance and mean structure analysis? b) Is it correct that on the within level, identification follows (p(p+1)/2), that is without the mean (intercept) information, because the intercepts are modelled only on the between-level? Many thanks! |
|
|
a) Yes. b) Yes. |
|
Student 09 posted on Friday, June 19, 2009 - 10:58 am
|
|
|
so this implies that the covariance structure of a twolevel model can be identified, while the mean structure on the between level is not identified? Thanks for your response |
|
|
No, it doesn't imply that. The sample means identify the p means in the model (which appear on Between). |
|
Back to top |
|
|