Variance explained by 2nd order factor
Message/Author
 André Beauducel posted on Thursday, February 27, 2014 - 5:50 am
Hello,

I want to calculate the variance explained by a 2nd order CFA factor:
Is it correct to calculate the mean square of the betas (factor and variable variances are 1) of the total indirect effects of the 2nd order factor on the observed variables?
(How) can I get a significance test for the variance explained?

Many thanks!
 Linda K. Muthen posted on Thursday, February 27, 2014 - 10:41 am
The STANDARDIZED option gives the variance of each first-order factor explained by the second-order factor. The R-square has a significance level.
 André Beauducel posted on Thursday, February 27, 2014 - 12:11 pm
Thanks!
However, I meant the overall amount of variance that the second-order factor explains in the solution. I have seen the R-squares for the primaries, but I would like to get the overall for the secondary (with significance level).
 Bengt O. Muthen posted on Thursday, February 27, 2014 - 12:19 pm
I don't know what you mean.
 André Beauducel posted on Thursday, February 27, 2014 - 11:35 pm
...I mean just the overall variance that is accounted for by the 2nd order factor.

I could compute the sum (or mean) of the squared loadings of the primary factors on the 2nd order factor (like in exploratory FA). However, is it possible to get a significance for this?
 Linda K. Muthen posted on Friday, February 28, 2014 - 10:18 am
You can use MODEL CONSTRAINT to compute this and will get a standard error.
 André Beauducel posted on Saturday, March 01, 2014 - 11:58 am
That's it! Thanks a lot!!!