Hi David
How many first order factors do you have?
If you have only three, then they will be identical, because you are
trying to explain 3 correlations between latents with three loadings.
What do you mean by "almost identical"?
Jeremy
2008/5/3 Moore, David <[log in to unmask]>:
> I'm trying to run a CFA on a 16 item scale which has been suggested to have either 3 or 4 factor. Some recent research has suggested that these lower order factors load onto a single higher order factor. I have tried to run the analysis with the lower order factors correlated and then with them all loading onto a single higher order latent variable. What I have found though is that the obleque rotation and the use of the higher order latent variable give very similar if not the same output when I do this. I am quite new to SEM and would really appriciate any advice anyone can give.
>
> Best
> Dave
>
--
Jeremy Miles
Learning statistics blog: www.jeremymiles.co.uk/learningstats
Psychology Research Methods Wiki: www.researchmethodsinpsychology.com
|