Factor scores change when adding 2nd ...
Message/Author
 Jan Ivanouw posted on Thursday, March 01, 2012 - 12:55 am
Hi,

I hope you can help me with the following problem:

I have a simple model with 4 factors, each measured by 8 categorical items (each with 5 categories).
In a second model I add two 2nd factors, each from two of the 1st order factors.

My problem is that the factor scores for the four 1st order factors are not the same in these two models.

1. Why is that?
2. Can the 1st order factor scores in the second model be understood as simple transformations of the factor scores from the first model (e.g. with some adjustment for the 2nd order factor scores)?

Thank you
 Bengt O. Muthen posted on Thursday, March 01, 2012 - 6:28 am
Doesn't your second model have one parameter less?
 Jan Ivanouw posted on Friday, March 02, 2012 - 12:51 am
The number of free parameters in the first model (without 2nd order factors) are 157
Number of free parameters in the second model with 2nd order factors are 156.
 Bengt O. Muthen posted on Friday, March 02, 2012 - 9:03 am
This means that the two models are different and therefore the factor scores will be different.