annonymous posted on Friday, December 30, 2005 - 12:51 pm
I have a conceptual question. I have a theory that suggests three measures make up latent variable x1 and four measures make up x2. However, some theorize that the measures of x1 and x2 make up a second-order factor of x3. All that we can say for the second-order factor is that x1 and x2 correlate, but they do not interact. This is a correct assumption?
I'm not sure what you mean. A model for a second-order factor with only two indicators is not identified, so you could only look at the correlation between the two first-order factors not test a second-order factor model. Is this what you mean?