finnigan posted on Tuesday, March 20, 2012 - 9:51 am
I am running a CFA using survey items.Is there a way in MPLUS to check linearity and multicollinearity assumptions?. In regression analysis this is straight forward,but I'm not sure about how this is done using the observed survey items.
I don't think linearity is checked for factors. You could add a squared term as is shown in Example 5.7. As far as multicollinearity, ask for TECH4 in the OUTPUT command and look for high correlations among the latent factors.