

Multigroup Multilevel CFA 

Message/Author 


Hi, I am trying to run a multigroup multilevel CFA analysis to test if the between level factor variances are the same across groups. I get the following error message: THE STANDARD ERRORS OF THE MODEL PARAMETER ESTIMATES COULD NOT BE COMPUTED. THE MODEL MAY NOT BE IDENTIFIED. CHECK YOUR MODEL. PROBLEM INVOLVING PARAMETER 44. THE CONDITION NUMBER IS 0.771D16. When i run the model separately for each group is does just fine, so i can't figure out the problem. Any suggestions? 


It sounds like you are freeing the first factor loading in the multiple group analysis. You should not mention the factor loading that is fixed to one in the groupspecific MODEL commands. 


Hi again, actually I have all the first order loadings fixed to zero, and they are not mentioned in the second and third model statements. However, I have second order factors as well, and the loadings onto these second order factors are also fixed to 1...but I mentioned them in the groupspecific model commands because I assumed they'd be treated as structural paths, but maybe they are not, and shouldn't be mentioned? In sum, I want the models to be the same across groups except for the first order indicator error variances and the second order factor variances. The only other thing I can think of is that perhaps some factor intercepts need to be released, but I thought they were free by default. 


Please send the output and your license number to support@statmodel.com. It is impossible to understand what you are doing without seeing the full picture. 

Back to top 

