Can you help me to determine how to best test for structural invariance when using complex survey data? When I have attempted to run my models using the "VARIABLES: GROUPING" option, I received an error message that "The SUBPOPULATION option is not available for multiple group analysis."
Given that the USEOBSERVATIONS option will result in improper standard error estimates when used with complex survey data, do you have any suggestions on testing for structural invariance?
For each group, analyze the data using each option, USEOBSERVATIONS and SUBPOPULATION. I think you will see little difference in the results. If this is the case, use USEOBSERVATIONS for the multiple group analysis.