I am testing various levels of measurement invariance across groups with dichotomous factor indicators. I have read most of the discussion on this subject in this list, Webnote #4, and Millsap & Yun-Tein 2004 article. I still need some assistance.
My questions are:
1) Is it possible to test a multigroup CFA model with factor loadings and thresholds estimated freely across all groups [using thata parametrization]?
2) If so, what are the constraints needed to identify the model?
I would appreciate it if anyone in the group can share complete examples of multigroup CFA models (testing all levels of invariance) with dichtomous indicators.