I am trying to do a multigroup CFA with dichotomous indicators. I am having difficulty in getting my models identified or converge to a solution. I am not sure the examples (5.16 & 5.17) in the user's guide are appropriate in my case.
I would like your help in specifiying the constraints needed for a multigroup ONE FACTOR CFA model with dichotomous variables [as is in F BY X1 X2 X3; with all x's having only two cateogies].
I'm fitting one dimension multigroup CFA model with categorical data (using Mplus 7.11). The number of groups is 34. I'v estimated model both with ML and WLSMV without problems. When I use WLSMV everything works fine except modindices. Although I've specified option MODINDICES(0) for some groups I do not get modindicies for some intercepts and factor loadings. In one group I've all of them in other some of them are missing. What can cause this situation? Might I assume that they are negligible while they are not reported?