I'm working on a multigroup cross-classified IRT model with 1 general factor, and the items are cross-classified by 6 content areas and 5 scenarios. Due to the large number of latent factors, I used integration = montecarlo in the estimation. I requested FSCORES in SAVEDATA as usual, but it does not give me the standard errors of the factor scores. I'm wondering if it's due to the montecarlo integration or it just doesn't provide standard errors in this situation?
Montecarlo integration is not the problem. The only case I am aware of that standard errors are not reported is if the factor variances are zero. If this is not the case send your example to firstname.lastname@example.org.