I have stored the estimates for a 2-class, 3-class, 4-class and 5-class model. Now I would like to compare these models. I know i can use estimates stats to get the AIC and BIC for all these models in a nice table, but it should be possible to do a LR test to see if the models with more classes add anything right?

But when i do lrtest twoclass threeclass, then I get an error saying that the dependent variables differ. I have seen the lrtest being used to compare models with the same number of classes but differing restrictions, but it should also be possible to compare models with a differing number of classes right?