marginal effect of latent class logit model

[tina liang]

I am trying to obtain the marginal effect of the membership variable in latent class logit model [ATTACH=CONFIG]n1496017[/ATTACH]

(e.g. if the age is more elder, what would be the probability of being in class 1 and class2 ) , i can't find any command

[tina liang]

i mean the Marginal effect on probability of membership to class

[Weiwen Ng]

I believe you are referring to a latent class regression model, where you have age as a predictor of the latent class but not as an indicator of the latent class. The

You can refer to this thread. In your case, you have age as a continuous covariate in your latent class model, rather than the binary values in the example there. You'd merely need to respecify the at(...) option to include representative values of age. The dydx(age) option may also work.