I am writing to ask a question on two-staged regression. My dependent variable is a latent variable with 15 items. I regress it on several sets of covariates, including demographic factors. One of the covariate (X1)is endogenous. It is a function of some demographic factors and is correlated with one item in the latent variable. X1 is a dichotomous variable. I found an intrumental variable for X1. I am wondering whether and how I can perform two-staged regression using mplus.