Jack Noone posted on Wednesday, September 12, 2007 - 3:54 pm
Hello I am new to SEM and mPLus and i have a couple of questions. 1. I have one latent outcome variable that is indicated by 7 dichotomous items and a number of latent independent variables that are indicated by a combination of continuous and ordinal variables. I am assuming that i would need to use categorical data modeling but can I also employ the techniques for dealing with missing data and complex survey designs?
2. I am expecting that the latent outcome variable will also have two second order variables - can you point me in the direction of any literature regarding second order variables in mPLus?
1. Yes, these can all be combined in Mplus. You can use either weighted least squares or maximum likelihood with TYPE=MISSING; and the complex survey data features described in the user's guide. See Complex survey data in the index.
2. See the second-order factor example in Chapter 5.