Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
Stata's -sem- does not support nonlinear latent variable models. However, you might be able to set this up in -gllamm- if t is an observed variable. Frankly, I am not sure your model is identified. On Wed, Dec 7, 2011 at 8:25 AM, Zhi Su <su.zh@husky.neu.edu> wrote: > Hi, dear Statalists, > > I want to use "sem" for a structural equation model, where Y is > outcome, x is colntorl variable, t is variable with interest, A is a > latent variable. The latent variable has two indicators :a1 a2. I want > to run a model as the following. > > Y<-x t t*A > A<-x > a1<-A > a2<-A > > Here is the problem. t*A is the interaction term between A and t, and > Stata return that "t*A" is an unknown variable. The interaction term > can not be generated to be an variable by using "generate" because A > is a latent variable. How can I use an interaction term between latent > variable and observed variable in "sem"? Thanks! -- Zhi Su 348 Holmes Hall Northeastern University 360 Huntington Avenue Boston, MA 02115 Office:1-617-373-2316 email:su.zh@husky.neu.edu * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/