Conducting path analysis with a binary dependent variable

[Monique]

Hello all,

I am relatively new to STATA but have read that you can use the gsem function to conduct a path analysis with a dichotomous outcome. The path model that I will be running will have two such variables (i.e., two binary endogenous variables). How do get STATA to treat this path model as a logit model? Or can you direct me to any resources or literature I can review.

Thanks for any assistance you can offer.

Monique

[Richard Williams]

Stata has a bunch of videos on sem and gsem. See, for example,

http://www.youtube.com/watch?v=CjsvhOl4ZAw

http://www.youtube.com/results?search_query=sem+stata

The manual is also good, as is Acock's book:

http://www.stata.com/bookstore/disco...g-using-stata/