Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | "JVerkuilen (Gmail)" <jvverkuilen@gmail.com> |
| To | statalist@hsphsun2.harvard.edu |
| Subject | Re: st: ordered logistic regression with endogenous variable |
| Date | Thu, 11 Oct 2012 09:26:59 -0400 |
On Thu, Oct 11, 2012 at 7:02 AM, Anat (Manes) Tchetchik <anatmanes@gmail.com> wrote: > Dear Statalisters, > I have a categorical (ranked ordered) dependent var. is there a way to > estimate such a model with IVs for endogenous covariate? How coarse is the ordinal variable? If it's got 10 categories or so and is reasonably centered and symmetric (the last part is crucial) you can get away with assuming it's continuous and just using an estimator for instrumental variables. You can probably get away with it even down to 5 categories, though you'll have some attenuation due to the discrete response. If it's got 3 points, is skewed or near the boundary of the response space, you're likely to run into trouble, but if you are intrepid you can use -gllamm- or -gmm- to get an estimate. * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/faqs/resources/statalist-faq/ * http://www.ats.ucla.edu/stat/stata/