1) Has anyone encountered the effect of data sorting (prior to running
an mlogit) on mlogit coefficients and test results? My sample consists
of 141908 obs of which 1259 are dropped for missing values for one
variable. Is there a relationship between these missing values and the
sorting when stata computes the mlogit, even though mlogit drops the
missing obs? There is nothing in the stata manual about sorting
issues. Thanks.
2) Does anyone know how to run fixed effects with mlogit in stata9? I
have not had any luck with this, despite obtaining results in mlogit.
I assume there is some additional programming to do in the ado file?
Has anyone done this? Thanks again.
Terri Nelson
UT Dallas
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
