st: RE: Using mixlogit as a substitute of xtlogit.

["Nick Cox" <n.j.cox@durham.ac.uk>]

The missing reference here is 

. search mixlogit

Keyword search

        Keywords:  mixlogit
          Search:  (1) Official help files, FAQs, Examples, SJs, and
STBs

Search of official help files, FAQs, Examples, SJs, and STBs


SJ-7-4  st0133_1  . . . . . . . . . . . . . . . . Software update for
mixlogit
        (help mixlogit if installed)  . . . . . . . . . . . . . . . A.
R. Hole
        Q4/07   SJ 7(4):593
        estimation speed improved and new options added for
        specifying weights and for obtaining robust and cluster-
        robust standard errors

SJ-7-3  st0133  . . Fitting mixed logit models using max. simulated
likelihood
        (help mixlogit if installed)  . . . . . . . . . . . . . . . A.
R. Hole
        Q3/07   SJ 7(3):388--401
        fits mixed logit models by using maximum simulated likelihood

To spell out the logic, which is important: 

1. Arne Risa Hole did a good job with -mixlogit- and deserves a nod of
recognition. (I'll tuck in here a comment that the Stata Journal is
always happy to get publicity like this.)

2. Someone, especially anyone new or inexperienced in Stata use, might
be intrigued by a posting like this and type -help mixlogit- and then be
mystified by the error message. Giving a one-sentence explanation that
this is a user-written command that you must install from the SJ site
would save such people from wasting their time trying to work out what
they did wrong. 

Nick 
n.j.cox@durham.ac.uk 

ippab ippab

I am wondering if there is any benefit of using mixlogit for a binary
dependent variable in a panel data.  Basically, to use mixlogit, I
will have to create two alternatives (which are complementary), an
alternative specific constant, and interactions between independent
variables and the alternative specific constant.  Is this is a wise
thing to do?

My vague understanding is that mixlogit allows for more heterogeneity
than xtlogit.  The other thing I noticed is that, for normally
distributed parameters, we can find out what percentage of the
population is on the other side of zero.  But, I am confused about
interpreting a population average beta obtained from xtlogit in light
of estimates from mixlogit.  Just to give an example, if the estimate
for x1 is 2.05 from xtlogit, the interpretation would be that
increasing x1 increases the the likelihood of y=1.  Now, getting a
similar estimate from mixlogit, e.g., mean is 2.5 for x1 with std
3.10, makes the interpretation complicated.  This means for about
20.9% of the sample, increasing x1 does not increase the likelihood of
y=1.  Am I understanding these correctly?  The more critical question
is, if more than 20% or 30% of the sample have different preference,
what does it mean to have a positive significant coefficient (mean)?


*
*   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/