Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: asclogit vs mixlogit


From   "JVerkuilen (Gmail)" <jvverkuilen@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: asclogit vs mixlogit
Date   Tue, 4 Dec 2012 14:41:23 -0500

On Tue, Dec 4, 2012 at 11:41 AM, elisabetta capobianco
<elisabetta.capobianco@gmail.com> wrote:
> Dear Statalist,
>
> I have done an asclogit and a mixlogit.
> What does it means that mixlogit coefficients are bigger than asclogit ones?
> And is there a command to evaluate the preference and have the
> coefficient for a product that is both bio and origine and trasport?
> Thank you for your help.

Caveat: It's not 100% clear to me that these models are actually
comparable due to a number of well-known indeterminacies. Hopefully
someone who runs these models on a regular basis can comment.

Note that the mixlogit log-likelihood is much better than that of
asclogit, so the former model fits much better.

What often happens with a mixed model is that its coefficients will be
more extreme than the population-averaged ones implied by the asclogit
model. This is because the asclogit coefficients have to compromise if
there is random variation between respondents and are pulled towards 0
as a consequence due to the nonlinearity of the logit link function.
*
*   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/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index