Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


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

Re: st: mixlogit and mixlbeta producing the same results


From   Arne Risa Hole <[email protected]>
To   statalist <[email protected]>
Subject   Re: st: mixlogit and mixlbeta producing the same results
Date   Tue, 28 Jan 2014 17:33:48 +0000

Erik,

This behaviour is to be expected. In your -mixlogit- syntax ECP, EWP,
EFR, EJS and EJI all have fixed coefficients so when you ask
-mixlbeta- to calculate the individual-specific coefficients it will
simply assign the fixed coefficient to every individual. To avoid this
allow some or all of the variables to have random coefficients, e.g.

mixlogit Choice ECP EWP EFR EJS, group(GID) id(INDID) rand(RR EJI)
mixlbeta EJI, saving(testfile)

Remember that the Statalist FAQ asks you to say if a program is
user-written: -mixlbeta- is a post-estimation command that is part of
the -mixlogit- package (SJ, SSC). See
http://www.stata-journal.com/article.html?article=st0133 for more info
about -mixlogit-.

Arne

On 28 January 2014 16:54, Erik Monsen <[email protected]> wrote:
> Dear Stata List,
>
> To calculate individual preferences for choice based conjoint data I have collected from 70 respondents who have made 8 choices each, I first run
>    mixlogit Choice ECP EWP EFR EJS EJI, group(GID) id(INDID) rand(RR)
> to compute the preference scores averaged across all respondents.
>
> To calculate the individual preferences for each of the 70 respondents, and then I run
>    mixlbeta ECP EWP EFR EJS EJI, saving(testfile)
>
> The problem is, mixlbeta assigned all 70 respondents exactly the same set of preference scores, instead of different scores for each respondent.
>
> How can I modify this command so that different preference scores are reported for each respondent?
>
> I have been combing the net for the answer, but I have not found it yet and would appreciate your help.
>
> I have also tried to run mixlogit separately for each individual respondent, but the results do not converge.
>
> Any insights to these various issues would be greatly appreciated.
>
> Best Regards, Erik
>
> =============
> Dr Erik Monsen
> Senior Lecturer and Director of Research
> Hunter Centre for Entrepreneurship
> Strathclyde Business School, University of Strathclyde
> Room 7.06, Sir William Duncan Building
> 130 Rottenrow, Glasgow, G4 0GE
>
> *
> *   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/
*
*   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 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index