Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: Conditional Logit Coefficients


From   "Guido Tiemann" <tiemann@euv-ffo.de>
To   statalist@hsphsun2.harvard.edu
Subject   st: Conditional Logit Coefficients
Date   Wed, 14 Feb 2007 16:38:30 +0100 (CET)

Dear all!


My problem concerns the estimation of a conditional logit model. I do have
a choice variable and two independent variables which contain the
utilities of the various alternatives. Formally, I have

Pi(choice_i = j) = beta1*ut1 + beta2*ut2,

in Stata, I used

clogit choice ut1 ut2, group(id)

and everything turned out fine. However, I am not that interested in the
unrestricted coefficients, but I do want to compare and evaluate the
impact of both utility variables. In order to obtain a somewhat
standardized coefficient, both utilities should sum up to one
(beta1+beta2=1). So, the model I am really interested in would be

pi(choice_i = j) = beta1*ut1 + (1-beta1)*ut2.


Does anyone know how to estimate something like that?

Moreover, my question could also be “elevated” to another level. In the
next step, I want to undertake a pairwise comparision of 4 alternative
specific utilities altogether. So, what about

pi(choice_i = j) = beta1*ut1 + (1-beta1)*ut2 + beta3*ut3 + (1-beta3)*ut4.

Does anybody have any idea? Probably, there is any kind of logistic
transformation involved, but I do not have any idea how to do this … Thus
suggestion would be highly appreciated, since I am completely stuck in
that problem.

Best regards,


Guido



*
*   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/



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