Re: st: nested logit

["Emla Fitzsimons" <efitzy75@hotmail.com>]

Yes, that does seem to be the problem.  For those vbles that I exclude from 
level3, and then include in level2, the coeffs and signif levels are the 
same whether or not I control for the inclusive value.  For the rest (ir the 
ones that I include in both level3 and level2), there are substantial 
differences in the coeffs depending on whether or not I include the 
inclusive value.

So am I right in saying that it is not really feasible to estimate a 3-level 
nested logit if the regressors are the exact same for all 3 levels?  For 
example I am looking at individual education choices across 3 years and 
using time-invariant individual and household characteristics as 
determinants of these choices.

Many thanks for your help,
Emla.


> From: "Ray Boston" <boston@cahp.vet.upenn.edu>
> Reply-To: statalist@hsphsun2.harvard.edu
> To: statalist@hsphsun2.harvard.edu
> Subject: Re: st: nested logit
> Date: Mon, 09 Jun 2003 14:47:17 EST
>
> Dear Emla,
> It seems to me that you may have hit in the problem.  Try removing
> one of the duplicated covariates and see if the problem goes away.
> Ray
> Ray C Boston, PhD
> Professor of Biomathematics and Biostatistics
> Dept. of Clinical Studies, NBC
> School of Veterinary Medicine
> 382 West Street Road Kennett Square, PA 19348
> Phone 610 444 5800 (X 2557)
>
> *
> *   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/
_________________________________________________________________
Sign-up for a FREE BT Broadband connection today! 
http://www.msn.co.uk/specials/btbroadband

*
*   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/