[Limdep Nlogit List] Repeated Nested-Logit Model

Leif leifand at u.washington.edu
Tue Sep 18 02:55:52 EST 2007


Professor Greene,
Just to clarify - it IS possible to estimate a repeated nested logit 
without random parameters heterogeneity and a repeated mixed logit model 
incorporating the heterogeneity you describe in NLOGIT, correct?

Leif Anderson
leifand at u.washington.edu



On Mon, 17 Sep 2007, William Greene wrote:

> Prof. Bujosa. The natural way to handle that would be with a mixed logit
> model in which the heterogeneity was constant across choice situations.
> However, NLOGIT does not support that model for the nested logit model.
> On the other hand, I do not see that Morey et al. actually did that either.
> /Bill Greene
> ----- Original Message -----
> From: "Angel Bujosa Bestard" <angel.bujosa at uib.es>
> To: limdep at limdep.itls.usyd.edu.au
> Sent: Monday, September 17, 2007 8:46:51 AM (GMT-0500) America/New_York
> Subject: [Limdep Nlogit List] Repeated Nested-Logit Model
>
> I agree with you, degenerate branches are not a problem for NLOGIT. My question was concerned with the "repeated" part of the problem. Morey et al's model use information on more than one choice per individual. Then, I was wondering if it is possible to estimate that model setting the choice variable as "frequency data"? Or if it is necessary to make some kind of modification to the usual nested logit command in order to consider more than one choice-ocasion per individual.
>
> Thank you very much!
>
> A.Bujosa
>
>
> ----- Original Message -----
>
> Prof. Bujosa: The model in that paper looks like a nested
> logit model with one (or more midel B) degenerate branches,
> which is no problem for NLOGIT. (The 1993 paper predates
> by several years the internet discussion by Hunt, Hensher
> and Greene which cumlinated in the papers by Hunt and by
> Hensher and Greene in which nested logit models with
> degenerate branches were developed at some length.  I believe
> that there were also a couple earlier papers on nested
> logit models with degenerate branches written by David
> Hensher.  These would also have appeared after Morey et al's
> paper.
> Regards,
> Bill Grene
> ----- Original Message -----
> From: "Angel Bujosa Bestard" <angel.bujosa at uib.es <http://limdep.itls.usyd.edu.au/mailman/listinfo/limdep>>
> To: limdep at limdep.itls.usyd.edu.au <http://limdep.itls.usyd.edu.au/mailman/listinfo/limdep>
> Sent: Monday, September 17, 2007 3:31:56 AM (GMT-0500) America/New_York
> Subject: [Limdep Nlogit List] Repeated Nested-Logit Model
>
> I am trying to estimate a Repeated Nested-Logit Model (Morey et al.,
> 1993) using Limdep. My question is: is it possible to estimate this
> model in Limdep? Maybe setting the choice variable to frequency data?
>
> Many thanks!!
>
> --
> A.Bujosa
> Balearic Islands University
>
> _______________________________________________
> Limdep site list
> Limdep at limdep.itls.usyd.edu.au <http://limdep.itls.usyd.edu.au/mailman/listinfo/limdep>
> http://limdep.itls.usyd.edu.au
>
>
> -- 
> Professor William Greene
> Department of Economics
> Stern School of Business
> New York University
> 44 West 4th St., Rm. 7-78
> New York, NY   10012
> http://www.stern.nyu.edu/~wgreene <http://www.stern.nyu.edu/%7Ewgreene>
> 212.998.0876
>
> _______________________________________________
> Limdep site list
> Limdep at limdep.itls.usyd.edu.au
> http://limdep.itls.usyd.edu.au
>
>
> -- 
> Professor William Greene
> Department of Economics
> Stern School of Business
> New York University
> 44 West 4th St., Rm. 7-78
> New York, NY   10012
> http://www.stern.nyu.edu/~wgreene
> 212.998.0876
>
> _______________________________________________
> Limdep site list
> Limdep at limdep.itls.usyd.edu.au
> http://limdep.itls.usyd.edu.au
>



More information about the Limdep mailing list