[Limdep Nlogit List] MNL: Quantities greater than one

[Fred Feinberg]

Prof. Greene is correct.  The MNL model is based on random utility theory,
and specifies the probability of a particular *single* item offering the
greatest (latent) utility value.  It presumes homoscedastic (double
exponential) errors, which is not always a desirable/realistic assumption.
It is NOT the correct model for more than one outcome.

The marketing literature has dealt with this issue extensively.  See the
classic papers by Chiang
(http://www.jstor.org/view/07322399/ap010040/01a00020/0) and Gupta
(http://www.jstor.org/view/00222437/sp040100/04x1951a/0).  Google scholar
can point you to the subsequent literature.  There is also some work by
Bockenholt in the psychometric tradition on modeling multiple outcomes.

Fred

=====

Prof. Fred Feinberg
Ross School of Business
University of Michigan
feinf at umich.edu

wgreene at stern.nyu.edu wrote:

> Dear Mr. Bau. I'm not persuaded that the MNL is the right model.
> The model describes *IF* the consumer makes a choice, not how
> many choices they make.  In principle, you can use "frequencies"
> data, but this, once again, would be for how many times the consumer(s)
> made the choice, not how many times a consumer made the choice.
> Perhaps you should consider a model for counts.
> Regards,
> Bill Greene
>
> ************************************************
> Professor William Greene
> Department of Economics
> Stern School of Business
> New York University
> 44 West 4th St., Rm. 7-78
> New York, NY   10012
> Ph. 212.998.0876
> Fax. 212.995.4218
> URL. http://www.stern.nyu.edu/~wgreene
> Email. wgreene at stern.nyu.edu
> ************************************************
>
> ----- Original Message -----
> From: Raimund Bau <Raimund.Bau at rwth-aachen.de>
> Date: Thursday, January 11, 2007 12:38 pm
> Subject: [Limdep Nlogit List] MNL: Quantities greater than one
>
> > Dear List,
> >
> > I am preparing a household panel data set for a MNL/RPL model.
> > Apparently, neither the NLogit documentation nor "Applied Choice
> > Analysis" cover the question how to deal with consumers buying
> > more
> > than one item of the same kind at the same location. Should I use
> > the
> > choice field or is this just a boolean value?
> >
> > Regards,
> >
> > Raimund Bau
> > RWTH Aachen
> >
> > _______________________________________________
> > Limdep site list
> > Limdep at limdep.itls.usyd.edu.au
> > http://limdep.itls.usyd.edu.au
> >
>
> _______________________________________________
> Limdep site list
> Limdep at limdep.itls.usyd.edu.au
> http://limdep.itls.usyd.edu.au