Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: Simple logit question - feeding data


From   Andreas Kuhn <akuhn@access.unizh.ch>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Simple logit question - feeding data
Date   Wed, 09 Jun 2004 18:32:53 +0200

Julia Gamas wrote:
Dear all,
I have a very basic question.  I am running a simple logit and want to know
how to "feed" the data to the logit command.
I have Y= whether a person took a private car (1)or public transit (0)on a
given trip.  My possible explanatory variables are, as reported by travelers
in an Origin Destination Survey:
-income per person (continuous),
-average distance between zones (average),
-travel time for the mode they used (continuous),
-number of vehicles available to the household this traveler belongs to
(0,1,2..)
-income bracket that this traveler's household belongs to (about 5
categories).
I will also be creating a
-cost variable (fare paid if public transit was used, parking and gas
expenses dependent on the average distance traveled.
Do I just give Stata this information as is, even if one variable is an
average, another is categorical and another is continuous (per person)?
I would hate to bunch all of them into categories because it seems to me I'd
lose a lot of information by doing this.
Any help with this would be appreciated, or if there is a book or paper I
can refer to, please let me know.
Thanks!
Julia A. Gamas
Program on Urban, Regional and Global Air Pollution
MIT


*
*   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/

the main point is that the choice of private versus public depends on both (i) attributes of the individual (e.g. income) and (ii) attributes of the choices themselves (e.g. cost and time). so I propose you first have a look at the so-called conditional logit model (see help clogit), but in order to estimate this model you will need access to cost and time information for both travel modes for each individual. references for this model are given in stata's manual.

mixing continuous and categorial data is not really a problem, as long as you are either using the - xi: - command or you are creating separate variables for the different values of the categorial variables (which can easily be done using the - table - command). the distance variable is also a continuous variable, albeit a noisy one (you do not know the exact distance). but I think there is not much you can do about that.

Andreas

*
* 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