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

RE: st: marginal effects for ordered logit


From   "Markiewicz, Agnieska" <Agnieszka.Markiewicz@econ.kuleuven.ac.be>
To   <statalist@hsphsun2.harvard.edu>
Subject   RE: st: marginal effects for ordered logit
Date   Wed, 10 Nov 2004 15:51:56 +0100

Dear Marteen,

I was following your discussion with Pradeep on calculation of marginal
effects of quadratic terms. I have the impression that the command nlcom
is not available in the 8 release. (I have only the 7) Is there any
other command I could use instead?

Aga

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of maartenbuis
Sent: Thursday, November 04, 2004 5:20 PM
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: marginal effects for ordered logit

Dear Aga,

The marginal effect you want is just the slope of a tangent line of
these graphs at a specific value of your explanatory variable (often
the mean of the explanatory variable), so the marginal effect just
shows part of the information represented by the graph. Furthermore -
mfx - won't calculate the marginal effects you want if you include a
quadratic term, so you'd probably have to do that manually.

Alternatively you could use the fact that the ologit is linear in the
log(odds), i.e. the log(odds) of belonging to one group versus all
`lower' groups is a linear function of the explanatory variables. So
you could plot or tabulate the odds ratio for various values of your
explanatory variable.

Have a look at chapter 5 of `Regression models for categorical
dependent variables using Stata', by Scott Long and Jeremy Freese if
you want to know more than I can tell you through this list.

Maarten

--- <Agnieszka.Markiewicz@e...> wrote:
> I was actually trying to plot my results but they don't look very
> convincing. Accordingly to your remark, the probability of
belonging to
> one of the classes changes and my graphs don't provide any clear
> explanation. Thus I still would like to interpret my results without
> graphs.




*
*   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/

*
*   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