Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: Problems with calculating marginal effects for categorical variables in ologit models


From   "Eva Poen" <eva.poen@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Problems with calculating marginal effects for categorical variables in ologit models
Date   Tue, 22 Apr 2008 12:23:23 +0100

I cannot reproduce your problem using the auto data (see example
below). Btw, it's unlikely that you have literally typed -Margeff,
dummies (B *)-. Instead, you probably wrote
-margeff, dummies(B*)-.

Also, you said "D are all dummies for which you want to calculate the marginal
effects". The option -dummies- does not mean that marginal effects
will be calculated only for these variables. The option ensures that,
for the process of calculating marginal effects, only one of the
dummies will be set equal to 1, while all others are set to 0.

If you send the exact code of your ologit model and the margeff
command to the list, people will be more likely to help.

Eva

** example
sysuse auto, clear
tab headroom, gen(hr) /* create dummies */
ologit rep78 foreign  hr1 hr2 hr3 hr4 hr5 hr6 hr7
margeff, dummies(hr*)
** end of example




2008/4/22, MANUELFERN.FLORES@rai.usc.es <MANUELFERN.FLORES@rai.usc.es>:
> Dear users,
>
>  My question is following one: I do not know how to calculate average
> marginal effects (AME) for categorical variables with more than two
> categories?
>  In my ordered logit model, I have a dependent variable with 4 categories
> and 5 categorical explanatory variables, 3 of them with more than two
> categories. I tried to implement the method of Tamás Bartus which is
> decribed in his article "Estimation of marginal effects using margeff"
> (2005). His suggestion is to use the following command:
>  Margeff, dummies (B *)
>  where D are all dummies for which you want to calculate the marginal
> effects.
>  The problem is that by trying to do this I got the following error:
> "[_cut1] not found" r(111)
>
>  I would appreciate very much any suggestions.
>
>  *
>  *   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