Re: st: Marginal Effects for Binary Response Models with Nonlinear Regressors

[Maarten buis <maartenbuis@yahoo.co.uk>]

--- On Sun, 30/5/10, Olivia Bacha wrote:
> Consider the following model 'logit y x(i)/sum_x(i)', where
> variable x(i) is expenditure for individual i, then sum_x(i) 
> sum up total expenditures for all individuals, it is a ratio.
> As stata uses delta method, mfx command will give me x(i)/
> sum_x(i) + delta , not x(i)+delta/[sum_x(i)+delta]. It is a
> sort of same problem we face to calculate marginal effects
> by including interactions in logit model.

I am not sure whether the this is the same problem, if you added the
proportion, then that means that you think that the proportion is the
unit of interest, in which case I see no problem. However, if you 
disagree you can always compute the derivative by -nlcom- and let 
-nlcom- take care of the standard errors.

Hope this helps,
Maarten

--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany

http://www.maartenbuis.nl
--------------------------


      

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/