st: Re: obtaining confidence intervals after mfx compute

[Kit Baum <kitbaum@mac.com>]

Jose said

Does anyone know how to obtain confidence intervals for the estimated Y
variable after the mfx compute command? Let's say you have an estimate 
of
your dependent variable, which is what (mfx compute) computing mean 
marginal
effects does. A very important issue in using regression estimates is 
the
ability to forecast the confidence interval of this dependent variable. 
As
far as I can tell, mfx compute gives you a confidence interval for the X
variables, but I need a measure of confidence for the predicted Y 
variable.

-whelp regress- indicates that after a regression, you may use 
-predict- to calculate either the standard error of prediction (stdp) 
or standard error of forecast (stdf). As any econometrics text (or the 
Stata manual) will demonstrate, these quantities may be used to 
generate an appropriate confidence interval for E[y | x]  or
E[y_0 | x_0].

mfx is a very useful tool, but confidence intervals for prediction from 
a regression is not what it is for.

Kit

*
*   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/