Statalist The Stata Listserver

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

Re: st: Marginal effects for panel data neg binomial (xtnbreg)

From   "Austin Nichols" <>
Subject   Re: st: Marginal effects for panel data neg binomial (xtnbreg)
Date   Mon, 5 Mar 2007 17:50:33 -0500

Corey Phelps --
Do you mean like this?
webuse grunfeld, clear
xtnbreg mva kstock inv, fe i(com)
mat li e(b)
predict xb if e(sample)
gen exp=exp(xb)
su exp, meanonly
mat mfx=r(mean)*e(b)
mat li mfx

On 3/5/07, Corey Phelps <> wrote:
I am trying to calculate marginal effects using the "mfx compute" command
for a xtnbreg model. Because mfx computes marginal effects from the
post-estimated linear predictions, the default dydx calculation returns the
regresssion betas as marginal effects. However, I am interested in the
marginal effect (exp^BX)Bj, evaluated at the sample means of the covariates.
How do I go about obtaining this?

Corey Phelps
*   For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index