# FW: st: mfx after xtnbreg and how to compute predicted Y

 From "Pek-Hooi Soh" To Subject FW: st: mfx after xtnbreg and how to compute predicted Y Date Mon, 20 Oct 2008 09:50:57 -0700

```Thanks for pointing this out. I am still learning to use statistical options
in Stata and the help menu is not too straightforward for me. Will
appreciate any help to my questions about IRR and Predict with xtnbreg.

Two follow up questions: 1) to calculate the marginal effect for a two-way
interaction term. Do I run IRR option and take the IRR coefficients to
compute the effect, equivalent to (β3 + β4 x2) as follow?

Say, E[y | x] = exp(β1 + β2 x2 + β3 x3 + β4 x2 x3 )

Then the proportionate change in the conditional mean due to a
one-unit change in x3 equals (β3 + β4 x2).

2)  I need to compute
Y2(X1 | other_X_at_mean_value)
Y1(X1+1std_dev | other_X_at_mean_value)
calculate Y2-Y1.

How do I use the options predict to have the values X1 and X1 + 1 std dev ?

Cheers,

Pek

--- Pek-Hooi Soh <phsoh22@gmail.com> wrote:
> I estimated the full model by xtnbreg and ran mfx.  The marginal
> effect table presents the same coefficients and standard errors as
> the table for xtnbreg. Is this possible?

Yes, if you type -help mfx- it tells you that: "By default, mfx
calculates the marginal effects or elasticities at the means of the
independent variables by using the default prediction option associated
with the previous estimation command." If you type
-help xtnbreg postestimation- it tells you that the linear predictor is
the default for -predict- after -xtnbreg-. This means that the marginal
effect -mfx- computes is just the same as the output of -xtnbreg-.

If you want to interpret incidence rate ratios you don't have to use
-mfx-, you can just use -xtnbreg- with the -irr- option. See:
-help xtnbreg-.

> If I do get the correct mfx output, how do I compute predicted Y? Do
> I have compute y = exp(XB) where B stores marginal effects separately
> in excel ?

See -help xtnbreg postestimation- and look at the -predict- section.

Hope this helps,
Maarten

-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands

Buitenveldertselaan 3 (Metropolitan), room N515

+31 20 5986715

http://home.fsw.vu.nl/m.buis/
-----------------------------------------

Send instant messages to your online friends http://uk.messenger.yahoo.com
*
*   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/

*
*   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/
```