Re: st: -mfx- after poisson

[Richard Williams <Richard.A.Williams.5@ND.edu>]

At 02:32 PM 4/29/2006, you wrote:
> I have a balanced panel of 3 years with approximately 800 individuals
> in each year (thus 2400 observations more or less).  I ran a poisson
> MLE with clustered and robust standard error correction.  There were
> several variables as regressors, but three worth mentioning (because
> I think they are causing the problem) are
>
> i.year
> i.fip
> i.year*i.fip
>
> That is, I am controlling for year, state and state-year fixed
> effects, and am using xi:poisson.
Some suggestions:

1) Those interaction terms may cause you grief no matter what you do.  See

http://www.stata.com/support/faqs/stat/mfx_interact.html

2) Use the -nose- option on mfx.  mfx runs much quicker if you don't 
calculate standard errors.  Other speeding up suggestions are at

http://www.stata.com/support/faqs/stat/mfx_speed.html

3) Download and try -margeff-.  It doesn't work with every command, 
but when it does work it tends to be much quicker than -mfx-.  It 
will also let you save the marginal effects in a format that is easy 
to input into -estout-.  Its -dummies- option might be a way of 
dealing with your interactions.

4) Yes, prchange is often a very good alternative to -mfx-. That 
might be the route I would go.

5) Try -findit marginal effects- to see if something else seems to 
meet your needs.


-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
FAX:    (574)288-4373
HOME:   (574)289-5227
EMAIL:  Richard.A.Williams.5@ND.Edu
WWW (personal):    http://www.nd.edu/~rwilliam
WWW (department):    http://www.nd.edu/~soc 
*
*   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/