# st: RE: question about xtlogit coefficient graphic interpretation

 From "Scott Merryman" To Subject st: RE: question about xtlogit coefficient graphic interpretation Date Mon, 30 Oct 2006 09:05:06 -0600

```-help xtlogit postestimation- will show you what is available.  There are no
canned routines to graph predicted probabilities after -xtlogit-, though it
is not too hard to roll you own:

//Prob of domestic car given different values of weight and mpg
//gear_ratio at mean
sysuse auto, clear
qui {
xtlogit fore mpg gear weight, i(rep) re
preserve
sum weight
local min = r(min)
local max = r(max)
collapse (mean) gear mpg
range weight `min' `max' `=`max' - `min' +1'
replace gear = gear[1]
forv i = 15(5)30 {
replace mpg = `i'
predict prob`i' ,pu0
}
}
twoway line prob* weight,  legend(off)
restore

Scott

> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-
> statalist@hsphsun2.harvard.edu] On Behalf Of Robin Luo
> Sent: Monday, October 30, 2006 1:39 AM
> To: statalist@hsphsun2.harvard.edu
> Subject: st: question about xtlogit coefficient graphic interpretation
>
> I need to interpret a coefficient from a xtlogit estimation
> graphically (it is a continuous variable). So, is a graph of predicted
> probabilities at changing values of this variable (holding all others
> constant) the only way to obtain a graphic interpretation? If so, is
> there a specialized postestimation command associated with xtlogit in
> Stata that could do the predicted probabilities easily?
> Thanks a lot!
> Robin

*
*   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/
```