I believe it should be "/sigma" not "_se".
For example:
. sysuse auto
(1978 Automobile Data)
. qui tobit mpg len, ll(20)
. qui sum len
. local meanlen=r(mean)
. nlcom
normden(-((20-(_b[len]*`meanlen'+_b[_cons]))/_b[/sigma]))*(_b[len]/_b[/sigma
])
_nl_1:
normden(-((20-(_b[len]*187.9324324324324+_b[_cons]))/_b[/sigma]))*(_b[len]/_
b[/sigma])
---------------------------------------------------------------------------
mpg | Coef. Std. Err. t P>|t| [95% Conf. Interval]
----------+----------------------------------------------------------------
_nl_1 | -.0254299 .0034924 -7.28 0.000 -.0323903 -.0184694
---------------------------------------------------------------------------
. mfx, predict(pr(20,.))
Marginal effects after tobit
y = Pr(mpg>20) (predict, pr(20,.))
= .45630887
----------------------------------------------------------------------------
variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X
---------+------------------------------------------------------------------
length | -.0254299 .00349 -7.28 0.000 -.032275 -.018585 187.932
----------------------------------------------------------------------------
Hope this helps,
Scott
> -----Original Message-----
> From: [email protected] [mailto:owner-
> [email protected]] On Behalf Of Bart Vanneste
> Sent: Friday, February 03, 2006 5:49 AM
> To: [email protected]
> Subject: st: inteff tobit
>
> Dear,
>
> How can I correctly compute an interaction effect in a
> Tobit model?
>
> On 14 Jan 2005, May Boggess provided an example (see
> http://www.stata.com/statalist/archive/2005-01/msg00264.html).
>
> However, using the exact same code:
>
> clear
> sysuse auto
> tobit mpg len, ll(20)
> predict xb,xb
> predict p, pr(20,.) /* Probability Uncensored */
>
> gen myp = norm(-((20-xb)/_b[_se]))
> sum p myp
>
> ----------
>
> I get the following error message:
> . gen myp = norm(-((20-xb)/_b[_se]))
> [_se] not found
> r(111);
>
> end of do-file
> r(111);
>
> ----------
>
> Is this error message due to different versions of
> Stata? I am using Stata 9.1 for Windows.
>
> Any suggestions or alternative approaches would be
> greatly appreciated.
>
> Many thanks,
> Bart
>
>
*
* 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/
