Suppose, in the auto dataset, domestic observations are truncated at 17
while foreign observations are truncated at 18. I believe you could do the
following:
. sysuse auto
(1978 Automobile Data)
. gen trunc = cond(foreign == 0,17, 18)
. truncreg mpg price, ll(trunc) nolog
(note: 20 obs. truncated)
Truncated regression
Limit: lower = trunc Number of obs =
54
upper = +inf Wald chi2(1) =
0.62
Log likelihood = -150.79259 Prob > chi2 =
0.4312
----------------------------------------------------------------------------
--
mpg | Coef. Std. Err. z P>|z| [95% Conf.
Interval]
-------------+--------------------------------------------------------------
--
eq1 |
price | -.0015613 .0019834 -0.79 0.431 -.0054486
.0023261
_cons | 19.6171 10.84309 1.81 0.070 -1.634962
40.86917
-------------+--------------------------------------------------------------
--
sigma |
_cons | 9.881807 3.537671 2.79 0.005 2.9481
16.81551
----------------------------------------------------------------------------
--
Hope this helps,
Scott
> -----Original Message-----
> From: [email protected] [mailto:owner-
> [email protected]] On Behalf Of Eric Sevigny
> Sent: Thursday, September 30, 2004 5:07 PM
> To: [email protected]
> Subject: Re: st: Defining observation specific truncation points with
> truncreg
>
> Right, but what if I have ll(10) for 1997 and ll(20) for 1996 and want to
> model these different truncation points? Can I do that with truncreg?
>
*
* 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/
