Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: RE: the order of variables makes difference in 'Heckprob'?


From   "Nick Cox" <[email protected]>
To   <[email protected]>
Subject   st: RE: the order of variables makes difference in 'Heckprob'?
Date   Wed, 27 Jan 2010 20:37:47 -0000

Show us an example with a dataset accessible to Stata. 

Nick 
[email protected] 

Chungshik Moon

I have a question about heckman's selection probit model (heckprob).
It sounds really weird, but it turns out that the outcome changes as the order of independent variables changes.

That is,
heckprob y1 x1 x2 x3, s(y2 x11 x12 x13) 
heckprob y1 x2 x1 x3, s(y2 x11 x12 x13) 

have different outcomes for x1 and x2 in both Stata 10 and 11

Even worse is that results vary across the version of Stata. Not only does each version show the same variable order issue, but even the same model shows different results across the version. 

Please help me out to address this issue.
I also wonder if you have any alternatives to heckprob.

*
*   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/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index