Maarten--
While -intreg- does seem tailor-made in some ways for this problem of
testing whether e.g. E(Y|X=A)>E(Y|X=B) when Y is discretized as Y*
(measured in large intervals), by simply typing e.g.
intreg ystar groupA
the assumption in -intreg- is that e (or Y-bX) is N(0,s) which seems
unlikely to hold true for incomes measured in dollars.
On 9/18/07, Maarten buis <[email protected]> wrote:
> You lose information that way, and if you just want to test, why not do
> that on the whole distribution? More interesting would be to use
> -intreg- as this command was specifically designed for this case.
*
* 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/
