Elizabeth said:
Using former posts from Statlist, I've found I need to use the command intreg because a Tobit will not produce robust st. errors.
I hope you also considered using user-written command -clad- (remember to refer to Stata commands and other syntax within ordinary text using dashes (- -) to flag words that you would type when using Stata), which "calculates Powell�s (1984) censored least absolute deviations estimator (CLAD) and bootstrap estimates of its sampling variance. It is a generalization of Tobit models which is robust to heteroscedasticity and is consistent and asymptotically normal for a wide class of error distributions." Note that it is not for interval data (you cannot specify censoring at both the lower and upper bound).
-clad- is package sg153 from http://www.stata.com/stb/stb58
P.S. Remember how to use -intreg- with data suitable for -tobit-:
Suppose you have a dependent variable like in:
tobit y x, ll(a) ul(b)
For use with intreg you should generate new variables like in:
gen depvar1 = y
*/* specify only if a lower limit is specified */*
replace depvar1 = . if y<=a
gen depvar2 = y
*/* specify only if an upper limit is specified */*
replace depvar2 = . if y>=b
intreg depvar1 depvar2 x
*
* 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/
