Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: intreg and missing standard errors follow up question


From   "Mai Bui" <mai.bui@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   st: intreg and missing standard errors follow up question
Date   Tue, 22 Jul 2008 21:18:51 -0700

Hello everyone,

I am working with data that needs correcting for 1.clustered standard
errors; 2.censored dependent variable. The Tobit command in Stata 9
doesn't have a cluster option, so I'm using Intreg, which does allow
clustering.  However, for model specifications with more than
approximately 7 independent variables (I have only 10 provincial
clusters), Stata appears unable to calculate all the standard errors (SE)
when using Intreg, displaying instead "." where SE are missing.

My question: Is there any way to both cluster and correct for
censoring using Stata 9, or is it a matter of choosing between
partially unobservable but clustered SE, and observable but
unclustered SE?

I've read the thread "st: intreg and missing standard errors" between
Schaffer and Lucas
<http://www.stata.com/statalist/archive/2004-01/msg00382.html>, and
I'm hoping for insight provided in addition to this helpful exchange.

Thanks for your consideration!

Mai
(Graduate student at UBC)
*
*   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 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index