Time to event is usually analyzed using survival analysis models, of
which Stata has plenty. I am not familiar with work on the treatment
effect (in econometric terms) for survival models, but I am pretty
sure there's been enough work done in biostatistics literature where
some similar topics include imperfectly randomized designs and
non-compliance with the study protocol. I am just bringing this up as
I believe survival models will be a better way to model your data.
Quantile regression is now hot in economics, but it would not seem the
tool of choice here, at least to me.
On 1/21/06, Marcello Pagano <[email protected]> wrote:
> For Jeanne:
>
> I am looking for Stata code that will run quantile regression under
> random censoring. (clad for fixed censoring won't work for my problem.)
> I have been unable to find Stata-based code through the usual sources,
> or via authors of articles on the topic (and I am not a programmer).
>
> I have a large dataset with a time to event variable, observed range of
> 0 to 730 days, with administratively censored times varying from 365 to
> 730 days (approx 10% of the data is censored). I am looking at the
> effect of a binary treatment on duration, and would like to use -qreg-
> or some version of quantile regression if I can find a Stata user ado
> file that can handle this type of censoring.
>
> Thanks for any help.
>
> Jeanne
> *
> * 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/
>
--
Stas Kolenikov
http://stas.kolenikov.name
*
* 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/
