On Fri, 11 Oct 2002, Brencic, Vera wrote:
> I would like to estimate a piece wise constant hazard in stata.
> the closest I got is the strata option but since this is available
> as part of stcox command, it seems not to be of much help.
> recommendation on the use of other statistical packages would
> also be greately appreciated;
For models in continuous time, think of this model as an Exponential model
with intercepts varying with survival time. So to estimate the model, you
simply need to episode split at the times the hazard varies, generate
appropriate dummy variables, and then include them in your Exponential
model estimated using -streg-. The data and estimation steps are discussed
in Lessons 3 and 5 at
http://www.iser.essex.ac.uk/teaching/stephenj/ec968/index.php
See also Cleves, Gould & Gutierrez "An Introduction to Survival Analysis
Using Stata" (More details about this book at Stata bookstore on-line).
For discrete-time models, the principles are very similar (see the URL
cited above)
Stephen Jenkins
*
* 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/
