Re: st: Goodness of fit using Cox-snell residuals

[rgutierrez@stata.com (Roberto G. Gutierrez, StataCorp)]

Emelda Okiro <eokiro@kilifi.kemri-wellcome.org> asks:

> Thank you for your helpful insight with this. I have one more question.
> Because the cumulative Cox-Snell residuals are only recorded in one
> observation per subject, sts generate is only using this observation/record
> to estimate the cumulative hazard. This is a problem in my data set as there
> is a failure outcome for each observation in the dataset. It means that only
> five failures out of the 952 failure (those that happen in the observation
> that has the cumulative cs residual) are actually used to graph the
> cumulative hazard.

This is strange, considering that the cumulative Cox-Snell residuals are
always stored in the last observation (chronologically) within each subject
ID.  As a result, ultimate failure or censoring for the subject is preserved
when -stset-ting the cumulative residuals.

If Emelda still feels that something inappropriate is occurring with her 
data, she can email them to me for closer inspection.

--Bobby
rgutierrez@stata.com
*
*   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/