Statalist


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

st: Testing goodness of fit


From   "Marie-Therese Cooney" <Marie-Therese.Cooney@amnch.ie>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: Testing goodness of fit
Date   Fri, 4 Jul 2008 13:16:17 +0100

Hi, 

I have derived a model for predicting 10 survival, based on certain 
risk factors. The follow-up time in my dataset is variable, so I have
used Cox 
methods in the derivation. 

But to test it I have been using Hosmer Lemeshow (the downloadable hl 
command in STATA). My problem is that I can shorten the follow-up on 
the longer observations to 10 years in order to run this, but there is 
nothing I can do with the shorter follow-ups. So, the test shows that 
the risk is over-estimated, because some of those who are included in 
the observed 10 year risk only have 5 years of follow-up. 


Does anyone know if there is a goodness of fit test that can use 
variable follow-up in STATA (similar to the post estimation command 
for Harell's C instead of AUROC)? 


Many thanks for your help, 
MTC

*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index