Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: Predicted survivor curve for model with interactions


From   "Stephen P. Jenkins" <stephenj@essex.ac.uk>
To   statalist@hsphsun2.harvard.edu
Subject   st: Predicted survivor curve for model with interactions
Date   Tue, 29 Jul 2003 08:52:46 +0100 (GMT Daylight Time)

A colleague forwarded me the following problem about how to treat 
interaction terms when drawing survivor curves. Any ideas please?
--------------------------------------
I have the model

B1 drug + B2 Age + B3  drug*Age

Imagine I want to plot the predicted survivor curves for the following cases

1) age=35, drug=0
2) age 50 drug=0
3) age=35  drug=1
4) age=50 drug=1

If I go about what you do, I will estimate the model using either the variable 
age-35 or age-50 and then plot the baseline in each case and obtain cases 1) 
and 2). In those cases since I have centered the age variable to 35 or 50, 
that variable assumes value 0 both for age and for age*drug. 
However, if I am interested in comparing those baselines with the case where 
drug=1, it seems to me that only the term B1 (drug) will vary here whereas the 
value of age*drug will continue to be zero since I have defined age to be 
age-35. To sum up, how do I plot cases 3) and 4)?

Interestingly I cannot find in the literature of survival analysis many 
references to interactive variables, any suggestions?
-----------------------


Stephen
----------------------
Professor Stephen P. Jenkins <stephenj@essex.ac.uk>
Institute for Social and Economic Research (ISER)
University of Essex, Colchester, CO4 3SQ, UK
Tel: +44 (0)1206 873374. Fax: +44 (0)1206 873151.
http://www.iser.essex.ac.uk

*
*   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