Hello everyone
I am trying to do a power calculation for a logistic regression analysis, and
have managed to track down a few journal articles on my research topic so
that I can get an idea what kind of odds ratio I should be expecting.
However, I notice that some of these articles don't use logistic regression but
something referred to either as Cox proportional hazards or as Cox regression
analysis. I have never come across this type of analysis before; all I have
managed to find out about it from reading the articles and doing a quick
Google search is that it does something similar to logistic regression, but is
somehow different.
So - before I head down to the library to set up camp in the epidemiology
section, can anyone recommend an accessible introductory text or references
on Cox, or maybe provide me with an outline of what it does? Specifically, I'd
like to know why one would use that rather than logistic regression and how
comparable the two types of analysis are.
Thanks!
Denham
|