Explained variation for survival models
Patrick Royston
Cancer Group
MRC Clinical Trials Unit
222 Euston Rd.
London NW1 2DA, UK
[email protected]
|
Abstract. This article introduces a new measure of explained variation for use with
censored survival data. It is a modified version of a measure previously
described by John O’Quigley and colleagues, itself a modification of
Nagelkerke’s earlier proposal for a general index of determination. I
describe Stata programs str2ph, which implements the new measure, and
str2d, which implements a measure proposed in 2004 by Royston and
Sauerbrei. I provide examples with real data.
View all articles by this author:
Patrick Royston
View all articles with these keywords:
censored survival data, regression models, index of determination, explained variation, explained randomness, information gain
Download citation: BibTeX RIS
Download citation and abstract: BibTeX RIS
|