Testing the additional predictive value of high-dimensional data CRAN Task View: Survival (http://CRAN.R-project.org/view=Survival) 'globalboosttest' implements a permutation-based testing procedure to globally test the (additional) predictive value of a large set of predictors given that a small set of predictors is already available. Currently, 'globalboosttest' supports binary outcomes (via logistic regression) and survival outcomes (via Cox regression). It is based on boosting regression as implemented in the package 'mboost'. Software Depends: R (>= 2.8), mboost (>= 2.0-0), survival Anne-Laure Boulesteix <boulesteix@ibe.med.uni-muenchen.de> Anne-Laure Boulesteix <boulesteix@ibe.med.uni-muenchen.de>, Torsten Hothorn <torsten.hothorn@stat.uni-muenchen.de>. GPL (>= 2) 2011-02-10 application/tgz http://CRAN.R-project.org/package=globalboosttest