Journal article
Linear signed-rank tests for paired survival data subject to a common censoring time
Lifetime data analysis, Vol.11(3), pp.351-365
09/2005
DOI: 10.1007/s10985-005-2967-6
PMID: 16133884
Abstract
In this paper we consider rank-based tests for paired survival data, in which pair members are subject to the same right censoring time. Linear signed-rank tests have already been developed for the two-treatment problem in which pair members receive the opposite treatments. Assuming a bivariate accelerated failure time model, we extend this class of linear signed-rank tests to the case of multiple covariates, making this methodology applicable to more complicated experimental designs. These tests can be reformulated as weighted sums of contigency table measures, giving an alternative method of computation and intuitive view of how these tests work. A simulation study of their small-sample performance relative to other tests demonstrates that the linear signed-rank tests have greater power in cases of moderately to highly correlated data.
Details
- Title: Subtitle
- Linear signed-rank tests for paired survival data subject to a common censoring time
- Creators
- Michael P Jones - Department of Biostatistics, College of Public Health, University of Iowa, Iowa City, 52242, USABongin Yoo
- Resource Type
- Journal article
- Publication Details
- Lifetime data analysis, Vol.11(3), pp.351-365
- Publisher
- United States
- DOI
- 10.1007/s10985-005-2967-6
- PMID
- 16133884
- ISSN
- 1380-7870
- eISSN
- 1572-9249
- Grant note
- R01 CA 52212 / NCI NIH HHS
- Language
- English
- Date published
- 09/2005
- Academic Unit
- Statistics and Actuarial Science; Biostatistics; Public Policy Center (Archive)
- Record Identifier
- 9983985921302771
Metrics
83 Record Views