Comparing Two Survival Time Distributions: An Investigation of Several Weight Functions for the Weighted Logrank Statistic |
| |
| Authors: | T Gaugler D Kim S Liao |
| |
| Institution: | 1. Department of Statistics , The Pennsylvania State University , University Park, Pennsylvania, USA tgaugler@stat.psu.edu;3. Department of Statistics , The Pennsylvania State University , University Park, Pennsylvania, USA |
| |
| Abstract: | In survival analysis, it is often of interest to test whether or not two survival time distributions are equal, specifically in the presence of censored data. One very popular test statistic utilized in this testing procedure is the weighted logrank statistic. Much attention has been focused on finding flexible weight functions to use within the weighted logrank statistic, and we propose yet another. We demonstrate our weight function to be more stable than one of the most popular, which is given by Fleming and Harrington, by means of asymptotic normal tests, bootstrap tests and permutation tests performed on two datasets with a variety of characteristics. |
| |
| Keywords: | Bootstrap test Censored data Permutation test Survivor function Weight function Weighted logrank statistic |
|
|
