|
Abstract:
|
Statistical procedures for assessment of treatments often involve complexities due to incomplete data. Here we propose a nonparametric method for testing the homogeneity of survival means based on right censored data. We develop a test based on Kaplan-Meier means. We derive asymptotic properties of this test statistic. Through simulations we compute and compare the power of this new procedure with that of the log-rank, Wilcoxon, and the Cox model. Our results indicate that performance of these estimation procedures depends on the level of censoring and whether the underlying assumptions are met. When the assumptions of log-rank and the Cox model are met, these procedures are more powerful than the other two tests. However, when the objective is testing homogeneity of survival mean, rather than survival curves, our new test statistic and Wilcoxon seems to have some advantages.
|
