Comparing Survival Curves Using Contingency Tables
Date of Award
Doctor of Philosophy
Joshua Naranjo, Ph.D.
Hyun Bin Kang, Ph.D.
Duy Ngo, Ph.D.
Mathew Rosales, Ph.D.
Chi-square test, compare k survival curves, compare two survival curves under non-pH models, contingency tables, test of homogeneity in survival analysis
Under proportional hazard assumption (PH) in survival analysis, log-rank test (Mantel, 1966) is optimal. When PH assumption is violated, Wilcoxon test (Gehan, 1965), weighted log-rank test (Fleming, Harrington, & O'Sullivan, 1987), Max- Combo test (Karrison, 2016), and restricted mean survival time (Royston & Par- mar, 2013) are alternative methods.
In this study, we propose a chi-square test of homogeneity for comparing survival curves. We propose an adaptive procedure for choosing between the number of classes in the contingency table. We also discuss an adaptive procedure based on Simes adjustment for multiple testing.
Restricted to Campus until
Kiatchupipat, Udomlux, "Comparing Survival Curves Using Contingency Tables" (2022). Dissertations. 3915.