Dissertation/Thesis Abstract

Comparison of Cox regression and discrete time survival models
by Ye, Hong, Ph.D., Wayne State University, 2016, 188; 10153426
Abstract (Summary)

A standard analysis of prostate cancer biochemical failure data is done by conducting two approaches in which risk factors or covariates are measured. Cox regression and discrete-time survival models were compared under different attributes: sample size, time periods, and parameters in the model. The person-period data was reconstructed when examining the same data in discrete-time survival model. Twenty-four numerical examples covering a variety of sample sizes, time periods, and number of parameters displayed the closeness of Cox regression and discrete-time survival methods in situations typical of the cancer study.

Indexing (document details)
Advisor: Sawilowsky, Shlomo
Commitee: Castronova, Frank, Piliawsky, Monti, Stevens, Craig
School: Wayne State University
Department: Educational Evaluation and Research
School Location: United States -- Michigan
Source: DAI-A 77/12(E), Dissertation Abstracts International
Subjects: Educational tests & measurements, Education
Keywords: Comparison, Cox regression, Discrete-time survival, Prostate cancer
Publication Number: 10153426
ISBN: 9781369085617
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy