Title

Full likelihood inferences in the Cox model: an empirical likelihood approach

Authors

Authors

J. J. Ren;M. Zhou

Comments

Authors: contact us about adding a copy of your work at STARS@ucf.edu

Abbreviated Journal Title

Ann. Inst. Stat. Math.

Keywords

Right censored data; Empirical likelihood; Maximum likelihood estimator; Partial likelihood; Profile likelihood; CENSORED SURVIVAL DATA; REGRESSION-MODELS; LARGE SAMPLE; EFFICIENCY; ESTIMATORS; Statistics & Probability

Abstract

For the regression parameter beta (0) in the Cox model, there have been several estimators constructed based on various types of approximated likelihood, but none of them has demonstrated small-sample advantage over Cox's partial likelihood estimator. In this article, we derive the full likelihood function for (beta (0), F (0)), where F (0) is the baseline distribution in the Cox model. Using the empirical likelihood parameterization, we explicitly profile out nuisance parameter F (0) to obtain the full-profile likelihood function for beta (0) and the maximum likelihood estimator (MLE) for (beta (0), F (0)). The relation between the MLE and Cox's partial likelihood estimator for beta (0) is made clear by showing that Taylor's expansion gives Cox's partial likelihood estimating function as the leading term of the full-profile likelihood estimating function. We show that the log full-likelihood ratio has an asymptotic chi-squared distribution, while the simulation studies indicate that for small or moderate sample sizes, the MLE performs favorably over Cox's partial likelihood estimator. In a real dataset example, our full likelihood ratio test and Cox's partial likelihood ratio test lead to statistically different conclusions.

Journal Title

Annals of the Institute of Statistical Mathematics

Volume

63

Issue/Number

5

Publication Date

1-1-2011

Document Type

Article

Language

English

First Page

1005

Last Page

1018

WOS Identifier

WOS:000293396600007

ISSN

0020-3157

Share

COinS