Premium
Case‐Weighted Measures of Influence for Proportional Hazards Regression
Author(s) -
Pettitt A. N.,
Daud I. Bin
Publication year - 1989
Publication title -
journal of the royal statistical society: series c (applied statistics)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.205
H-Index - 72
eISSN - 1467-9876
pISSN - 0035-9254
DOI - 10.2307/2347680
Subject(s) - statistics , regression dilution , mathematics , regression , regression analysis , proportional hazards model , econometrics , polynomial regression
SUMMARY In this paper we consider the development of diagnostics and the assessment of case influence for the proportional hazards model of Cox. We critically examine various residuals which have been defined for this model and suggest that, instead of using these to assess certain model assumptions, diagnostics based on case‐weighted measures of influence should be used routinely. The ideas of Cook on local influence are applied to the proportional hazards model and useful statistics and plots are derived. We show that certain plots involving these residuals and diagnostics will in most cases be dominated by deterministic pattern and suggest their use with caution. Other plots are recommended and shown to be useful. Two data sets are considered to illustrate the techniques.