The identifiability of the competing risks model

James J. Heckman, Bo E. Honoré

Research output: Contribution to journalArticle

166 Scopus citations

Abstract

This paper considers the consequences for identifiability of introducing regressors into the competing risks model of multistate duration analysis. We establish conditions under which access to regressors overturns the nonidentification theorem of Cox and Tsiatis for both proportional and accelerated failure time models.

Original languageEnglish (US)
Pages (from-to)325-330
Number of pages6
JournalBiometrika
Volume76
Issue number2
DOIs
StatePublished - Jun 1 1989
Externally publishedYes

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Mathematics(all)
  • Agricultural and Biological Sciences (miscellaneous)
  • Agricultural and Biological Sciences(all)
  • Statistics, Probability and Uncertainty
  • Applied Mathematics

Keywords

  • Competing risks model
  • Identifiability
  • Regressor

Fingerprint Dive into the research topics of 'The identifiability of the competing risks model'. Together they form a unique fingerprint.

  • Cite this