Estimating the association parameter for copula models under dependent censoring

Weijing Wang*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

66 Scopus citations


Many biomedical studies involve the analysis of multiple events. The dependence between the times to these end points is often of scientific interest. We investigate a situation when one end point is subject to censoring by the other. The model assumptions of Day and co-workers and Fine and co-workers are extended to more general structures where the level of association may vary with time. Two types of estimating function are proposed. Asymptotic properties of the proposed estimators are derived. Their finite sample performance is studied via simulations. The inference procedures are applied to two real data sets for illustration.

Original languageEnglish
Pages (from-to)257-273
Number of pages17
JournalJournal of the Royal Statistical Society. Series B: Statistical Methodology
Issue number1
StatePublished - 1 Oct 2003


  • Archimedean copula models
  • Bivariate survival analysis
  • Competing risk
  • Cross-ratio function
  • Estimating function
  • Frailty models
  • Identifiability; Kendall's τ
  • Log-rank statistic
  • Multistate process
  • Semi-competing-risks data
  • Semiparametric inference


Dive into the research topics of 'Estimating the association parameter for copula models under dependent censoring'. Together they form a unique fingerprint.

Cite this