Key features of survival data. Types of censoring and truncationn. Definition of hazard and other basic concepts. Parametric models such as exponential and Weibull, and the general accelerated failure time model. Non-parametric methods: Kaplan-Meier and Nelson-Aalen estimators and the log-rank test. Semi-parametric methods: the Cox proportional hazards model and extensions to stratified models and time-dependent covariates. Hypothesis testing, residuals and influence, and checking the proportional hazards assumption. Introduction to counting processess. Introduction to frailty and other methods for multivariate survival data. Extensive use will be made of the statistical computing language R in analysing real datasets.

Note: Course not offered every year - contact School for more information.