ABSTRACT
This chapter examines the relationship between the survival time Y and values of an explanatory variable X. For example, if Y denotes 'survival time following heart transplant', X may denote 'age'; the aim is to use X in modelling for Y. In this general framework for the regression of Y on X, other terms sometimes used to describe X are covariate and regressor variable. The chapter shows that in an accelerated lifetime model, the logarithm of lifetime is a linear function of x. This naturally occurs when lifetimes follow a Weibull distribution. Compare the standard deviations of the data in each group. The chapter present two types of modelling processes — 'proportional hazards' and 'accelerated lifetimes' and shows that only in the case of the Weibull models do these two kinds of modelling processes coincide.
