Key points are not available for this paper at this time.
Survival data are analysed where it is assumed that the logarithm of the odds against survival beyond a certain time is equal to a linear term, involving covariates in a regression model, plus a function of time. It is shown that estimates for the regression parameters can be obtained by replacing survival times by their ranks. Comparisons are made with estimates found by Bennett (1983a, b), who used the same model but different techniques of estimation.
A. N. Pettitt (Sun,) studied this question.