The proportional hazards model is one of the most common model for analyzing survival data. Only proportional hazards assumption is required to apply this model. Using appropriate sampling methods is an important part of modelling data and estimation of parameters. In literature there is a few studies based on sampling methods in survival analysis and most of them are related with non-parametric estimations of survival functions, sample size calculation etc. The main innovation of our approach is to examine the sampling methods for the proportional hazards model. This paper describes usage of ranked set sampling design in the proportional hazards model. In order to analyze the performance of our methods, we use a real data and conduct a simulation study. We conclued that ranked set sampling is more efficient than simple random sampling.
Mathematics Subject Classification (2010). 62N01, 62D05