SURVIVAL ANALYSIS IN HETEROGENEOUS POPULATIONS - STATISTICAL MODELS AND CONCEPTS

摘要: Publisher Summary This chapter describes the statistical concepts and models for analysis of survival data in heterogeneous populations. Standard techniques methods are not directly applicable to event history data, because censoring problem. Data said be censored if it is possible record exact time occurrence an each individual. There usually two reasons why involve observations. Ignoring observations analyzing only cases with information on timing events can cause serious biases. However, biases also arise from treating as though had occurred at last observation. Thus, models, which allow when histories, needed. A variety sophisticated procedures already exists, analyst able choose among various alternative estimation methods.