MODELING LIFETIME DATA WITH MULTIPLE CAUSES USING CAUSE SPECIFIC REVERSED HAZARD RATES

In this paper we introduce and study cause specific reversed hazard rates in the context of left censored lifetime data with multiple causes. Nonparametric inference procedure for left censored lifetime data with multiple causes using cause specific reversed hazard rate is discussed. Asymptotic properties of the estimators are studied. Simulation studies are conducted to assess the efficiency of the estimators. Further, the proposed method is applied to mice mortality data (Hoel (1972)) and Australian twin data (Duffy et al. (1990)).