Cross-sectional and Longitudinal Approaches in a Survival Mixture Model

In this article, we explore the performance of estimators in mixture survival model using simulated data recorded cross-sectionally and longitudinally. We use the maximum likelihood estimation approach in estimating the unknown model parameters. We found, in particular, that the maximum likelihood estimator for the proportion of long-term survivors in longitudinal setting gain better efficiency and precision for a certain distance of recording time.