Abstract:
|
This paper explores the use of time-to-event models under right censoring in the Generalized Difference estimator framework. Model-assisted estimation has been widely used for sample surveys. One class of model-assisted estimators is Generalized Difference estimators. These estimators have been explored for Generalized Linear Models but not time-to-event models under right censoring. Adaptations to the current theory need to be made through the use of a counting process approach to time-to-event analysis. Here, both parametric and semi-parametric time-to-event models will be explored.
|