Regression Analysis Of Doubly Truncated Data With Nonignorable Missing Covariate

Doubly truncated data exist in many fields such as astronomy,economic and medicine.It has high academic research value.Based on the standard linear model,we consider the nonignorable covariate missing mechanism and extend the Mann-Whitney-type estimation equation in this paper.We estimate the parameters by the inverse probability weighting method.Then,we introduce the standard linear model with symmetric random effect.Furthermore,we prove the consistency and asymptotic normality of parameters estimation.Through numerical simulation,the results show that our method is effective in parameter estimation.Finally,we apply this method to a real survival data.