The main goal of this research project is to develop efficient statistical techniques for regression analysis of time-to-event data with complex censoring such as current status data or interval-censored data. Specifically, the investigator will study three types of problems that often occur in medical and public health researches. One is to develop efficient score function-based estimation procedures for univariate current status or interval-censored data that apply to more general situations and give practitioners more flexibilities in analyzing these data. The second is to develop efficient score function-based estimation procedures for bivariate current status or interval-censored data and the third is to develop efficient sieve maximum likelihood estimation procedures for bivariate current status or interval-censored data. The second and the third parts will investigate the same problem but employ different modeling and inference strategy to allow one to focus on different aspects of the data analysis or different questions.
The results of this study will advance not only the theoretical research in statistics pertaining to survival analysis but also provide statistical methods for data analysis in a wide range of scientific disciplines including public health, medicine, demographics, economics and psychology.
