STA 490/690: Analysis of Time-to-Event Data

Time-to-event data arise in many common settings, for instance in modeling time-to-graduation, or time-to-death, or any other similar outcomes. Core to these data is the concept of censoring, which represents a unique type of missingness requiring careful consideration. For instance, we might know that a patient has survived at least 3 years after a particular surgery, but might not know the exact time it takes for them to die. This special topics course will focus on foundational concepts in the analysis of such data, including empirical likelihood estimators, the log-rank test, and parametric and semi-parametric regression models.