This dissertation investigates three interesting problems in clustered event history data. The first project considers a positive stable frailty proportional hazards model for clustered failure time data, with the frailty distribution allowed to depend on cluster-level covariates. We connect the...