1134 - Lu Wang

Analysis of Clustered Event Times With Right-Censoring

Motivated by an infectious disease study at the BC Centre for Disease Control, this project is concerned with clustered event times where the observation is subject to right-censoring, and the cluster size is random. We formulate the dependence of the event times within each cluster with a copula model, and assume a parametric survival model for the margins. Inference on the model parameters are made via MLE (maximum likelihood estimation). In addition, we explore patterns of the cluster sizes and their association with the individuals who define the clusters. The motivating infectious disease study is used throughout this project to illustrate the research.