Identification Of Graphical Models With Missing Data

Nonresponse mechanisms are the keypoints to the statistical inference on incomplete data. Identification of parameters in these models is an important and difficult problem. In this dissertation, we use decomposable graphical models to describe the mechanisms of nonignorable nonresponse in contingency tables classified by two binary variables, and we discuss identification of parameters in these models. For an unidentifiable model, we introduce covariates which are always observed such that these models become identifiable. Moreover, a simulated presentation of the method is provided for visual explanation.