Non-Random Missingness in Categorical Data: Strengths and Limitations

Abstract

There have recently been substantial developments in the analysis of incomplete data. Modeling tools are now available for nonrandom missingness and these methods are finding their way into the broad statistical community. The computational and interpretational issues that surround such models are less well known. This article provides an exposition of several of these issues in a categorical data setting. It is argued that the use of contextual information can aid the modeler in discriminating among models that are indistinguishable purely on statistical grounds.