Introduction

This introduction presents an overview of key concepts covered in the subsequent chapters of this book. The book is concerned with the analysis of a data matrix when some of the entries in the matrix are not observed. It considers basic approaches, including analysis of the complete cases and associated weighting methods, and methods that impute (that is fill in), the missing values. The book also considers more principled approaches based on statistical models and the associated likelihood function, and provides applications of these methods. The generally preferred philosophy of inference can be termed “calibrated Bayes,” where the inference is Bayesian, using models that yield inferences with good frequentist properties. For example, 95% Bayesian credibility intervals should have approximately 95% confidence coverage in repeated sampling from the population. The method of multiple imputation has such a Bayesian justification but can be used in conjunction with standard frequentist approaches to the complete-data inference.