Introduction

This introduction presents an overview of the key concepts discussed in the subsequent chapters of this book. The book addresses methods of statistical analysis involving distinct, sometimes complex distributional assumptions that are necessary for integrated data obtained from multiple sources. It discusses a somewhat different tack when considering how inference should be carried out with data that have either been fused, or where missingness makes key parameters non-identifiable. The book describes a wide variety of issues that arise when integrated register data are used for statistical purposes. It shows that the set can be made more compact by inclusion of extra-sample information, particularly in the form of logical constraints about parameter values. The book explains the research aimed at the identification and fitting of models for population-size estimation that address this reality, as well as the design of data collection strategies.