This book is intended to be read sequentially, as each chapter builds on the material that comes before. For example, Part I should be read by all, since it includes an overview, a formal statement of the ecological inference problem, and the notation used throughout the book. Nevertheless, for a first reading, some readers may wish to skip certain passages by following one of the roadmaps provided here.

Although Part II introduces several new results and provides motivation for many features of the solution offered in this book, readers uninterested in prior approaches to ecological inference may wish to skim this part by reading only pages 39-45 and Section 3.5 (on page 55) in Chapter 3, along with the indented, italicized remarks in Chapter 4.

Those readers who wish a quicker introduction to the proposed methods should read Part I, and skim Part II as described above. Then, a brief summary of the most basic form of the statistical model requires the information about the data and bounds in Chapter 5 (especially the explanation of Figure 5.1), and the introduction to the model and interpretation on pages 93-98 in Chapter 6. See also Chapter 10 for an application.

All readers should be aware that the solution to the ecological inference problem put forth and verified in this book is more than the basic statistical model that lies at its core. It also includes various extensions to avoid specific problems, a variety of new diagnostic procedures, graphical techniques, and methods of interpretation. Each of these, discussed in the rest of the book, is an integral part of making valid inferences about relationships among individual variables using only aggregate data. Many of these features of the methodology are demonstrated during the verification of the method in Part IV. Especially important is Chapter 16, which provides a checklist for those who intend to use these methods.

** Next:** Background
**Up:** Preface
** Previous:** Outline