A note on the data

Every dataset used in this book is either built into R, supplied with the book, or generated directly by code that you can read and run yourself. This means that every table, figure and numerical result can be reproduced by running the accompanying script from beginning to end.

Reproducibility is a fundamental principle of statistical work. If an analysis cannot be recreated from the original data and the accompanying code, its results cannot be independently verified. Good statistical practice therefore means leaving behind a complete record of every step that produced the final result.

A result that cannot be reproduced by running a script from top to bottom is not yet a result.

One dataset deserves a special mention. The class survey used throughout the book was collected from students enrolled in this course. Before it was included here, all identifying information was removed and each student was assigned an anonymous identification number. Protecting the privacy of participants is one of the first responsibilities of anyone working with data about people.

Later in the book we return to questions of sampling, data collection and research ethics. For now it is enough to remember that every dataset represents real individuals, and good statistical practice begins with treating their information responsibly.