Introduction

Statistics is often used for more than describing a dataset — one of its most powerful applications is estimating the truth about an entire population from just a small sample of it. This process is called statistical inference. It only works when we can trust that the sample genuinely reflects the population it was drawn from — the same idea behind an exit poll: pollsters interview a few thousand voters on election day and use that small sample to predict, often with striking accuracy, how millions of people actually voted. This chapter builds the toolkit for that kind of reasoning — how to …