商品詳細

This text is both an introductory text for those new to the concepts and debates encountered in survey sampling and a balanced guide to the controversial question of survey sampling inferences. It tackles the question of whether sampling inferences should be made primarily in terms of inclusion probabilities, or whether instead they should be based on estimates of the parameters of a realistic population model. This text argues cogently that the two approaches are complementary rather than competitive, the former being appropriate for large samples and the latter for small ones. It also shows how they can neatly be combined. To achieve these ends, the title uses a story-telling style that aims to bring life to the tools used by the two approaches and shows how they can be fitted together. The early chapters draw on the fable of Basu's elephants as an illustrative example, while the late chapters broaden out to cover other instances of (perhaps more common!) popultions from businesses, households and everyday life. Exercises are included throughout so readers can put into practice the ideas encountered in each chapter. The back and forth discussions of the paradoxes that arise in survey sampling, depending on whether the design-based or model-based approach is used, should be illuminating not only to beginning students but also to experienced practitioners who have used these tools themselves, but perhaps without exploring all their ramifications.