[.ca] Model Assisted Survey Sampling (ISBN 0387406204)

thorough and practical account of survey sampling:
This book provides a graduate level text for survey sampling based on models including superpopulation models. It is a very up-to-date and modern approach and includes material on resampling methods as well as all the classical material on random sampling, stratified sampling and sampling proportional to size. They also provide methods for handling regression problems in the finite population context and even provide names of available software packages. Important topics (particularly for censuses) such as measurement error, nonresponse and imputation methods are all nicely treated. The book contains some practical examples and several classroom exercises.

control for the probabilities of inclusion!:
This book has certainly changed my life, and probably will eventually turn the entire empirical economics priesthood on its head! This idea - that the way the sample was created is just as, if not MORE, important as how the estimate was created - is so obvious it's revolutionary! It means survey researchers have to pay close attention to inclusion probabilities from now on. I have to say I already consider the work a classic. If taken as a course with a good instructor, the book is approachable and very applied. I recommend it to anyone interested in doing economic, political, social, market, environmental, or population research. (Don't be afraid of the math - it's just the most elegant way to write the ideas down.)