Next
Sampling / Steven K. Thompson. Thompson, Steven K., 1945- author.
New York : John Wiley & Sons, [1992]
xv, 343 pages : illustrations ; 25 cm.
Estimating an elusive human population, estimating the abundance of an animal population, predicting the amount of fossil-fuel resource at a new site, or estimating the prevalence of a rare disease requires statistical methods that go beyond the material traditionally found in sampling texts. Designed specifically for scientific researchers and others who use sampling, this important reference is the first to explore the host of sampling methods useful for investigating populations that are inherently difficult to sample, elusive, rare, clustered, or hard to detect. Balancing both classical and modern sampling methods, Sampling offers systematic, thorough, and broad-based coverage of basic and standard sampling design and estimation methods. Organized into six parts containing twenty-six chapters, the book is a comprehensive one-volume seminar on using sampling methods to develop effective sampling designs for real life situations. In a clear progressive format, the book examines basic sampling from simple random sampling to unequal probability sampling in Part 1. The use of auxiliary data with ratio and regression estimation is discussed in Part 2 as well as the ideas of sufficient data and of model and design in practical sampling. Part 3 covers major useful designs including stratified, cluster systematic, multistage, double, and network sampling. Part 4 examines detectability methods for elusive populations, focusing on detectability visibility and catchability as well as specific methods of line transects, variable circular plots, capture-recapture, and line intercept sampling. Spatial sampling is covered in Part 5, with discussion of prediction methods of geostatistics, efficient spatial designs, and comparisons of different observational methods including plot shapes and detection aspects. Part 6 introduces adaptive sampling designs, illustrating how the adaptive cluster sampling designs described are highly effective for sampling rare, clustered populations. Each chapter contains numerous exercises, designed to pique interest and foster technical mastery, followed by answers in the appendix. A text suitable for the diverse needs of researchers and students in statistics and mathematics as well as the natural and social sciences, Sampling is an unparalleled introduction to the host of sampling methods available for studying challenging, difficult to sample populations.