Basic textbooks treating survey sampling invariably begin with the simple random samples, a type of sample rarely seen in the conduct of actual surveys; often they progress no further. On the other hand, there are many books written for specialists which discuss in detail the issues regarding complex sample designs, but which require a statistical sophistication which many people working in survey research simply do not have. Statistics for Real-Life Sample Surveys was written to bridge the gap between those two presentations: it explains the logical, mathematical and practical aspects of the issues regarding non-random samples in a manner requiring no more than elementary statistical knowledge.
Statistics for Real-Life Sample Surveys performs two great services: it collects in one volume a great deal of up-to-date technical information about non-random sampling which has practical implications for the working researcher, and presents this material with admirable clarity. The least technically-skilled research assistant could benefit by reading the discussions of the issues involved, while those with more statistical background can use it as a handbook, consulting the sections relevant to a particular problem they are working on. Basic statistical formulas are presented in the text, although it is possible to comprehend the main points while skipping the formulas, while mathematical proofs are included in several appendices.
The chapters devoted to weighting and to the statistical effects of sampling and weighting are particularly useful since those topics are relevant to many sampling projects and yet are not covered sufficiently (if at all) in most textbooks. The section on evaluating statistical software, including a test which allows the user to infer how weights are handled within the software (basic information which, oddly enough, is not always clearly documented), is a model of common sense and practical utility.
Sergey Dorofeev and Peter Grant have over fifty years of commissioning and conducting survey research between them. Sergey Dorofeev is Technical Director of Roy Morgan International, the global arm of Roy Morgan Research, an Australian market and social research, public opinion and political polling organization. Peter Grant was Research and Development Manager in the Melbourne head office of Roy Morgan Research.
Sarah Boslaugh (email@example.com) is a Performance Review Analyst for BJC HealthCare and an Adjunct Instructor in the Washington University School of Medicine, both in St. Louis, MO. Her books include An Intermediate Guide to SPSS Programming: Using Syntax for Data Management (Sage, 2004), Secondary Data Sources for Public Health: A Practical Guide (Cambridge, 2007), and Statistics in a Nutshell (O'Reilly, forthcoming), and she is Editor-in-Chief of The Encyclopedia of Epidemiology (Sage, forthcoming).
Preface; 1. Sampling methods; 2. Weighting; 3. Statistical effects; 4. Significance testing; 5. Measuring relationships; Appendices; Bibliography; Index.