Practical Methods for Design and Analysis of Complex Surveys (2nd ed.), Risto LEHTONEN and Erkki PAHKINEN, Chichester, U.K.: Wiley, 2004, ISBN 0-470-84769-7, viii + 349 pp., $89.95.
The first edition of this book was one of the first books in the excellent Wiley U.K. series on Statistics
The authors mention "improved usability" and "user feedback at meetings" (Preface, p. vii) as the principal reasons for updating their book. I am not sure that they have entirely accomplished the improved usability objective, however. They deleted a nice appendix that did a software review for survey analysis. They now handle "computational aspects" with and "have inserted technical materials" into a new "web extension of the book" (ibid.). They note that because of this feature, "examples and case studies can be worked out in an interactive environment." I think that a website with additional materials is great, but I am not enthusiastic about integrating a website and the book; after all, I am doing this report on a cross-Atlantic flight. The book has more than a dozen references to material and activities that should be pursued on the website. Despite this emphasis, the URL is not provided anywhere in the book. Buried deep in the index is this entry under "Web extension" (p. 345): "URL: See website of John Wiley & Sons Ltd." That is some rather petty marketing!
The primary addition to the book is a new chapter, "Model-Assisted Estimation for Domains," that discusses the methods for the estimation of the statistics for population subgroups. These estimates are focused on design-based procedures, which in general are the book's primary focus. Principles, estimators, and models are all discussed. The authors note (p. 207) that the relationship among the estimators is considered in more detail in the web extension of the book.
One new case study, "Monitoring Quality in a Long-Term Transport Survey" is added to the three case studies from the first edition. The authors present 20 quality indicators to monitor the data collection process that occurs from month-to-month for sampling and nonsampling errors. The recent book by Biemer and Lyberg (2003), reported by Ziegel (2004), focuses entirely on survey quality.
REFERENCES
Biemer, P., and Lyberg, L. (2003), Introduction to Survey Quality, Hoboken, NJ; Wiley.
Ziegel, E. (1996), Editor's Report on Practical Methods for Design and Analysis of Complex Surveys, by R. Lehtonen and E. Pahkinen, Technometrics, 38, 89-90.
_______ (2004), Editor's Report on Introduction to Survey Quality, by P. Biemer and L. Lyberg, Technometrics, 45, 277.