Data Analysis by Resampling: Concepts and Applications

Rating:
Summary: Applied resampling techniques
Review: I took several multivariate analysis courses from Dr. Lunneborg as a grad student at the University of Washington and they were pretty difficult being filled with matrix algebra and derivations. With that background, I was pleasantly surprised to find this book so well organized and applied. I do a great deal of resampling in my research and Dr. Lunneborg has done an excellent job of summarizing the various areas where resampling can save your butt, and where it can kick you in the butt if you are not careful. He provides the algorithms for Resampling Stats (a major resampling software package) and S-plus. I would have appreciated if he had included the code for Sas as well, but in most cases you can easily back it out from the S-plus code. If you are a student or an applied statistician and want to either learn how to use resampling techniques or actually apply it in your work, then this is an excellent book. If you are more mathematically oriented, then you would be better off going to the technical journals and reading the original works by Efron et al. to understand the logical, statistical and mathematical bases of this methodology. I have used the Resampling Stats Excel add-in for several years, so it was very useful to find a book that provides the algorithms for this software.

Rating:
Summary: Applied resampling techniques
Review: I took several multivariate analysis courses from Dr. Lunneborg as a grad student at the University of Washington and they were pretty difficult being filled with matrix algebra and derivations. With that background, I was pleasantly surprised to find this book so well organized and applied. I do a great deal of resampling in my research and Dr. Lunneborg has done an excellent job of summarizing the various areas where resampling can save your butt, and where it can kick you in the butt if you are not careful. He provides the algorithms for Resampling Stats (a major resampling software package) and S-plus. I would have appreciated if he had included the code for Sas as well, but in most cases you can easily back it out from the S-plus code. If you are a student or an applied statistician and want to either learn how to use resampling techniques or actually apply it in your work, then this is an excellent book. If you are more mathematically oriented, then you would be better off going to the technical journals and reading the original works by Efron et al. to understand the logical, statistical and mathematical bases of this methodology. I have used the Resampling Stats Excel add-in for several years, so it was very useful to find a book that provides the algorithms for this software.

Rating:
Summary: The latest elementary account on resampling methods
Review: Professor Lunneborg covers bootstrap methods, permutation methods and subsampling methods and contrasts them in terms of the sampling design. This is a good introductory text at a fairly elementary level. Like Efron and Tibshirani (1993), Davison and Hinkley (1997) and Chernick (1999), he emphasizes the value of resampling in the age of modern fast computing and explores the variety of applications. This book is unique in that it could be used as an introductory text for students with only high school algebra. It also views the appropriateness of methods according to the experimenters sampling design. Use bootstrap when the data constitute random samples, permutation methods in randomized designs such as randomized trials and subsampling for non-random studies. While this is an interesting way to view the methods it is not universally accepted and both the bootstrap and the permutation tests have been applied in wider contexts.