正态检验也能成书，长见识。赶紧在Amazon试读了一下，STATISTICS:Textbooks and Monograps 部分列出了很多重要文献。忍不住分享一下Michael R. Chernick的书评：
It is designed for practitioners with most of the theory avoided but all the references necessary to learn the theory are given. The title does not do the book justice. He includes outlier detection (univariate and multivariate) and testing for mixtures of normals. Even influence function measures that do not necessarily relate to the Gaussian model are discussed. To my pleasant surprise some of my work on the use of influence functions to detect multivariate outliers is mentioned briefly.
A nice feature of the book is the thoughtfully chosen quotes from statisticians at the beginning of each chapter. The first one provides the theme for the book. It is a quote from Egon Pearson in 1935 where Pearson (who with Neyman worked on optimal tests) says that it is permissible for statisticians to use less than efficient tests when simplicity or ease of computation are a consideration.
