This second edition of the book first published in 1960 (see the review Zbl 0095.139) is approximately 200 pages larger than the first one. The authors claim that they have gone over the first edition word by word and paragraph by paragraph in a concentrated effort to expand the explanatory aspects and thereby simplifying the content. The book now contains 25 chapters, an exhaustive appendix of 24 tables and an index.
Among the new techniques added are Fisher’s least significant difference, Scheffé’s test and Walter-Duncan’s Bayesian k-ratio t test for multiple comparisons; Satterthwaite’s procedure for computing quasi F tests; the use of orthogonal polynomials to produce response surface equations in factorial experiments; and Kolmogorov-Smirnov one-and two-sample tests of goodness of fit.
The material on linear and multiple regression has been presented in matrix notation. The material of other chapters has been reorganized. Each chapter contains a number of exercises largely based on data from biological sciences. The book is nicely written and I hope it will be welcomed by biostatisticians and others.