The merits of a bare bones approach

According to its preface, this book is intended for "students in universities, polytechnics and colleges of education". Polytechnics became universities in 1992, which made me check the book's publication details more closely. It was published first in 1986 and republished with corrections in 2005. I mentioned the book to Peter Nuesch, the distinguished Swiss statistician of EPF-Lausanne, who discovered a review of the first edition in the Mathematical Gazette . This referred to errors in the text, both minor and critical, and a number of these remain. For example, the Newton-Raphson formula is still given incorrectly, and there are sign and bracket errors presumably remaining from the original edition.

Although this criticism sounds negative, the book is not all bad. Probability and Random Variables was written before the personal computer created opportunities for investigation of the subject. However, there will always be a need for clearly written "no-frills" texts that explain underlying mathematical principles, and this book does exactly that.

A particular strength is that at the end of each chapter there is a section devoted to "Brief solutions and comments on the problems", which will be of assistance to those studying the material independently.

The selection of material is largely standard, and the previous reviewer commented on the good, if rather brief, treatment of estimation and on the lack of discussion of significance testing. However, when the text was first published, the short final chapter on the generation of random variables was a novelty that made the book stand out. It would have been valuable to have updated and extended these chapters. A few references are given, but there have been many worthwhile books on elementary probability published since 1985 that are not referred to.

In the final analysis, this is a good book, particularly as a library reference. For this purpose, as Nuesch commented to me, some careful rereading and correcting would have rendered a solid text an excellent one.

Perhaps the publisher will now produce a fully updated new edition? I believe the effort would be worthwhile.

Nigel Steele is emeritus professor of mathematics, Coventry University.