John Wiley & Sons
2012 | 275pp | £90 (HB)
Experiment! is perhaps the most informative and guiding book ever written for the practising experimentalist. Although it is aimed at PhD students, it is relevant for anyone involved in research. Indeed, it is going to be of great value to me for years to come, regardless of my 15 years as a practicing chemist post-PhD.
Rather than immersing the reader in the subject of scientific research as a task based, methodical discipline, where scientists and engineers venture into the unknown following a few simple rules, the approach here begins with historical reference and context. This description of the history of the scientific method and the struggle to establish a stable experimental method and framework firmly sets the foundations for the beginning of the book – Understanding the world – and is so rich in content that it could easily form a great book all by itself.
Part two of Experiment! then moves on to Interfering with the world, focusing on the role of statistics in modern experimental design and interpretation. This section would serve as a fine topic to be taught in all universities engaged in scientific training and research. There is much to be gained from a clear understanding of statistics in experimental design. In particular, as automated techniques produce a wealth of data that is becoming more and more overwhelming, statistics and statistical tools will undoubtedly be critical in translating data into useful knowledge.
Overall, Experiment! is an excellent book to read from a historical point of view but its strength clearly lies in the craftsmanship of its presentation. It is an incredibly well written account of modern scientific practice and how that is now designed and steered by statistical methods. I can highly recommend this book to all who are involved in scientific research and development.