“Design of experiments is a powerful tool for understanding systems and processes….Our view is that optimal design of experiments is an appropriate tool in virtually any situation that suggests the possible use of design of experiments,” the authors say in the book.
So what’s an optimal design?
“An optimal design is an experimental plan that maximizes the obtainable information in some sense. Optimal designs do not necessarily have to be computer generated. There are textbook designs that have been shown to be optimal. However, the textbook designs tend to place substantial restrictions that interfere with the ability to match the process under study with a design. By using a computer based approach, it is possible to make the design exactly match the process or system under study within the available resources and still optimize the information content of the experiment,” Jones explains.
The book is meant for both new and experienced DOE practitioners, and the authors hope it will be used as a handbook and will empower more people to experiment more.
Stu Hunter, professor emeritus at Princeton University and a preeminent industrial statistician, says, “Statisticians and para-statisticians alike should enjoy this book. Clearly, a new day is dawning in the art and practice of experimental design.”
The format of the book is appealing and accessible. Each chapter presents a case study, as the title of the book indicates. What’s unusual is that the details of each case study unfold in a dialogue between two consultants named Brad and Peter, who work for a fictitious consulting company called Intrepid Stats.
The case studies come from different industries, including biotechnology and consumer products. And the chapters cover such topics as comparative experiments, screening experiments, response surface experiment and mixture experiments.