Design and Analysis of Experiments with R presents a unified treatment of experimental designs and design concepts commonly used in practice. It connects the objectives of research to the type of experimental design required, describes the process of creating the design and collecting the data, shows how to perform the proper analysis of the data, and illustrates the interpretation of results.
Drawing on his many years of working in the pharmaceutical, agricultural, industrial chemicals, and machinery industries, the author teaches students how to:
Make an appropriate design choice based on the objectives of a research project Create a design and perform an experiment Interpret the results of computer data analysis
The book emphasizes the connection among the experimental units, the way treatments are randomized to experimental units, and the proper error term for data analysis. R code is used to create and analyze all the example experiments. The code examples from the text are available for download on the author s website, enabling students to duplicate all the designs and data analysis.
Intended for a one-semester or two-quarter course on experimental design, this text covers classical ideas in experimental design as well as the latest research topics. It gives students practical guidance on using R to analyze experimental data.
This is a very clearly written book on experiment design, which is important whenever data collection is limited or expensive. It is not always an easy book to read, but that is because some of the concepts it presents are inherently complex; I found the book to be uniformly clear and well organized. I especially recommend this book for data scientists who run A/B tests - there is wisdom here which you probably haven't heard of and could use.
You might think that experiment design would be an easy topic which you should read about first, but this is not the case - this is actually a rather advanced book, and it will go better if you already have an understanding of multilevel regression before you read it. I would suggest Gelman and Hill as a minimum prerequisite.