Ken Russell, honorary professor in NIASRA, has written a book on the design of experiments when the data to be collected will be analysed by a generalized linear model (GLM).  The book concentrates on situations where the predictor variables are ‘interval’ or ‘ratio’ in nature.

There are numerous books on the analysis of data by GLMs, but it is believed that this is the first book written solely on the topic of design.  The target audience includes scientists as well as statisticians.  Unlike much of the other material on this topic, it is not assumed that the reader has a mathematics background roughly equivalent to Honours level.  Programs in R to perform the necessary calculations are included in the text or are available on a website of supporting material.

Chapter headings are 1. Generalized Linear Models;  2. Background Material;  3. The Theory Underlying Design; 4. The Binomial Distribution;  5. The Poisson Distribution;  6. Several Other Distributions;  7. Bayesian Experimental Design.

Design of Experiments for Generalized Linear Models” has been published by CRC Press in its Chapman & Hall/ CRC Press Interdisciplinary Statistics. For more details see: www.crcpress.com/9781498773133