Practical Data Analysis for Designed Experiments

This book is aimed at statisticians and scientists who want practical expe rience with the analysis of designed experiments. My intent is to provide enough theory to understand the analysis of standard and non-standard experimental designs. Concepts are motivated with data from real experi ments gathered during over a dozen years of statistical consulting with scientists in the College of Agriculture and Life Sciences, augmented by teaching statistics courses on the 'Theory and Practice of Linear Mod els' (Stat 850) and 'Statistical Consulting' (Stat 998) at the University of Wisconsin-Madison. Students and colleagues have taught me much about what I tend to assume and about how to blend theory and practice in the classroom. I had hoped to find a textbook geared to this subject. I began by using Scheffe's Analysis of Variance to establish the theoretical framework, and Milliken and Johnson's Analysis of Messy Data to provide the practical guidelines. What I wanted was half-way in-between. Searle's Linear Models for Unbalanced Data has much the flavor I desired, but seems too detailed in some aspects for the classroom setting. Several other texts have noteworthy strengths, in particular Neter, Wasserman and Kutner's Applied Linear Statistical Models {3rd edn., Irwin, Boston, 1990), but do not cover the material with my preferred emphasis. This book can be used as a first or second semester text on linear models.