Second edition

This edition differs from its predecessor in several important ways:

  • A holistic approach to linear models vs. a box of named tools
  • We have brought the estimation of effects very much to the front
  • P-values remain one way of identifying signals, but not the only one, and we have tried to illustrate parallel approaches

In the first edition, we were agnostic concerning the choice of software packages. Biologists used a wide range of primarily commercial packages with different interfaces. We used one (SYSTAT), but we didn’t provide code. The world has moved on, and most biologists now use the package R because it’s free and supported by a massive user community. We’ve decided to analyze all our worked examples using R

We’ve provided lots of worked examples taken from the biological literature. For all our worked examples, we’ve run our own analysis. The website will include the code for all these examples, including lots of the exploratory steps and a link to the data and the original paper. The website will also include additional worked examples suitable for teaching, plus additional exercises.

Table of Contents

  1. Introduction
  2. Things to know before proceeding
  3. Sampling and Experimental Design
  4. Introduction to Linear Models
  5. Exploratory Data Analysis
  6. Simple Linear Models with One Predictor
  7. Linear Models for Factorial (Crossed) Designs
  8. Multiple Regression Models
  9. Predictor Importance and Model Selection in Multiple Regression Models
  10. Random Factors in Factorial and Nested Designs
  11.  Split-plot (Split-unit) Designs: Partly Nested Models
  12. Repeated measures designs
  13. Generalized Linear Models for Categorical Responses
  14. Introduction to Multivariate Analyses
  15. Multivariate analyses based on eigenanalyses
  16.  Multivariate analyses based on similarities or distances
  17. Telling stories with data



List of Acronyms

There’s a more detailed ToC here

Chapter 12 online material examples (not publicly available)