Sometimes the only thing you can do with a poorly designed experiment is to try to find out what it died of.
—Ronald A. Fisher
In the last chapter we saw how to use statistical methods to find correlations and differences in data. Chapter 6 showed that experimental method helps us to connect effect with cause. Combining experimental and statistical method makes it easier to answer scientific questions both more economically and with better precision than otherwise. In this chapter, statistical approaches will be put into an experimental context. After a brief look at single factor experiments we will look at the very important scenario where several variables are suspected to simultaneously affect a response. The purpose is to explain some basic thoughts and introduce some classic designs using practical examples. Exercises should be completed when encountered in the text to confirm that you understand the methods before proceeding. Further reading is suggested at the end for those who want a more comprehensive treatment.