### Introduction

In Sections 44.1 and 44.2 we have considered how to analyse data from experiments of certain types. Of course, before we can analyse any data we must conduct the experiment and before we can conduct an experiment we must design it. The work of applying statistical ideas to engineering experiments does not begin with the analysis of data. It begins with the design of the experiment. It is important to give proper consideration to experimental design to make sure that our experiment is efficient and that it will, in fact, give us the information we require. A badly designed experiment may give poor or misleading results or may turn out to be an expensive waste of time and money.

#### Prerequisites

• explain the concepts and terminology of the one-way and two-way ANOVA
• be familiar with the $F$ -distribution
• understand the general techniques of hypothesis testing

#### Learning Outcomes

• explain the basic concepts of experimental design
• apply randomized blocks and Latin Square designs
• analyse the results from randomized blocks and Latin Square designs