Experiments can be complex, but some core concepts and methods can be introduced in the context of extremely simple experiments in which:
- Nothing is known about the characteristics of the experimental units before the experiment is started.
- The researcher will only adjust the levels of a single controlled factor.
In this chapter, we describe:
- the best design for such an experiment (random allocation of treatments to experimental units)
- a model for the response
- summarising the effect of the factor with parameter estimates
- analysis of variance to test whether the factor does affect the response.
These methods will be the basis of design and analysis for more complex experiments in later chapters.