Science is built upon a large body of reliable facts.
A common method for establishing facts is the experimental method.
An experiment consists of a carefully worked-out and executed plan for data collection and analysis.
The plan is based on contrasting two treatment conditions. Subjects are treated identically, except for one feature that is different. Thus observed differences in behavior can be unambiguously attributed to critical differences among treatment conditions.
A well-designed experiment permits the inference of causation.
An experiment design permits us to make a small number of observations, but infer facts that would normally require an exhaustive set of observations.