Unavoidable variation in experimental units
Identical experimental units result in the most accurate estimates of the effect of a factor. In practice however, we usually have little choice and the available experimental units are often very variable.
We now illustrate this grouping of individuals into matched pairs.
Feed supplement and beef from cows
Matching of experimental units
In the above example, the experimental units were grouped into pairs using a numerical measurement — ages of animals or the previous year's yield from fruit trees could be used in a similar way. However pairing is often done in a less formal way — it is acceptable to construct the matched pairs of experimental units by any subjective or objective method.