The need to construct a physical setup gives rise to the question of cost. In many applications, the cost of instrumentation may also be quite high. Thus, the central issues such as experiments versus utilization of published data, simple and inexpensive versus complex and expensive experiments, short versus long, scaled versus prototype and detailed versus cursory experiments have to be carefully addressed. From an engineering point-of-view, economic viability will play a decisive role in the nature of experiment to be conducted.

NOTES:

1. The procedure that is needed to construct the experimental apparatus itself involves design.

2. The phrase design of experiments used in later sections refers not to that of the apparatus, but to the experimental procedure that will maximize the information derived.

3. Discussion on specific issues in the present chapter uses terminology that would be familiar to engineers, rather than statisticians.