Affordable Access

Publisher Website

Chapter 5 The Concept of Optimal Design

Elsevier Science & Technology
DOI: 10.1016/s0922-3487(08)70515-x
  • Design


Publisher Summary This chapter discusses the concept of optimal design. It is by no means easy to choose the best strategy for performing measures. The chapter examines the problem using a series of weighings. It is assumed that an investigator uses an old two-pan balance to weigh two objects, A and B, having masses of ma and mb. The investigator makes two weighings and obtains two results with an error of ±σ. The chapter also discusses criteria of optimality of experimental designs. Four criteria are mentioned, and two of them are applied to the weighing design and resistance measurement examples. A research strategy can be evaluated, and there is an optimum strategy. There is a best strategy: the optimal strategy. The positioning of experimental points within the experimental domain is of great importance to obtain the best precision on the effects and interactions. When only two trials are performed per factor, the experimental points must lie at the limits of the domain in such a fashion that all factors are involved in all trials. This is the case for the factorial designs; the investigator who uses such designs can be sure of having the best possible strategy.

There are no comments yet on this publication. Be the first to share your thoughts.