INFORMATIVE EXPERIMENTAL-DESIGN FOR ELECTRONIC-CIRCUITS

This paper presents the case against two widespread practices in desig ning engineering experiments, which are (i) to vary one factor at a ti me (the OFAT approach) and (ii) to generate experimental points by ran dom selection (the Monte Carlo approach). These approaches do not prod uce good experimental designs, defined as generating maximum informati on per run (IPR), and should be replaced by designs that do. These lat ter designs (i) vary many factors at a time and (ii) use a patterned s et of experimental points rather than a random set. An example from ci rcuit design is used to illustrate this approach. The limitations of t he random approach are well known amongst statisticians but often not among engineers. (C) 1998 John Wiley & Sons, Ltd.