Bhote, Keki R. (1990, ASQC) Motorola Inc., Schaumburg, IL
The orthogonal array, developed by Taguchi, is difficult to understand ,time-consuming and costly. Most important, it can give poor results if interaction effects between input variables are present. To a lesser extent, these shortcomings are also present in the classical approach to the Design of Experiments (D.O.E.).
The author and his company, Motorola--one of the first companies to win the Malcolm Baldrige National Quality Award -- have used, instead, techniques formulated by Dorian Shainin for D.O.E. This paper briefly compares all three approaches to D.O.E. and shows why the Shainin techniques are simpler, less costly and statistically more powerful than the other two. There are at least seven Shainin tools, each applicable in solving a unique quality problem. Two of these -- the MultiVari and Variable Search -- are described in detail, with actual examples to demonstrate their power and ease of implementation.
Comparison,Shainin techniques,Design of experiments (DOE),Taguchi method,Electronics industry