in monetary units, to assure that both upper-level management and production management understand the concept. Thus, loss in the Taguchi context refers to product performance as a deviation from target performance (Atkinson, 1990, p. 17). The emphasis in quality improvement is on product design, process design, and tolerance design (Atkinson, 1990, p. 17). Product design should seek to satisfy end user requirements, while minimizing susceptibility to production process variations (Taguchi and Clausing, 1990, pp. 65-75). Process design depends upon the application of the findings of statistical experiments to identify the production parameters that will cause the least variation in product performance (Kackar, 1985, pp. 176-209). Tolerance design requires (1) the identification of those factors that most affect product performance, and (2) the determination for such factors of the tolerance levels that will assure product performance at the required level (Schmidt and Meile, 1989, pp. 49-56).
Dr. Genichi Taguchi studied classical statistics in England. Taguchi determined, however, that classical statistics were too theoretical for use in manufacturing (Walton, 1988, p. 123). Taguchi, thus, developed "a more practical approach for use in the design engineering phase to counteract . . . problems that were expensive and difficult to eliminate" (Walton, 1988, p. 123). Taguchi's statistical experiments rely heavily on factor analysis.
Factor analysis is a statistical procedure with data-reduction capabilities that is used to determine the underlying pattern of relationships among a set of variables or conditions which may be taken as source variables, accounting for the observed interrelations in the data. Factor loadings are the coefficients of factors identified in factor analysis, and are used as measures of the degree of the relationships between factors and variables.
Many research studies generate vast quantities of da...