Assessing Component Effects in Formulation Systems

ABSTRACT

The purpose of Snee (2011) was to discuss and illustrate how using Six Sigma concepts and approaches can lead to a better understanding of formulation systems. Central to the recommended approach is estimating and interpreting the effects of the components in the formulation on the response variable. Knowledge of components that have no effect, small or large effects (positive or negative), and similar effects is very helpful in understanding how the formulation system works. This understanding helps one decide (1) which components having large effects should be increased or decreased and (2) what to do about components with no effects or similar effects.

Correspondence between the authors following Snee (2011), along with further review of the literature, has pointed out that the approaches discussed in Snee (2011), and the interpretation of component effects in general, need further clarification. That is the focus of this article.

KEYWORDS:

• component effects
• component effect directions
• formulation systems
• mixture experiment models
• mixture experiments
• Scheffé linear mixture model