DISPERSION EFFECTS FROM UNREPLICATED DESIGNS IN THE 2(K-P) SERIES

Quality improvement requires reduction of undesired variation. Design of experiments has become a useful tool for discovering such minimum v ariance conditions, but most existing techniques require a large exper imental effort involving replicated experiments. In this article we de velop a method for significance testing of dispersion effects from unr eplicated two-level fractional factorial experiments using principles close to those for identifying location effects. The method provides a n F-distributed test statistic and appears to be useful for identifica tion of dispersion effects as early as the screening stage when fracti onated unreplicated experiments are frequently employed.