Detecting selected non-random patterns with individuals control charts

Western Electric (1956) and Nelson (1984) proposed rule sets to detect visually obvious process upsets on a control chart. Performance evaluation of these rules require simulating situations that correspond their intended detection purposes such as linear drift, cycling, seesaw sustained mean shifts. These non-random patterns become in the presence reduced variation. Contrary previous assessments where variance is assumed constant, are shown be preferred over Shewhart X CUSUM charts for detecting reduction wide ranges slopes, cycle period/amplitude combinations, alternating shift sizes. One real-data example from Deming's book Out Crisis provided affirms implications extensive simulation analyses.