EFFECT OF PROCESS FAILURE MECHANISMS ON ECONOMIC X¯ CONTROL CHARTS

Abstract

There has been considerable research on the economic design of Statistical Process Control models, for example on x¯ charts. Much of this research has assumed either a discrete event or continuous time process. However, virtually no research has been reported on continuous time models which formally investigate the effects of the choice of process failure mechanism on its design parameters. Using a continuous time model, we show empirically that such model design parameters are affected by the choice of process failure mechanism. The significance of this influence depends on the size of a process shift that one wishes to detect. The smaller the shift, the more significant the effect on the design parameters. For process shifts of 1 standard deviation or more, the design parameters are essentially robust with respect to choice of process failure mechanism.