Abstract
The on-line quality-monitoring procedures for attributes presented by Taguchi, Elsayed, and Hsiang are examined. Their method for obtaining monitoring strategies is compared with the work of Gibra and Barlow, Hunter, and Proschan. An economic model under a geometric process failure mechanism is developed, and a comparison is made between methods optimal under that model and the methods of Taguchi et al. In some cases, there are important practical differences in the two approaches. KEY WORDS: Economic model; Process failure mechanism; Statistical process control.