ABSTRACT
To evaluate the performance of a control chart, one of the key elements is the average run length (ARL), which is difficult to calculate. However, if the underlying observations can be embedded into a finite Markov chain, then an exact ARL can be found if the observations are discrete, and approx imations are available if they are continuous. In this chapter I provide a systematic review of many quality-related topics in situations in which a finite Markov chain can be employed.
