ARL properties of a sample autocorrelation chart

There are several statistical process control (SPC) methods for detecting the presence of special causes of variation when process observations are inherently autocorrelated. Most of these methods, however, focus on studying changes in the mean or variance of a time series as a signal of the presence of these special causes. It is seldom emphasized in the quality literature that such causes of variation are manifested not only by changes in the mean or variance of a time series but also by the changes in its stochastic behavior. A method that specifically focuses on monitoring this tupe of change is the sample autocorrelation chart (SACC). The SACC is claimed to be capable of detecting changes in mean, variance and stochastic behavior of a series, but no detailed studies have been reported concerning such properties. In this paper, we conduct Monte Carlo experiments to analyze the average run length (ARL) properties of the SACC. The results show that, in comparison with the existing techniques for monitoring autocorrelated processes, the SACC is less sensitive in detecting mean and variance shifts but very competitive in detecting changes in the parameters of an ARMA model.