In statistical process control it is usually assumed that the observations
taken from the process of interest are independent, but in practice the obs
ervations in many cases are actually autocorrelated. This paper considers t
he problem of monitoring a process in which the observations can be represe
nted as a first-order autoregressive process plus a random error. The probl
em of detecting special causes which may produce changes in the process mea
n and/or variance is considered. Several types of control charts and combin
ations of control charts are evaluated for their ability to detect changes
in the process mean and variance. Some of these control charts plot the ori
ginal observations and have control limits adjusted to account for the auto
correlation in the observations, and others plot the residuals from the for
ecast values of a fitted time series model. The results of these investigat
ions show that there is no combination of charts that gives optimal perform
ance across a wide variety of situations, but, for reasonably good overall
performance, an exponentially weighted moving average chart of the observat
ions used with a Shewhart chart of the residuals can be recommended for pra
ctical applications.