ABSTRACT
The graphical display of data plays an important role in the de tection of changes in the level of a variable characterizing a process. In fact, in many practical instances with a single attribute monitored, sta tistical quality control consists mainly of plotting the data on a control chart assessing levels and/or trends in the variable based on the graph ical display, and taking action when it is determined that an assignable cause is present. Furthermore, even in the cases when several attributes are monitored, the most common practice is to monitor them on several separate displays. Thus, although we focus on multivariate observations, it is important to review briefly the univariate control charts which in many practical cases will at least supplement the multivariate charts.
