Wheeler, Donald J. (1991, ASQC)
This abstract is an edited version of the author's original.
Several myths about control charts hinder and inhibit their effectiveness. Among these myths are mistaken notions about the need for normally distributed data, the role of the central limit theorem, the independence of observations, and the right way to compute control limits.
In contrast to these myths, four foundations of Shewhart's control charts are listed and explained. These foundations are: (1) the use of three-sigma limits, (2) the computation of limits based upon the average variation within the subgroups, (3) the use of rational sampling and rational subgrouping, and (4) the importance of the rational utilization of the knowledge gained from the charts.
To further illuminate the misconceptions surrounding Shewhart's control charts, the different ways that Shewhart's charts may be used are explained. These uses include the evaluation of process trials, control charts as report cards, using control charts for process adjustments, using multiple control charts for extended monitoring of many variables, and the use of control charts for the continual improvement of processes and product.
Finally, Cusum techniques and Exponentially Weighted Moving Averages (EWMA) techniques are compared with the control chart. While these techniques may be substituted for a control chart for some uses, they are poor substitutes when it comes to other uses.
Central Limit Theorem,Control charts,Cumulative sum control chart (CUSUM),Exponentially weighted moving average control charts (EWMA),Shewhart control chart,Statistical process control (SPC),Statistics