MULTI-VARI CHARTS AND STATISTICAL PROCESS CONTROL
Multi-vari charts are not part of the classic tools for data analysis but they are conceptually very simple and in many cases they are the best way to identify the different sources of process variability. Naturally, and once again, for this to work, data should have been collected in an appropriate manner. Multi-vari charts are usually considered an exploratory data analysis technique, but we have preferred to place it in Part 4, along with control charts, because it shares with them the objective of fighting against variability.
Statistical Process Control (SPC) is one of the best known and more widely used quality tools. Its aim is to keep the process working at its ‘normal state’ – with only random variability causes affecting it – by releasing alarm signals when something strange – a special cause of variability – is disturbing the process. In this sense they are usually considered a control tool, but it is clear that they are also – and perhaps mainly – an improvement tool, as the opportunity is given to learn from the process, by showing what causes affect it and in which way; thus, providing the information needed to improve the process by incorporating the good and eliminating the harmful ones.
The generic name of SPC is an umbrella for several types of control charts that all work with the same general principle stated above. Control charts are classified into two big groups depending on whether they are aimed ...