Multivariate Control Charts Overview
data           

Display statistics from two or more related measurement variables. A multivariate chart shows how several variables jointly influence a process or outcome. For example, you can use multivariate control charts to investigate how the tensile strength and diameter of a fiber affect the quality of fabric.

If the data include correlated variables, the use of separate control charts is misleading because the variables jointly affect the process. If you use separate univariate control charts in a multivariate situation, Type I error and the probability of a point correctly plotting in control are not equal to their expected values. The distortion of these values increases with the number of measurement variables.   

Multivariate control charting has several advantages over creating multiple univariate charts:

·    The actual control region of the related variables is represented (elliptical for bivariate case).

·    You can maintain a specific Type I error.

·    A single control limit determines whether the process is in control.

However, multivariate charts are more difficult to interpret than classic Shewhart control charts. For example, the scale on multivariate charts is unrelated to the scale of any of the variables, and out-of-control signals do not reveal which variable (or combination of variables) caused the signal.

To determine whether you should use a univariate or multivariate chart, create a correlation matrix of your variables. If the variables are correlated, consider creating a multivariate control chart.  

For more information about control charts, see Control Charts Overview.

Choosing a multivariate control chart

Minitab offers four multivariate control charts:

·    T2 and Generalized Variance -  a T2  chart and Generalized Variance chart displayed in one window

·    T2 - a multivariate chart of means

·    Generalized Variance - a multivariate chart of process variability

·    Multivariate EWMA -  a multivariate chart of exponentially weighted moving averages