6. Process or Product Monitoring and Control
6.5. Tutorials
6.5.4. Elements of Multivariate Analysis

6.5.4.4.

Hotelling's T2 Chart

Continuation of example started on the previous page

STEP 6. Plot the T ² 's against the sample numbers and include the UCL.

Mean and dispersion
T ² control charts

The following figure displays the resulting multivariate control charts.

Interpretation of control charts

Inspection of this figure suggests an out-of-control conditions at sample numbers 1, 2, 9, 10, and 11 since these points exceed the T-square for means UCL of 14.52. Fortunately, the process was apparently back in control after the periods 9-12.

The problem may be caused by one or more univariates, and it could be easier to remedy than when the covariance plays a role in the creation of the undesired situation. To study the cause or causes of the problem one should also contemplate the individual univariate control charts.