A simple data transformation of auto-correlated data for SPC

Abstract

Statistical process control procedures that use one-step-ahead forecasts, including special cause control charts, tend to have outstanding average run length properties when applied to a negatively auto-correlated data. However, nearly all industrial data, which is auto-correlated, exhibits a positive auto-correlation structure. In this paper, a simple transformation on a positively auto-correlated data is proposed which results in a data set with the same mean and nearly the same variance. The resulting auto-correlation coefficient is opposite in sign and approximately equal to that of the original data set. Two simulated examples using the special cause control chart are used to illustrate the proposed procedure.

Keywords:

• Auto-correlated data
• Special cause control chart
• Statistical process control

Acknowledgements

The author wishes to gratefully acknowledge the referees of this paper who helped to clarify and improve the presentation.

Notes

Email: [email protected]

Additional information

Notes on contributors

F. K. Wang

Email: [email protected]