On the Performance of EWMA Chart in the Presence of Two-Component Measurement Error

ABSTRACT

Control charts are extensively used to monitor the stability of analytical processes. The presence of measurement error can seriously affect the outcome of any process and reduces the ability of control charts to detect a shift of a particular magnitude in the process parameters. Research has shown the superiority of exponentially weighted moving average (EWMA) control charts over Shewhart charts when it comes to quick and timely detection of small process shifts. In this study I examine the performance of the EWMA control chart in the presence of two-component measurement error due to its immense importance in analytical chemistry and environmental settings. It has been shown that effect of measurement error can be reduced by taking multiple measurements at each sample point.

KEYWORDS:

• analytical chemistry
• EWMA
• Monte Carlo simulations
• run length distribution
• two-component measurement error

ACKNOWLEDGMENTS

The author is grateful to the editor and referees for their constructive comments that led to substantial improvements in the article. Special thanks are also extended to Dr. Arden Miller, Department of Statistics, the University of Auckland, Auckland, New Zealand, for his valuable suggestions and guidance throughout this study. This research is partially supported by Education New Zealand under New Zealand International Doctoral Research Scholarship.

Notes

ARL = average run length; MDRL = median of the run length; SDRL = standard deviation of the run length.

ARL = average run length; MDRL = median of the run length; SDRL = standard deviation of the run length.

ARL = average run length; MDRL = median of the run length; SDRL = standard deviation of the run length.

ARL = average run length; MDRL = median of the run length; SDRL = standard deviation of the run length.

ARL = average run length; MDRL = median of the run length; SDRL = standard deviation of the run length.

ARL = average run length; MDRL = median of the run length; SDRL = standard deviation of the run length.

ARL = average run length; MDRL = median of the run length; SDRL = standard deviation of the run length.

ARL = average run length; MDRL = median of the run length; SDRL = standard deviation of the run length.

ARL = average run length; MDRL = median of the run length; SDRL = standard deviation of the run length.