Main Article Content

Abstract

The exponentially weighted moving average (EWMA) chart is very popular in statistical process control for detecting the small shifts in process mean and variance. This chart performs well under the assumption of normality but when data violate the assumption of normality, the robust approaches needed. We have developed the EWMA charts under different robust scale estimators available in literature and also compared the performance of these charts by calculating expected out-of-control points and expected widths under non-symmetric distributions (i.e. gamma and exponential). The simulation studies are being carried out for the purpose and results showed that amongst six robust estimators, the chart based on estimator Q_n relatively performed well for non-normal processes in terms of its shorter expected width and more number of expected out-of-control points which shows its sensitivity to detect the out of control signal.

Keywords

exponentially weighted moving average (EWMA) robust estimator expected points out-of-control (EPO) expected width (EW) interval width (I.W).

Article Details

Author Biographies

Nadia Saeed, Punjab University

Assistant Professor, College of Statistical and Actuarial Sciences, Punjab University Lahore

Shahid Kamal, Punjab University

Professor, College of Statistical and Actuarial Sciences, Punjab University Lahore
How to Cite
Saeed, N., & Kamal, S. (2016). The EWMA control chart based on robust scale estimators. Pakistan Journal of Statistics and Operation Research, 12(4), 659-672. https://doi.org/10.18187/pjsor.v12i4.1475