A robust dispersion control chart based on modified trimmed standard deviation

Control Chart is a widely used on-line process control techniques to control
variability. This paper focuses on variability due to dispersion of a quality
characteristic. Classical methods of estimating parameters of the distribu-
tion of quality characteristic may be aected by the presence of outliers. In
order to overcome such situation, robust estimators, which are less aected
by the extreme values or small departures from the model assumptions, are
introduced in industrial application. This article introduced a modication
to trimmed standard deviation to increase its eciency, and is used in con-
trolling process dispersion. Authors constructed a phase-I control chart de-
rived from standard deviation of trimmed mean, which is robust. Simulation
study is conducted to assess its performance at phase-II. This robust control
chart is compared with s-chart in terms of its eciency to detect outliers or
assignable causes of variation as well as its Average Run Length.