The problem of estimating σ from individuals’ data under out-of-control and nonnormal conditions is investigated. The mean squared errors for 10 estimators proposed in the past 20 years are computed by simulation, both before and after screening for outliers with median-absolute-deviation-based control limits. Also considered are the effects of contaminated data on the performance of exponentially weighted moving average control charts (EWMA). It is shown by simulation that Boyles’ dynamic linear model estimator is frequently the best way of estimating σ. An approach that involves robustly testing for shifts prior to final estimation of σ can be approximated by an automatic method based on smoothing the data first and estimating σ from the residuals. In some situations where noisy data are not amenable to accurate estimates of σ, engineering considerations and careful data analysis are required to obtain accurate estimates.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados