An Improved R (Range) Control Chart for Monitoring the Process Variance

摘要

An R chart is often used to monitor shifts in the process variability. However, the range, R_i, i = 1, 2, ..., statistics from a sampling distribution are highly skewed. Hence, the classical R chart based on the ± 3σ control limits will not give an in-control average run length of approximately 370, or equivalently a type I error, α = 0.0027. In this paper, an approach is shown to obtain the control limits of an improved R chart based on a desired type I error from the density function of the R_i statistics.