A Comparison of EWMA Charts for Monitoring the Process Mean Using Interval‐Valued Data

ABSTRACT In this paper, we compare the performance of EWMA charts based on the interval sampling, midpoint, and conditional expected value methods for monitoring the process mean using interval‐valued data with fixed and variable widths, assuming a normally distributed process. The steady‐state average run‐length profiles of the EWMA charts are estimated using Monte Carlo simulation. The three EWMA charts exhibit comparable performance, with the midpoint‐based EWMA chart being the easiest to implement. An application illustrates the implementation of the EWMA charts. In addition, the neutrosophic EWMA chart is critically reviewed, and its limitations in handling interval‐valued data are highlighted, leading us to advise against its use for process monitoring.