Data Sheet 1_Monitoring time-between-events data: evaluating the effectiveness of a new control chart.pdf

The findings indicate that, under the numerical settings investigated, the proposed chart exhibited better performance relative to the benchmark CQC-type charts in lower, upper and both states of the process. This performance may partly reflect the distributional assumption underlying the proposed chart rather than its intrinsic advantages. Therefore, we conducted a robustness study to investigate it.

The results demonstrate that the performance of the proposed chart declines when the underlying distributional assumption is violated. This sensitivity to distributional assumption was observed across all CQC-type charts examined rather than being unique to the proposed chart. Furthermore, we also investigate how Phase I estimation uncertainty propagates into the false-alarm behavior of every chart examined, extending the evaluation from an idealized comparison to one that reflects realistic conditions where the in-control parameters are unknown.

The results of Phase I estimation indicate that increasing the sample size vastly enhances the reliability of the proposed and benchmark control charts. Two case studies further illustrate the practical applicability of the proposed chart.</p>