This is the authors' abstract. We don't add key points for this paper.
The process capability index (PCI) is an important tool for determining the yield of the manufacturing process. Traditional PCIs are frequently used for normally distributed data. However, in practice, the data may follow a non-normal distribution, and traditional PCIs cannot be directly used. In this paper, we propose four PCIs for the Weibull distribution. The proposed PCIs have interpretability for process analysis and improvement and provide an exact functional relationship between the indicator values and non-conforming proportion. Additionally, when the parameters are unknown, we provide two estimation methods for the proposed PCIs: Modified approximate SC estimation (MASC) and bias-corrected MASC estimation (CMASC). These methods directly assess the PCIs without estimating the distribution parameters. The simulation results show that, under different skewness and sample sizes, the CMASC method outperforms the other methods and exhibits greater robustness to distributional skewness. Finally, an example demonstrates the process analysis and improvement strategy proposed in this study.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Bai et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: