The PO bootstrap approach for comparing process incapability applied to non-normal process selection. Issue 2 (4th March 2022)

• Record Type: Journal Article
• Authors: Leony, Florence; Lin, Chen-ju
• Is Part Of: Quality technology & quantitative management. Volume 19:Issue 2(2022)
• Page Start: 215
• Page End: 233
• Publication Date: 2022-03-04
• Subjects: Production loss -- inaccuracy -- imprecision -- resampling -- Bonferroni adjustment; Quality control -- Periodicals; Quality control -- Statistical methods -- Periodicals; Industrial management -- Periodicals; Industrial management; Management -- Research -- Methodology -- Periodicals; Qualitative research -- Periodicals; Management; Quality control; Quality control -- Statistical methods; Periodicals; 658.00721
• DOI: 10.1080/16843703.2021.2015827
• Languages: English
• ISSNs: 1684-3703

Abstract

ABSTRACT: Process selection has been a focal task in operation management. This research focuses on finding alternatives to the current process that have to be at least as capable as the current process. Having multiple alternative processes available enables the manufacturers to have better resource utilization and scheduling flexibility. However, selecting the right process under non-normal data remains a challenge. Quality loss is a popular criterion because of its direct relationship with cost objectives. In this research, we propose the Cpp -based PO bootstrap approach to evaluate candidate processes based on quality loss by utilizing the incapability index. The Cpp index represents Taguchi's Loss function k ( x – T ) 2, which is suitable for the nominal-the-best type of quality characteristic. It measures production loss caused by process inaccuracy and imprecision. The experiments show that the proposed method can loosen up the reliance on normal assumption by controlling type I error and providing higher power compared to the extended method from the literature. The application to amplifier circuits manufacturing showed that the proposed method is effective to identify the inferior processes despite the severe departure of data from normal, while the opposed method built under normality assumption fails to do so.