This paper proposes an alternative p-value test approach to proceed statistical tests of mean for normal population with fuzzy data. In real-world applications, the approach is often used in tests of average quality characteristics in manufacturing environments. In the proposed approach, the membership function of fuzzy test statistic is first constructed based on the α-cuts of fuzzy numbers. Since the function relationship between fuzzy test statistic and fuzzy data under known population variance is different from that under unknown population variance, we use two methods to construct the membership function under known and unknown population variance environment respectively. An alternative p-value is then calculated according to this membership function. Consequently, a statistical decision can be made by comparing the alternative p-value with the significance level. Two numerical examples are given for illustration.
為了持續優化網站功能與使用者體驗,本網站將Cookies分析技術用於網站營運、分析和個人化服務之目的。
若您繼續瀏覽本網站,即表示您同意本網站使用Cookies。