Abstract
For sensory quality characteristics, such as appearance, softness and color, control charts by attributes are used reluctantly. On the contrary, Wang & Raz tried to resolve this problem by choosing quality-related characteristics from various linguistic variables and presented control charts constructed by using the member-ship functions associating with the linguistic values. This paper presents the new concept of control charts for linguistic data by means of expanding the Wang & Raz's chart, and a method for constructing the charts based on theories of fuzzy sets and mathematical statistics. Our control charts are aimed at directly controlling the underlying probability distributions of the linguistic data in order not only to control the process average but also to control the process variability.
