Premium
Case studies of industrial applications of the “M+N” classification method with larger M and N for visual and manual inspections
Author(s) -
Kuesten Carla,
Hou Amy,
Wu Jenny,
Bi Jian
Publication year - 2017
Publication title -
journal of sensory studies
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.61
H-Index - 53
eISSN - 1745-459X
pISSN - 0887-8250
DOI - 10.1111/joss.12245
Subject(s) - sorting , cronbach's alpha , quality (philosophy) , product (mathematics) , product testing , sensory system , two alternative forced choice , reliability (semiconductor) , statistics , sensory analysis , perception , computer science , artificial intelligence , mathematics , pattern recognition (psychology) , psychology , cognitive psychology , engineering , psychometrics , operations management , algorithm , philosophy , geometry , power (physics) , physics , epistemology , quantum mechanics , neuroscience
In recent sensory literature, Bi and Kuesten (2015a. Food Quality and Preference , 43, 47–52; 2015b. Journal of Sensory Studies , 30, 461–471) have proposed the “M + N” classification method with larger M and N (>3) for discrimination testing based on Fisher's exact test. Industry studies using the method are shared in this article to illustrate application of the method. The tests and measures are taken for non‐fatiguing visual and manual inspections conducted across a variety of product categories. Further, within the examples provided, a validation between labs is documented and results of the method are compared to results from other methods (2‐Alternative Choice and 2‐Alternative Forced Choice) based on common conventional sensory measurement Thurstonian d′ . Estimations are given of Intraclass Correlation Coefficient and Cronbach's coefficient alpha for evaluation of panel performance in the test using the “M + N” classification method with larger M and N (only for M = N > 3). Practical applications Sorting methods are popular and often used for rapid investigation of product quality, product differences, or consumer perception and are used to help guide product maintenance and/or development efforts. The “M + N” classification method with larger M and N, a type of sorting technique, can be applied for sensory discrimination and perceptual tests with generation of various indices, including Thurstonian discriminal distance delta (or d′ ) thereby offering the ability to compare results with other sensory discrimination methods.