Use of attribute agreement analysis (AAA) in the validation of sensory evaluation methods: Case study for the visual determination of parasites in fish

Zoran Petrović; Jelena Ćirić; Jelena Babić-Milijašević; Milan Milijašević; Mirjana Lukić; Jelena Jovanović; Aleksandra Nikolić

doi:10.18485/meattech.2024.65.1.7

Zoran Petrović Institute of Meat Hygiene and Technology, Kaćanskog 13, Belgrade, Serbia https://orcid.org/0000-0003-2016-5681
Jelena Ćirić Institute of Meat Hygiene and Technology, Kaćanskog 13, 11 000 Belgrade https://orcid.org/0000-0002-8118-7676
Jelena Babić-Milijašević Institute of Meat Hygiene and Technology, Kaćanskog 13, Belgrade, Serbia https://orcid.org/0009-0003-8923-7046
Milan Milijašević Institute of Meat Hygiene and Technology, Kaćanskog 13, 11000 Belgrade, Serbia https://orcid.org/0000-0003-4269-236X
Mirjana Lukić Institute of Meat Hygiene and Technology, Kaćanskog 13, 11000 Belgrade, Serbia https://orcid.org/0009-0001-7749-864X
Jelena Jovanović Institute of Meat Hygiene and Technology, Kaćanskog 13, Belgrade, Serbia https://orcid.org/0000-0003-0301-729X
Aleksandra Nikolić Institute of Meat Hygiene and Technology, Kaćanskog 13, 11 000 Belgrade, Serbia https://orcid.org/0009-0009-0012-3418

DOI: https://doi.org/10.18485/meattech.2024.65.1.7

Keywords: Attribute agreement analysis (AAA), Sensory methods, Validation, Fish, Parasites

Abstract

Validation of sensory evaluation employs a process similar to any other method vali‑ dation procedure in analytical chemistry. However, the parameters often measured in sensory testing are different, because the human sensory apparatus and brain are the instruments being calibrated. A total of 10 fish samples were examined by an expert 5‑member evaluation panel for the visual presence of parasites in frozen fish. From each frozen hake sample, a group of six slices of fish muscle was formed by longitu‑ dinal sectioning. The total of 60 sliced samples were divided into 10 plates with six samples each. The first plate contained six slices, each of which contained parasites. The attribute agreement analysis showed strong agreement in the overall ratings. It was found that at least 75% of all tests achieved the highest level of agreement. The results of the tests were presented in the form of tables and graphs summarizing the subjective test results using Fleiss’ Kappa and Cohen’s Kappa statistics.