<scp>CATA</scp> vs. <scp>FCP</scp> for a rapid descriptive analysis in sensory characterization of fish

Abstract

Abstract Storing non‐marketable fish aboard fishing vessels takes up valuable storage space and, consequently, it is usually discarded. Therefore, the strategy to minimize the amount of discarded fish and boost the fishing economy should entail measures to turn unexploited species into marketable fish. The use of sensory descriptive analysis is usually the first step in the characterization of foods and assess their baseline appeal to the final consumer. In this study, the sensory characterization of five unexploited fish species was conducted, using a semi‐trained panel, with the objective of establishing the appeal and marketability of these species. Check‐all‐that‐apply (CATA) and Free Choice Profiling (FCP) were used to obtain descriptive attributes. Correspondence Analysis (CA) was used to examine CATA data and to compare both sensory methods, while Principal Component Analysis (PCA) was used for FCP data analysis. The results obtained with CATA method allowed to differentiate species by appearance, odor and flavor, while appearance and odor were statistically significant descriptors in FCP. CATA method was found to be more suitable to discriminate fish species, presenting a higher discriminatory power than FCP. Practical Applications For CATA and FCP methods, using a panel of tasters with short training may increase the quality of a descriptive profile by avoiding the use of less accurate descriptors and ensure that the most relevant descriptors are included, especially in the case of CATA. In this study, both methods were suitable for fish profiling, although they provide different sensory characterization for the different samples. The statistical methods applied allowed to statistically identify significant attributes for each fish species and to differentiate from each other. When comparing both tested sensory methods, CATA reveals a greater discriminatory character compared to FCP. This can be explained by the absence of a support list of descriptors in the FCP that limits the choice of attributes by the panelists.