Consistency between sorting task and sensory descriptie analysis: application to cheese products

Sensory descriptive analysis (DA) is the incontrovertible method to obtain a complete description of a product. In such methodology, trained panellists evaluate a series of attributes that significantly discriminates products in a set. Sensory sorting task (ST) is often used with consumers in order to obtain a sensory map representing the global differences between products. The two methods are based on very different methodology, but each attends to show the main differences and similarity amongst products. Thus, one can wonder whether these two methods show the same differences, or more precisely, do the results obtained by DA resemble the ones obtained with ST? In the present study, we examined the consistency between these two methods in terms of global similarity between different cheese products. Seven commercial semi-hard cheeses were chosen for the study. A ST with 60 untrained judges was carried out to compare the cheeses according to their similarities by orthonasal and retronasal stimulation. A DA was performed by a panel of 13 trained judges evaluating 13 odour attributes by orthonasal stimulation and 12 attributes by retronasal stimulation (aroma, taste, texture). The results obtained with the ST method were analysed by multidimensional scaling whereas principal component analysis was applied on the DA results. Hence, each analysis led to construction of multidimensional sensory maps whose proximity was estimated by the normalised RV coefficient (NRV). The orthonasal profile was very similar to the orthonasal sorting results and the NRV reached 4.01 indicating very close relationship between the two datasets. The retronasal sorting results were also very close to the retronasal profile of the cheeses (NRV = 2.06). In conclusion, our results indicate a very good consistency between and ST, showing that the DA can highlight the global differences between products perceived by consumers