Álgebra intervalar na análise de correspondência: um estudo de caso em testes de aceitação sensorial com erros de medida

Abstract

Applicability of affectionate tests on the sensory analysis is given from specific evaluations to the consumers’ acceptability and preference related to one or more products. So that, different scales which relate sensory perceptions to hedonic experiences are used. However, the contextual interference, through which tasters are involved are not despised. Two divergent groups of researchers fit into this context: the experimental psychologists and sensory researchers, who share the use of hedonic scale including “likes” and “dislikes” degree evaluation. In order to improve the resulting analysis, inserting this scale variability, we aim to propose building percentage maps using the correspondence analysis technique together with interval algebra. Through a case study on Soya genotypes acceptability [Glycine Max (L.) Merril] named Black (MGBR07-7141), Brown (BRSMG-800-A) and Yellow (BRSMG-790-A), sensory features like appearance, global impression and texture were considered. To summarize, information related to the accuracy of tasters’ response were provided while considering the measurement error inherent to the taster’s mark for the expected response to the use of interval algebra in the correspondence analysis.