Análise sensorial: um estudo sobre procedimentos estatísticos e número mínimo de julgadores
Data
2009-08-28
Autores
Guaglianoni, Dalton Geraldo [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
A partir do início dos anos 70, ficou estabelecido que a qualidade sensorial é o resultado da interação entre o alimento e o homem, portanto varia de forma subjetiva. Diferentes escalas foram então associadas às análises sensoriais, surgindo como conseqüência procedimentos estatísticos específicos nesta área. Os experimentos envolvendo análise sensorial iniciaram assim a discussão quanto ao número necessário de julgadores, ao seu treinamento e aos procedimentos apropriados à análise dos dados extraídos dessas escalas ordenadas. Este trabalho foi baseado em um teste sensorial de aceitação do atributo “Impressão Global” de quatro amostras de cachaças avaliadas por 120 julgadores. O objetivo fundamental foi realizar comparações entre os procedimentos estatísticos, paramétricos ou não-paramétricos, com vistas a compará-los e estabelecer o número mínimo de julgadores. Com base nos dados dos testes sensoriais foram criadas 5400 amostras sistemáticas com números diferentes de julgadores, sobre as quais foram aplicados os procedimentos estatísticos estabelecidos. Os resultados assim obtidos mostraram ser possível a utilização de um número menor de julgadores. Revelaram ainda que, na maioria dos casos, o índice de concordância é o fator mais importante na determinação do número mínimo de julgadores, sendo que, para valores de concordância de Kendall maior ou igual a 0,18, o número de 60 julgadores mostrou-se suficiente.
Since begin of 1970s, it was well established that the sensory quality of a food was a result of the interaction between the food and humans, thus varying subjectively. This led to the creation of scales associated to sensory analyses, end as a consequence, specific statistical procedures in this area. The experiments related to sensory analysis started the discussion about the number of judges required, their training, and the appropriate procedures for the analysis of data extracted from these ordered scales. The present work was based on a sensory acceptance test of the attribute “Global Impression” cachaça samples evaluated by 120 judges. The basic objective was to compare the standard statistical procedures, parametric and nonparametric, in order to determine the minimum number of judges necessary. From the initial data, 5400 systematic samples with different numbers of judges were created, and the standard statistical procedures were applied. Bored on the obtained data it was evaluated if the base experiment could be reproduced with fewer judges. The results indicated that, in most cases, the index of agreement is the most important factor to determine the necessary number of judges. In this case, if the Kendall concordance value among judges is ≥ 0,18, the total number of judges could be only 60.
Since begin of 1970s, it was well established that the sensory quality of a food was a result of the interaction between the food and humans, thus varying subjectively. This led to the creation of scales associated to sensory analyses, end as a consequence, specific statistical procedures in this area. The experiments related to sensory analysis started the discussion about the number of judges required, their training, and the appropriate procedures for the analysis of data extracted from these ordered scales. The present work was based on a sensory acceptance test of the attribute “Global Impression” cachaça samples evaluated by 120 judges. The basic objective was to compare the standard statistical procedures, parametric and nonparametric, in order to determine the minimum number of judges necessary. From the initial data, 5400 systematic samples with different numbers of judges were created, and the standard statistical procedures were applied. Bored on the obtained data it was evaluated if the base experiment could be reproduced with fewer judges. The results indicated that, in most cases, the index of agreement is the most important factor to determine the necessary number of judges. In this case, if the Kendall concordance value among judges is ≥ 0,18, the total number of judges could be only 60.
Descrição
Palavras-chave
Avaliação sensorial, Estatísitica - Análise, Cachaça, Sensory analysis, Number of judges, Statistics in sensory analysis
Como citar
GUAGLIANONI, Dalton Geraldo. Análise sensorial: um estudo sobre procedimentos estatísticos e número mínimo de julgadores. 2009. 125 f. Tese (doutorado) - Universidade Estadual Paulista, Faculdade de Ciências Farmacêuticas, 2009.