Development of the attributes and panel screening for texture descriptive analysis of milk gels aided by multivariate statistical procedures

Abstract

Univariate and multivariate procedures were used in the development of the attributes and the selection of panel members for texture descriptive analysis of gelified dairy desserts containing kappa carrageenan-locust bean gum. The evaluation of the samples was carried out in three stages: compression and cutting with a spoon, and in the mouth. Initially, 16 attributes were defined as a result of the application of Kelly's repertory grid method. Subsequently the panel evaluated the intensity of these attributes in four samples with four replicates. By applying principal component analysis and stepwise discriminant analysis, 11 attributes were selected. For the panel screening, analysis of variance, multivariate analysis of variance, correlation coefficients and cluster analysis were used. As the judges discriminant power and reproducibility were similar, agreement among them was the criterion for selection. In order to evaluate this aspect, cluster analysis was better than correlation coefficients.