1Department of Business Science, Area: Technology and Resource Valorisation, University of Bologna, Piazza Scaravilli 2, 40126, Bologna, Italy
2Department of Statistics, University of Bologna, Piazza Scaravilli 2, 40126, Bologna, Italy
In this study, we tested application of a statistical method to classify foods in clearly defined homogeneous categories, useful when dealing with products whose merchandising-quality characteristics are not easily identifiable.
The method based on Principal Components Analysis (PCA), Cluster Analysis and Discriminant Analysis was applied to various types of dry, savoury and sweet bakery products, a particularly dynamic sector of the market with the introduction of new products with quality and merchandising characteristics (crostini, filled biscuits, snacks) differing from those of traditional products (breadsticks, crackers, French toast, plain biscuits).
Using their composition in nutrient factors together with the new differentiating factor of apparent density, it was possible to establish correlations able to place the various types of product in different merchandising areas typical of their overall quality.
Statistical method, merchandising categories, identification, overall quality, foods
