Premium
SHORT‐TERM TRAINING AND ASSESSMENT FOR PERFORMANCE OF A SENSORY DESCRIPTIVE PANEL FOR THE OLFACTOMETRIC ANALYSIS OF AROMA EXTRACTS
Author(s) -
BIANCHI GIULIA,
ZERBINI PAOLA ECCHER,
RIZZOLO ANNA
Publication year - 2009
Publication title -
journal of sensory studies
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.61
H-Index - 53
eISSN - 1745-459X
pISSN - 0887-8250
DOI - 10.1111/j.1745-459x.2008.00200.x
Subject(s) - ripeness , reliability (semiconductor) , aroma , statistics , quality (philosophy) , odor , sensory analysis , product (mathematics) , mathematics , variance (accounting) , artificial intelligence , pattern recognition (psychology) , computer science , psychology , food science , chemistry , ripening , power (physics) , physics , philosophy , epistemology , geometry , quantum mechanics , accounting , neuroscience , business
A study on the performances of a panel on the ability to discriminate between odor notes characteristic to nectarine fruits of different maturity (from unripe to fully ripe) was carried out by computing, from analysis of variance results of standardized scores, the agreement coefficient, which describes variation of ratings for a single product, and the reliability coefficient, which reflects discrimination ability of a panel for a group of products. Panelists, after short‐term training for recognition of the main odor attributes of peach aroma, scored the odor profile of both model solutions and Spring Bright nectarine aroma extracts from pool of fruit having well‐defined maturation degrees. Panel performances were good in terms of both agreement and reliability for the descriptors related to unripeness (sharp, herbaceous) while were good for agreement but fair for reliability for descriptors related to ripeness (fruity, ripe).PRACTICAL APPLICATIONS The use of agreement and reliability coefficient, which are based on classical reliability theory and recent advances in psychological measurement, is quite a novel technique in evaluating performances of sensory panels. In this study, these coefficients were applied to monitor and compare performances of both the whole panel in evaluating attributes and the single panelist in evaluating each attribute. The results of the study underlined the usefulness of the graphical presentation of the two indices on the rating quality space in comparing individual panelist performances. In this way the selection of panelists according to their ability in rating a distinctive feature of a product can be easily accomplished.