Description

Abstract: The effect of a warm-up sample on reliability of attribute intensity ratings from descriptive panel evaluations was studied by: (1) using a warm-up sample prior to test samples (WU), (2) using a warm-up sample prior to test samples and providing panel consensus attribute ratings for the warm-up sample to panelists (WUC), and (3) providing no warm-up sample (NWU). The standard deviation of ratings from 7 and 4 of 11 attributes for all samples was smallest in WUC and WU, respectively. In paired t-tests, significant differences were found between duplications in 0, 1 and 4 of 11 attributes in WUC, WU and NWU, respectively. Regression analysis of percent fat vs attribute intensity ratings showed the largest R2 from 6, 2, and 2 of 10 significant regression models in WUC, WU, and NWU, respectively. To maximize reliability of descriptive analysis ratings, WUC should be used.