A WARM‐UP SAMPLE IMPROVES RELIABILITY OF RESPONSES IN DESCRIPTIVE ANALYSIS

Authors:

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.

This publication has 14 references indexed in Scilit: