2026 Volume 95 Issue 4 Pages 330-339
We investigated the selection of panelists after the test for analysis in the eating quality test of cooked rice by the former Food Agency method. A panel consisting of unselected panelists (“all”) evaluated four test samples per test in six items (“overall evaluation”, “appearance”, “aroma”, “taste”, “stickiness”, and “hardness”) by comparing them with one reference. Eating quality tests were conducted with four replicates for each of the six different samples. The evaluation trends of each panelist were analyzed from two perspectives: agreement (simple correlation coefficient between each panelist’s average score for each sample and the average score for each sample of the panel) and discrimination (F-value of one-way ANOVA with sample as a factor). The panelists whose agreement and discrimination scores of “overall evaluation” were both p<0.05 were classified as “selected” group and scores of “selected” group for each evaluation item were calculated. Additionally, the relationship between the “overall evaluation” and other evaluation items was examined using regression analysis. The results showed that the “selected” group, compared to “all,” had an improved ability to detect differences from the reference in “taste,” which in turn improved their ability to detect differences from the reference in “overall evaluation,” while “stickiness” also influenced the decision on the “overall evaluation.”