Abstract:

Repeatability, reproducibility and consistency are considered as the indexes for the ranking capability of panels
and panelists for sensory evaluation. The type, difference and number of samples are very important in ranking performance
measurement. Rank is identified as the basic data style to evaluate the ranking capability. If the order of samples is unknown,
the optimal evaluation order should be firstly confirmed by the methods combining Friedman test and least significant
difference (LSD) after unsuitable panelists who have less repeatability and reproducibility are rejected. This paper was focused
on the application approach of the two-variable Spearman rank correlation and Kendall coefficient of concordance for multiple
variables in evaluating the repeatability, reproducibility, consistency of panels and panelists. Meanwhile, a good manufacturing
practice to evaluate the ranking capability of panels and panelists has been established, which will provide a theoretical
foundation for proficiency testing of sensory ranking panels from different laboratories. It is not only helpful for monitoring the
performance of panels and panelists, but also for improving management capacity of sensory evaluation laboratories.

Key words: sensory analysis, ranking, panels and panelists, performance evaluation