感官分析评价小组及成员表现评估技术动态分析

Overview of Assessment Techniques for Sensory Panel and Panelist Performance

作为感官评价仪器的"评价员/评价小组"是获得可靠感官分析数据的关键,其性能表现评估技术是有效管理该仪器的重要手段。本文将信度与效度作为该"仪器"的评估指标,发现21世纪以后该类研究进入高峰期,其超过85%的研究成果发表于Food Quality and Preference与Journal of Sensory Studies这两个感官研究类权威杂志中,其中欧美国家在此技术中占主导地位。透过技术内容发现,基于定量描述能力的评价小组及成员性能表现评估技术研究频繁并趋于成熟,技术手段重点采用单参数或多参数方差分析、多元统计方法(主成分分析、广义普罗克分析等),涌现了以PanelCheck、Compusense Five等为代表的评估软件,同时国际标准化组织和美国材料与试验协会机构也进入相关标准研制阶段。而有关差别与排序的评估技术研究相对缺乏。本文提出性能评估所用样品性质及数量、性能评估所用数据表现形式、性能评估感官实验设计要求、性能评估期望值及置信区间、性能评估所用感官分析方法选择、性能评估的数学统计方法选择等关键要素将成为该领域未来努力的方向,并有助于形成统一的、系统的感官分析评价小组及成员表现评估技术规范。

In sensory evaluation,panels and panelists are used as instruments that are the key tool to obtain reliable sensory data.Assessment techniques for their performance can provide an important approach for managing them effectively.In this paper,the validity and reliability are used as the main indexes for performance assessment,which have reached a plateau of research since the beginning of the 21st century.More than 85％ of the research papers have been reported in the top journals in sensory science,such as Food Quality and Preference and Journal of Sensory Studies.Especially,the majority of the research reports come from the USA and Europe and have tended to be frequent and mature in assessing the performance of descriptive sensory panels-panelists.The various methods include univariate or multivariate variance analysis,and multivariate statistical methods such as principal component analysis （PCA）,generalized procrustes analysis （GPA）,etc.At the same time,softwares such as PanelCheck,Compusense Five,and so on have been developed,while the International Organization for Standardization and the American Society for Testing and Materials are also formulating the relevant standards on quantitative descriptive analysis.By contrast,relatively insufficient research has been done on performance assessment in ranking and discrimination sensory panels-panelists.Six key research contents are proposed for future study.They include the number and property of sample,the form of data presentation,the design of sensory experiment,expected value and confidence interval,and the selection of sensory analysis methods and statistical methods.This paper will be helpful to establish the unified and systemic framework for assessing performance in panels and panelists.