葡萄酒感官评价结果的统计分析方法研究

Studies on the Statistical Analyses Methods for Sensory Evaluation Results of Wine

在葡萄酒的感官评价中,由于品酒员间存在评价尺度、评价位置和评价方向等方面的差异,导致不同品酒员对同一酒样的评价差异很大,从而不能真实地反映不同酒样间的差异。因此,在对感官评价结果进行统计分析时,必须对品酒员的原始数据进行相应的处理,以真实反映样品间的差异。对有关数据处理方法的比较分析结果表明:标准化法不仅没有消除品酒员间的异质性,反而加大了品酒员间的差异;而置信区间法对原始数据进行调整,能有效地降低品酒员间的差异,真实地反映酒样间的客观差异。

In sensory evaluation of wine, the values given by different tasters for the same sample are very variable, because the tasters have different scales, positions and directions of evaluation, Therefore the difference existing among samples of wine are often concealed. So the primitive data must be pre-treated for statistical analyses, to truly reflect the difference among samples. The results of comparative analysis among different methods for the treatment of primitive data showed that, the method of stadardization does not eliminate variance of tasters, but enlarge difference among tasters, however, data transformed by confidence interval method, tasters＇ variance can be reduced and truly reflect the difference existing among samples evaluated.