Evaluation of the consistency of wine quality assessments from expert wine tasters

Publisher Website
Google Scholar
Abstract
The within-assessor consistency of 571 experienced wine tasters in assigning quality ratings to red and white table wines was determined. Wine quality scores were collected over a 15 year period from tasters undertaking an advanced training course in wine quality assessment. Consistency was measured by correlating the scores given to duplicate presentations of wines,c alculatingt hep ooled variation inr epeats cores, andb y assessingt het asters’a bilityt o allocate duplicate presentations of the same wine to the same quality category. The distribution of individual assessor correlation coefficients for both red and white wines was left skewed with a median of 0.48 and 0.37 respectively. The ability of the tasters to consistently allocate red wines to the same or similar quality categories was particularly good. Consistency was improved by combining the independently assigned scores of three assessors as is done in the Australian wine show system. Assessors generally showed greater reproducibility in scoring red wines compared with whites, and in general, the ability of a tastert o consistentlys corer edw ines was ap oorp redictoro ft heira bility to consistentlys core white wines, and vice versa. Lastly, while the majority of wine tasters showed statistically significant scoring consistency, there was considerable variation between individuals in their ability to do so. Keywords: wine quality, wine scoring, wine assessment, taster consistency