Browsing Research Articles (Statistics and Actuarial Science) by Author "Brand, J."
Now showing 1 - 1 of 1
Results Per Page
- ItemSorting in combination with quality scoring : a tool for industry professionals to identify drivers of wine quality rapidly(South African Society for Enology and Viticulture, 2018) Brand, J.; Kidd, M.; Van Antwerpen, L.; Valentin, D.; Naes, T.; Nieuwoudt, HeleneQuality plays an important role in the criteria directing wine product development. The evaluation of sensory characteristics associated with wine quality, as perceived by industry professionals, is therefore important. We investigated the suitability of the free-sorting sensory evaluation method, in combination with wine quality scoring using a 20-point scoring system, to determine the drivers of quality. Eight commercial South African Sauvignon blanc wines were assessed by a panel of 24 wine industry professionals. Free sorting with a verbalisation step to describe the groups, followed by quality scoring using score sheets routinely used in the wine industry, was performed. A multivariate sensory map was constructed using DISTATIS to explain the similarities and differences amongst the set of wines. Correspondence analysis (CA) was applied to the group descriptors, and CA deviates were calculated. Pearson’s correlation coefficients between CA deviates and the quality scores were calculated to identify the drivers of quality. Significant differences in quality were observed between the wines. The sensory attributes “passion fruit”, “green pepper”, “peas”, “asparagus” and “green” were frequently cited by the panel for the wines that received the highest average quality scores, and these attributes were identified as drivers of quality. In this study, a procedure is presented that combines sorting and quality scoring to investigate the relationship between sensory attributes and quality scores to identify the drivers of wine quality. Industry professionals and research environments can use this procedure to determine drivers of wine quality in a single evaluation session.