|
Abstract:
|
We conducted a study to reduce sugar content in a beverage while keeping essential sensorial properties similar to commercially available control product. Effect of the three calorie-reducing ingredients was evaluated. A factorial design included 10 blinded variants with different ingredient levels. The bench respondents compared sensorial properties of individual samples to unmasked control under Overall Difference, Sweetness, Mouthfeel Syrupy, and Aftertaste. The gap was recorded as a Degree of Difference on the scale 0-7 with zero for No Difference to the control. Furthermore, masked control was added to the mix, to measure the "noise" in Degree of Difference for each attribute for bias reduction and model accuracy adjusting. Response Optimizer and PCA BiPlot were applied to map variants and attributes. The mapping was also evaluated by ordinal logistic modelling where normality was less concerned. Results suggested the two ingredients were significant on Overall Difference and Sweetness. Specifically, at the optimal points of these two ingredients 2.7 Degree of Difference was achieved for Sweetness, comparing to 1.4 of overall "noise" level.
|
