Relations between chemical components and sensory properties of vinegars (Part III)

Application of multiple regression method

Abstract

Multiple regression analysis was applied to the correlation matrices of all 28 variables, including 18 chemical analyses and 10 sensory evaluations carried out on 35 vinegars available on the market. The preference, criterion variable, could not be explained by chemical properties adopted as predictor variables, when all vinegars were included. However, for rice and wine vinegars, comparatively high values of determination coefficients were found. Therefore, it will be possible to explain each vinegar if some suitable chemical properties are selected as the predictor variables. When sensory properties were adopted as the predictor variables, the preference was explained with a determination coefficient of 72. 6% for all vinegars and more than 95% for respective vinegars. The sensory properties were explained by chemical properties as the predictor variables with a determination coefficient of more than 80% in sweetness for all vinegars and in many sensory properties for rice and cider vinegars. Evaluation values for sweetness, sourness, saltiness and vmami (taste) were not satisfactorily explained by total acid, sugars, salt and total nitrogen.