Factors involved in the consumer’s sensorial evaluation of fruit qualityand the construction of the subjective evaluation models of the ‘Fuji’ apple

Abstract：【Objective】To work out the consumer’s preference on the fruit quality, the relationship between the subjective quality indicators and objective sensorial scores, and to set up an exact predictionmodel of the consumer’s sensorial evaluation of fruit quality scoring based on the quality indicators.【Methods】The‘Fuji’apple fruit harvested from 53 national major producing counties were collected,their subjective fruit quality concerning the appearance, internal flavor quality and integrated sensorialquality were assessed by randomly picking 50 consumers of deferent ages ranging from 20-60 years old.Meanwhile, eight objective quality indicators were measured. Their correlation between the subjective sen⁃sorial integrated evaluation scores and the objective indicators were analyzed by the partial least square re⁃gression method (PLSR). The effect of the fruit objective indicators X1-X8 and their weights on the subjec⁃tive evaluation scores Y1-Y4 were estimated.【Results】To find out the correlation of the consumers’subjective scores and the objective quality indicators, a multiple co-linearity diagnosis between the consumers’subjective scores and the objective quality indicators were carried out. Results showed that multipleco-linearity existed between the eight objective indicators, further multiple co-linearity regression analy⁃sis were necessary to determine their relationship regression analysis. Partial least squares regression, oneof the most newly developed multiple co-linearity regression analysis techniques, showed that single fruitmass (X1), coloration coverage (X3), fruit firmness (X4), SSC: TA ratio (X8) and SSC (X6) contributed to theconsumers’subjective score with positive loading weight and factors, while the fruit index (X2), water con⁃tent (X5) and titratable acids (X7) contributed to the consumers’subjective score with negative loadingweight and factors. Among the eight indicators, the coloration coverage (X3) and SSC (X6) harboring biggerpositive loading weight and factors and titratable acids (X7) poses bigger negative loading weight and fac⁃tors, which implied that consumers preferred fruit with larger coloration coverage and high SSC but dis⁃liked the fruit with high titratable acids content. The relationship between the consumers’subjectivescores and three different categories of consumers’sensorial quality scores including fruit appearancequality, internal quality and flavor quality were analyzed. The results showed that fruit appearance quali⁃ty, internal quality and flavor quality contributed 30.04%, 27.42% and 42.52% for the consumers’subjective integrated sensorial scores, respectively. This implied that consumers’paid more attention on flavorquality than appearance quality and internal quality. This is the first quantitative description of effects ofdifferent quality categories on the consumers’subjective evaluations. With the objective fruit quality indicators as independent variables (X1-X8) and the consumer’s subjective evaluation score as dependent variables (Ý), a universal model for predicting the general consumer’s subjective integrated evaluation scorewas constructed. The equation is as follows: Ý=95.711 2+ 0.0155X1-6.525 4X2+ 0.099 6X3+ 0.428 1X4-0.595 3X5+ 0.987 6X6- 4.992 2X7+ 0.049 6X8 (in which the X1, mean fruit weight; X2, fruit shape index; X3,colored area; X4, fruit firmness; X5, water content; X6, soluble solid content; X7, titratable acid content; X8,SSC: TA ratio; and Ý, predicted consumer’s scores). The correlation analysis of the consumers’scoresand the predicted consumers’scores showed a good linear distribution validating the usability of the predicting equation.【Conclusion】The result presented the consumers sensorial preference of the‘Fuji’apple objective quality indicators and a predicting model for consumers sensorial score based on the qualityindicators were established, which will provide an important understanding and platform for fruit subjective quality comparisons, cultivation method evaluations and even optima cultivation region selections.