| Sensory evaluation is more and more frequently used in the field of new productdevelopment, quality control and marketing to support product research and to meet the needof consumers. At present, research on descriptive sensory analysis, consumer acceptabilityand their relationship is relatively less. This research compared Spectrum and QDA, two keydescriptive sensory methods, by the descriptive sensory analysis of dried bean curds.Simultaneously, hedonic scale and JAR scale were selected to test consumer acceptability ofdried bean curds. In order to provide solutions to improve sensory quality of products, dataanalysis methods were investigated to build the relationship between descriptive sensoryanalysis and consumer acceptability.Descriptors generated by Spectrum and QDA methods were almost the same. Thequantity of descriptors of QDA method was less than that of Spectrum, and the loss of QDAdescriptors might be due to the disagreement during group discussion. The main qualitydiffenence of descriptors of the two methods was reflected in the attributes naming. Thismight because QDA panelists hadn’t received standardized sensory training. Therefore,QDA’s descriptors might link to consumer language.The reliability of the sensory data from Spectrum and QDA methods was compared byPanel Check software. Results showed that all panelists performed well on discrimination andrepeatability, and good consistency of each panel was detected. It must be mentioned thatpanelists of Spectrum had better sensory evaluation ability than panelists of QDA. Principlecomponent analysis was used to derive a multidimensional space of samples and attributesbased on Spectrum and QDA data separately. The first three dimensions of QDA dataexplained90.26%of the total variance, which was in close proximity to that of Spectrum data,94.30%. In addition, distribution and loading of attributes, especially those flavor attributes,on PCA map were similar between the two methods. Based on these results, QDA’s sensoryevaluation result was acceptable in spite of short-time training.In consumer research, hedonic scale and JAR scale were selected to test the consumeracceptability of dried bean curds. ANOVA and internal preference mapping allowed us toobtain the significant difference of the degree of liking among samples and differentconsumer groups based on variety of liking. Serpentine figure and ideal point method wereused to indicate the intensity distance from each attribute to the “just about right” point, thediscrepancy and direction between each real product and ideal product. In brief, independentdata analysis of hedonic data or JAR data could achieve effective information helping toimprove sensory quality of products.In this paper, relationship between consumer acceptability and descriptive analysis wasdiscussed. External preference mapping was used in order to link hedonic acceptability todescriptive analysis. This mapping gained insights into consumer perception pointing out thepreferred products’ attribute characteristics, and predicted the marketing acceptance of theproducts approximately. Results showed that the consumer perceptual dimensions ofSpectrum and QDA methods were similar and QDA’s prediction was a little more conservative than Spectrum’s. Relationship between hedonic data and descriptive data wasalso established by matching the preferred samples to the descriptive attribute intensity ranges.This matching method was also used to link “just about right” samples to descriptive attributeintensity ranges. Therefore, the preferred and “JAR” attribute intensity ranges were obtainedquantitatively. Besides, penalty analysis connected hedonic data with JAR data obtainingproblematic attributes of each product, direction and priority of improvement. Availableimprovement design of products could then be determined according to the penalty analysisresults and quantitative attribute intensity ranges. |
PDF Full Text Request