Title Page
Abstract
Contents
Ⅰ. INTRODUCTION 13
Ⅱ. LITERATURE REVIEW 16
1. Sensory evaluation 16
1.1. Descriptive analysis (DA) 17
2. Consumer-based sensory profiling 18
2.1. Check-all-that-apply (CATA) 19
2.2. Rate-all-that-apply (RATA) 20
2.3. Intensity scales 21
3. Preference mapping (PM) 22
3.1. Internal preference mapping 23
3.2. External preference mapping 24
4. Product optimization 25
4.1. Euclidian Ideal Point Mapping (EDIPM) 26
4.2. Response Surface Ideal Point Mapping (RSIPM) 27
Ⅲ. Materials and methods 28
1. Samples 28
2. Sensory profiling methods 30
2.1. Descriptive analysis (DA) 30
2.2. Consumer-based sensory profiling 32
3. Product optimization 36
4. Validation test 37
5. Statistical analysis 38
Ⅳ. RESULTS AND DISCUSSION 39
1. Descriptive analysis (DA) 39
2. Consumer-based sensory profiling 44
2.1. Consumers' overall liking (OL) 44
2.2. Rate-all-that-apply (RATA) 47
2.3. Intensity scales 51
3. Product optimization 54
3.1. Euclidian Ideal Point Mapping (EDIPM) 54
3.2. Response Surface Ideal Point Mapping (RSIPM) usng DA data 57
3.3. RSIPM using RATA data 59
3.4. RSIPM using intensity scales data 61
4. Comparison of methodologies 63
5. Validation test 68
Ⅴ. SUMMARY AND CONCLUSION 72
REFERENCES 74
APPENDICES 85
〈Appendix 1〉 Descriptive analysis test 85
〈Appendix 2〉 Rate-all-that-apply (RATA) test 87
〈Appendix 3〉 Intensity scales test 89
〈Appendix 4〉 Validation test 93
국문초록 95
Table 1. Product description of seven apple juice samples. 29
Table 2. Sensory attributes, definition, reference, and its intensity of developed lexicon for the apple juice samples. 31
Table 3. Demographic information of preliminary online survey for consumer-based evaluation. 33
Table 4. Means for the sensory attributes obtained from descriptive analysis (DA) for seven apple juice samples. 42
Table 5. Overall liking scores of seven apple juice samples for the overall and clustered groups by K-means cluster analysis. 45
Table 6. Means for the sensory attributes obtained from RATA for seven apple juice samples. 49
Table 7. Means for the sensory attributes obtained from intensity scales for seven apple juice samples. 52
Table 8. Sensory profiles of the ideal products derived by EDIPM. 55
Table 9. Sensory profiles of the ideal products derived by RSIPM. 66
Table 10. Rv coefficient of EDIPM from descriptive analysis (DA) and consumer-based methods. 67
Table 11. Rv coefficient of RSIPM from descriptive analysis (DA) and consumer-based methods. 67
Table 12. Overall liking (OL) scores for three ideal products based on RSIPM and the product that scored highest in OL in the initial study for each cluster. 70
Table 13. Results of Friedman analysis for the ranking method in the validation test. 71
Figure 1. Principal component analysis (PCA) plot displaying the first two principal components using the sensory attributes obtained from descriptive... 43
Figure 2. Preference patterns of two consumer clusters classified by K-means. 46
Figure 3. Principal component analysis (PCA) plot displaying the first two principal components using the sensory attributes obtained from RATA. 50
Figure 4. Principal component analysis (PCA) plot displaying the first two principal components using the sensory attributes obtained from intensity scales. 53
Figure 5. Configuration of the ideal product using the preference data of cluster 1 based on EDIPM (n=83). 56
Figure 6. Configuration of the ideal product using the preference data of cluster 2 based on EDIPM (n=77). 56
Figure 7. Configuration of the ideal product using the sensory data of descriptive analysis (DA) based on RSIPM. 58
Figure 8. Configuration of the ideal product using the sensory data of RATA based on RSIPM. 60
Figure 9. Configuration of the ideal product using the sensory data of intensity scales based on RSIPM. 62
Figure 10. Correlation analysis plot for the ideal sensory profiles of DA and two consumer-based methods: (a) DA-RATA; (b) DA-Intensity scales. 66