**2.4 Quantitative descriptive sensory analysis**

The approach of quantitative descriptive analysis (QDA) is based on the concept of the ability of panelists to verbalize impressions of the commodity in a reliable manner. The process embodies a systematic screening and training technique, the creation and use of a sensory language, and the labeling of items in a repeat test in order to achieve a complete, quantitative definition. We followed the method of stone and side, a panel of 10 assessors (5 males and 5 females, age range 20–50 years). Initially, 90 descriptive attributes to coffee beverage were Aroma (attributes, and to test beverage, get the six attributes after getting consensus among 10 assessors on the definition of terms as shown in **Figure 1**). The rating scale method shown below was evaluated by using a two-way analysis of variance (ANOVA).

### **2.5 Sensory test with a rating scale method**

With a line scale, the members of the panel rated the strength of the coffee aroma by labeling the horizontal line. The marked point corresponded to the influence of the stimulus perceived. The length of the line was 14 cm with anchors on both ends. The left end of the scale corresponded to "none," while the right end corresponded to such descriptive terms. The panel consisted of 5 males and 5 females. Their ages ranged from 20 to 50 years with a mean of 24 years.

#### **2.6 Statistical analysis**

Analysis of variance (ANOVA) was applied for the determination of the main effects of the investigated independent factors (flavor and percentage) on the sensory attributes as they were rated during the QDA procedure. Duncan's multiple range test was used to separated means of sensory attributes when significant differences (p < 0.05) were observed.

A Shimadzu GC-MS solution of 2.5 SU1 and an Agilent ChemStation D.02.00.275 were used to collect data. To visualize sample groups and compare information provided by each sampling, Principal component analysis (PCA) was used. In normalized ISTD data, PCA based on Pearson correlation coefficient was

performed. XLSTAT (version 2015.5.01.23164) copyright Addinsoft 1995–2015 was used for statistical analysis, including one-way ANOVA and PCA. Non-polarity.
