Statistical Analysis between Volatile Components and Sensory Data of Brewed Coffee

Abstract

The relationship between gas chromatographic (GC) data and subjective sensory evaluation was analyzed in 31 arabica coffee by multivariate analysis. Using Quantification Theory 3 for sensory data, the samples were classified into six groups. The relationships between GC data and scores on axis 1 and 2 obtained by Quantification Theory 3 were analyzed by multiple regression analysis. The multiple correlation coefficient (R) was O.85 at step 7 and 16. On the basis of the R and F-value, the multiple regression models were determined for estimating the scores on axis 1 and 2. The result suggested that the aroma profiles of the coffee samples were closed to the sensory evaluation. © 1989, Japan Society for Bioscience, Biotechnology, and Agrochemistry. All rights reserved.