Abstract
The relationship between sensory data and gas chromatographic (GC) was investigated for 29 samples of sesame seed oil by multivariate analysis. The samples were classified into six and four groups by the couple of principal components analysis and cluster analysis of sensory and GC data, respectively. The clusters were found to correspond well to each other except in two cases. GC data were examined for their relating to the sensory scores on ten flavor attributes (strong, light, aromatic, mild, nutty, bitter, astringent, complicated, degraded, aftertaste) by stepwise multiple regression analysis. All multiple regression models had statistical significance (P<0.01).
