Multivariate analysis of chemical components of tobacco leaves

This study examined the Multivariate analysis of Chemical Components of Tobacco Leaves using Canonical Correlation Analysis which seeks to identify and quantify the association between two sets of variables. The paper focused on using Canonical Correlation Analysis to analyze data on chemical components of 25 tobacco leaf samples. The multivariate data satisfied the normality assumption. The data for this study, which contains three criterion measures and six predictor variables, were analyzed using the “SAS” statistical software package. Based on the results obtained, and the hypotheses carried out, it was revealed that out of the three sample canonical correlations, the first two (  are significant, while the third one ( = 0.373) is insignificant. The analysis also revealed that the first sample canonical variate,  , of the criterion measures is a “better” representative of its set than the first sample canonical variate,  , of the predictor variables of its set.