Principal Component Analysis in Horticultural Research1

Abstract

Principal component analysis, a technique which reduces the dimensionality of multivariate data by removing intercorrelations among variables, has a number of potentially useful applications in horticultural research. It can be used to order multivariate commodity quality data in 1 or 2 orthogonal dimensions called principal components, which express most of the variance of the original data. Scores on these principal components can be used as an index of commodity quality to replace subjective visual quality ratings in conventional statistical analyses. Interpretation of the pattern of variable loadings on these principal components may aid in the elucidation of interactions among variables in the data. Plotting of multivariate data in 2 or 3 dimensional principal component space can be useful for displaying relationships among cultivars or species in taxonomic studies.