Many herbivorous, predaceous, and parasitic insects use constitutive and herbivore-induced volatile organic compounds (VOCs) to locate their respective host plant, prey, and hosts. Multivariate statistical tools (e.g., factor analysis) are recognized increasingly as an appropriate approach for analyzing intercorrelated data such as presence/absence or quantities of VOCs. One challenge of implementing factor analysis is determining how many new variables (factors) to retain in the final analysis. I demonstrate a method proposed by Johnson and Wichern to mitigate this problem by using VOC data published in Chen et al. The advantage of using loading (or weight) transformation in interpretation of new variables was also illustrated in the example. Factor analysis found similar nitrogen fertilization effects on VOC production as those in Chen et al. Similarities were 1) nitrogen fertilization interacted with herbivore damage status on VOC production: at low nitrogen (42 ppm) level, beet armyworm, Spodoptera exigua (Hübner) (Lepidoptera: Noctuidae), damage elicited increases in VOC production, whereas at high nitrogen (196 ppm) VOC production was suppressed; 2) nitrogen fertilization did not affect limonene, alpha-pinene, and beta-pinene production. The seven individual VOCs significantly affected by nitrogen fertilization in Chen et al. were (Z)-3-hexenal, (E)-2-hexenal, (E)-beta-farnesene, (E)-4,8-dimethyl-1,3,7-nonatriene (DMNT), alpha-bergamotene, gamma-bisabolene, and bisabolol, of which only three ((E)-beta-farnesene, gamma-bisabolene, and bisabolol) weighed heavily on factor 1 in the current study.