The loading values represent the importance of each variable to form a PC. For example, aij indicates the degree of importance of
the jth variable in the ith PC. This idea can be extended to k PCs of interest. Determination of the number of PCs (i.e., k) to retain is subjective. Typically, the scree plot that effectively visualizes the variability of each PC is used (Johnson & Wichern, 2002). A PCA loading value for the jth original variable can be computed from the first k PCs. Thus, an overall PCA weight for jth variable can be represented as follows: