Proving discriminant loading as a covariance between the discriminant score and observed variable, establishing orthogonal rotation in discriminant axes and analysing perceptual map

This study aims to introduce an approach that extends the application of a widely recognised algorithm for factor analysis (varimax rotation) to discriminant axes, thereby offering a notable contribution to the field of classification. This study proved that discriminant loading represents the covariance between the discriminant score and the observed variable. The data pertain to head measurements for designing helmets using six essential variables. This study also explored the application of a perceptual map to visualise the relationships between variables and to uncover latent structures within the data. Using this approach, we identified distinct dimensions that represented the underlying patterns and relationships among the data. The process began with discriminant axis extraction, followed by orthogonal rotation using the varimax criterion. This approach ensures that each variable's loading is predominantly associated with one discriminant axis, thereby improving the clarity and interpretability of the perceptual map.