Efforts to develop existing infrastructure facilities are highly regarded in order to keep up with capacity demand and upgrade changes in data trends. To obtain the best users’ interest in the facilities that aligns management development plan and schedule, a questionnaire is commonly conducted. Datasets acquired from questionnaire featuring satisfactory level such as Likert scale tends to be ordinal. Ordinality using standard Pearson correlation lean towards weak relationship. Traditional PCA, relying on Pearson correlation, may struggle to capture the nuanced relationships within such ordinal data, leading to a loss of valuable information. Through a comparative analysis of PCA results using both covariance matrix and conventional Pearson correlation, this paper demonstrates the efficacy of the proposed methodology in uncovering latent patterns and relationships within the questionnaire responses.
