A statistical model for developing body size charts for garments

A simple easy to follow statistical approach has been proposed for developing body size charts from anthropometric data. It has been possible to cover 95 percent of the population using 11 size charts. Multivariate analysis was carried out to detect relationships between variables. Principal component analysis was carried out to identify the key body measurements which can form the basis for classifying the population. Bust for the upper body and hip for the lower body were identified as the critical dimensions affecting garment fit. Body shapes and their distribution within the population have been identified. Validation of size charts was done by calculating the aggregate loss of fit.