Discriminant analysis of anthropometric measurements: A unisex ready-made trousers Casd study

Abstract

Fit to size is an important factor for consumers engaged into wearing unisex ready-made trousers. This study demonstrates how discriminant analysis can be used to improve the standard sizing system of ready-made trousers in Nigeria. To achieve this, the researcher measured 400 randomly selected students (200 male and 200 female) of Michael Okpara University of Agriculture, Umudike on seven anthropometric measurements; Waist, Hip, Length, Lap, Knee, Flap, and Base used in sewing trousers students. The data grouped into three (3) different sizes (small -s, large -l, and medium -m) on the basis of existing standard sizing measurement system. The assumptions of normality were established while that of equal covariance matrices was invalidated. Thus, the quadratic classification rule was employed per pair of the variables considered. On the overall, the analysis showed that about 3% of the sizes sampled were misclassified but when the derived classification function was evaluated it correctly classified about 97% of the sizes under study.