Evaluation of the validity of anthropometric design assumptions

Abstract

Some designers assume that if an individual's stature is a certain population percentile, his/her other body dimensions will also be the same population percentile. It is also frequently assumed that anthropometric dimensions are normally distributed. This study aimed at evaluating the validity of these two assumptions. The subjects, 70 male and 70 female, were grouped according to their stature in the 5th, 10th, 25th, 50th, 75th, 90th, and 95th percentile categories. Each category had 10 males and 10 females. A set of 12 body dimensions was measured and recorded on each individual. A comparison between sample mean percentiles and corresponding population percentile values given in the published literature indicated that the assumption that a certain percentile stature person will also have his/her other body dimensions the same percentile was not validated. The normality assumption was evaluated by pooling all data for each dimension, separately for men and women, and performing the Kolmogrov–Smirnoff test. The results indicated that for males, 4 of the 12 dimensions (chest breadth, foot breadth, forearm-hand length and shoulder-elbow length) were not normally distributed (p<0.01). For females, 5 of the 12 dimensions (acromial height, chest breadth, foot breadth, foot length and hand breadth) were not normally distributed (p<0.01). The skewness for the non-normal dimensions were all negative (tail to left). For males, they were −1.4, −0.1, −2.7, and −0.5, respectively. For females, they were −0.2, −0.3, −0.3, −0.2, −0.3, respectively. For both men and women, the actual percentile values (5th to 95th) of the non-normally distributed body dimensions are provided. Relevance to industry This study shows that the body dimensions are not always normally distributed. Furthermore, the assumption that an individual has the same percentile body dimensions as his/her stature is also invalid. Such assumptions could lead to serious design compromises.