Work

On the Validity of Principal Component Analysis as a Method of Analysis of the Variations in Human Physique

Year: 1979

Type: article

Abstract: Principal component analysis (PCA) and cluster analysis were applied to a correlation matrix based on 44 somatic measurements taken on 112 males aged from 18 to 27 years, and the results were compared with those of previous studies. The purposes of this study are: 1) to clarify the reason why the 3rd principal component is varied in meaning from case to case, and 2) to confirm that PCA offers effective criteria for the grouping of the measurement items. Initial selection of the measurement items is considered to be responsible for the varied manifestation of the 3rd principal component. All the studies reviewed show similar grouping of the measurement items based on the first three principal components, except for the location of sitting height. Cluster analysis supported the grouping based on PCA on the whole. Selection of measurement items was attempted on the basis of above-mentioned grouping, and tentatively 20 items were selected.

Source: The Journal of Anthropological Society of Nippon

DOI: 10.1537/ase1911.87.423

Author: Makiko KOUCHI

Institution: The University of Tokyo

Language: English

Topic: Statistical Methods and Applications

Subfield: Statistics and Probability

Field: Mathematics

Domain: Physical Sciences

SDG: Good health and well-being

Open Access status: diamond

Funder: University of Tokyo