Canonical Correlation Analysis for Estimation of Relationship between Anthropometric Measurements and Office Furniture Measurements in Public Institution
Adu, G; Adu, S; Asare, K.P; Darkwa, N.A

Abstract
Canonical correlation analysis (CCA) was adopted to develop relationships between anthropometric measurements and furniture measurements to enhance some healthy sitting posture. Estimation of relationship between two groups of measurements in public institutions is achieved through CCA. The objective of this research was to apply canonical correlation analysis to establish the relationship between human body measurements (Anthropometric Measurements) such as popliteal to floor height, buttock to popliteal length, elbow to seat height, sitting shoulder height, knee height and width of bitrochanter and furniture measurements such as seat height, seat depth, armrest height, backrest height, desk clearance and seat width in public institutions. In this study, canonical correlation analysis was applied to incorporate many variables within a single research analysis to develop relationships between those variables. The results showed that six canonical correlations were obtained (0.830, 0.666, 0.350, 0.283, 0.173 and 0.056). Among the estimated correlation coefficients, the first three pairs were significant (p < 0.001 in each). In the first pair of canonical variables, the estimated relationships were between seat width and width of bitrochanter and, seat width and sitting shoulder height. For the second pair of canonical variables, the relationship were between desk clearance and width of bitrochanter; seat width and width of bitrochanter; and between desk clearance and seat width within the furniture variables. Finally, the third pair of canonical variables was between seat height and popliteal to floor height.

Full Text: PDF