Affective Response Dimension Selection for Product Design: A Comparison of Cluster Analysis and Procrustes Analysis

Abstract

In r ecent years, the relat ionship between consumers’ aff ective res ponses (CARs) and product form features (PFFs) has been an important iss ue in the in dustrial desig n field. Responding to con sumers’ feelings towards a product’s app earance, CARs are usually presen ted in the f orm of a choice of adjectives. Based on the Kan sei Engin eering (KE) con cept, t his stud y condu cted Cl ustering A nalysis (CA) and Procrustes Analysis (PA) to f ind the CARs of a p roduct’s shape, and com pared the res ults of CA and PA. In the initial stage of the study, 75 samples of mobile phones were collected from the Taiwan market p lace. Twen ty-two p airs of adj ectives d escribing the cell p hones were used for a S emantic Differential (SD) experiment. Two-stage clustering was implemented to find the clustering segmentations of the affective responses according to the factor loading from the Factor Analysis (FA), and to obtain representative pairs of adjectives within t he clustering segmentations. P A was also used to decide adjective priorities according to the sorting rule. The KJ (Kawakita Jiro) method was used to verify both CA and PA. Finally, these two methods were analyzed and compared.