Multi-Level Modelling of Haptic Perception Based on Physical Properties

In this doctoral thesis, a field-tested experimental method for measuring subjects' material perception is presented. The fully automated measurement process made it possible to utilize N = 48 subjects sample within a short period to develop a highly detailed differentiation of six representative polymers in two surface finishes. Based on psycho-physical, affective, and quality perception dimensions, the process reaches a variance explanation of about 60% and is extended by material performance perception. The extracted perceptual components show statistically significant differences across all materials in performance and psycho-physical perception, with strong effect sizes and a weak effect size for quality perception. In detail, the materials substantially differ in fineness, roughness, adhesion, hardness, and durability perception. Between surface finishes, significant differences in performance perception with a strong effect size and quality perception with a weak effect size are found. Polymers differ strongly in psycho-physical perception and weaker in quality and affective perception. The extracted perceptual components are used as criteria for multilevel analyses. Multilevel analyses are performed to control the individual frame of reference of sensory perception. The nested data structure is confirmed by an average ICC of .22. In the first modezling step, psycho-physical from physical parameters are predicted. In the second step, the affective-, quality, and performance perception components are predicted from psycho-physical/physical parameters.