2101 主観確率分布にもとづく期待効果の数理モデル視覚予測が触覚知覚に与える影響を用いた検証

Abstract

A user's experience of a product involves a set of transitions from one sensory state to another. For example, in the user's perception of a product's qualities, the state of their sensory modality may shift from vision to touch. Between such state transitions, users have expectations regarding their subsequent states based on their experience of a current state event. A disconfirmation between prior expectation and posterior experience evokes emotions such as surprise, satisfaction, and disappointment. Such emotions affect the user's perception of the value of a product. A noteworthy phenomenon in the perception of expectation disconfirmation is that the expectation affects the perceived experience itself. This psychological phenomenon called the expectation effect affects the user's perception of expectation disconfirmation. This effect is a key element for designing the affective experience of a product. Although experimental findings on the expectation effect exist in a variety of research disciplines, general and theoretical models of the effect have been largely neglected. In this paper, we propose a theoretical model of the expectation effect using information theory. We hypothesize that Shannon's entropy of the prior subjective probability distributions of posterior experience determines the occurrence of the expectation effect and that the amount of information gained after experiencing a posterior event is positively correlated with the intensity of the expectation effect. To verify these hypotheses, we conducted an experiment in which participants responded to the tactile qualities of surface texture. In the experiment, we extracted the visual expectation effect on tactile roughness during a sensory modality transition from vision to touch and analyzed the causes of the effect based on our hypotheses. The experimental results indicated the appropriateness of the proposed model of the expectation effect.