Abstract
Visual appearance of a product is an important factor to attract customers. Before purchasing a product, customers see a product in different visual conditions such as a photo in a catalog and a real product from certain distance. Impressions of a product appearance may differ depending on the visual conditions. If the appearance in a condition prior to seeing real one (e.g. a photo in a catalog) looks better than appearance of a real one in use environment, customers may be disappointed. Thus, minimizing such visual disconfirmation is a vital issue to keep customer's interest. In this paper, we proposed a method for analyzing visual factors with respect to disconfirmation of product appearance between different visual conditions. We used housing interior materials as a case study and examined factors of visual expectation disconfirmation between small samples and full scale samples. As a result, we found a significant effect of hybrid images (spatial characteristic) and optical illusions of color on visual expectation disconfirmation. We further explored a possibility for analyzing visual expectation disconfirmation by the use of eye gaze features of an observer. We examined relations of awareness patterns such as "as expected", "mistake", "new finding", "unexpected" and eye gaze features using low-pass pictures and original pictures. As a result, we found that Kullback-Leibler Distance (KLD) has a potential to discriminate between "as expected/mistake" and "new finding/unexpected".
