CLASSIFICATION OF MENUS ON HOME DINING TABLES BASED UPON HUMAN MEAL Kansei

Abstract

To evaluate human meal Kansei methods of paired comparisons are considered to be used. This paper concerns a method to classify menus on home dining tables by using the method of paired comparisons. There are many methods of paired comparisons proposing and among those methods the method proposed by Thurstone is used for this study because the burden of the subjects in the experiment to obtain the data on humanKansei is comparatively light. In Case V of Thurstone's method of paired comparisons a model is assumed that human sensations are normally distributed with equal standard deviations, and the scales are obtained so that the units of them are the standard deviations of the standard normal distribution hence the scales are useful to compare Kansei among many stimuli. It is however necessary for the use of Case V to test goodness of fit to the model Thurstone has assumed. An experiment by paired comparisons to obtained data on human meal Kansei was conducted using 200 menus on home dining tables and goodness of fit to the model was tested. For the results it was concluded that meal Kansei for 151 menus among 200 menus does not fit the model assumed in Case V of Thurstone's method. It is therefore decided that the scales obtained for human meal Kansei are not effective hence a method is considered to classify menus on home dining tables into several groups according to human meal Kansei based upon the scales obtained by the method of paired comparisons. Here it is described how to classify menus on home dining tables according to human meal Kansei and some of the results of classification are shown.