A domain is defined describing the product type in question; the target group and the context situation of the product. This domain is then described using verbal expressions erecting the Semantic Space (Osgood; Suci;& Tannenbaum; 1957). In parallel the Space of Properties is set; defining relevant product traits. Kansei Engineering then uses various tools in order to establish links between the two spaces. These tool use userdata often collected by questionnaires or observational studies. After validation a prediction model is build quantifying the relationship between affective meaning and product trait. The relationship can be either mathematical or descriptional. In practice these steps are undertaken using manual methods; but recently also data collection on the internet has been proven successful.
Engineering methods (including Kansei Engineering) used in product development are often reductionistic; i.e. they simplify reality into less complicated models in order to increase understanding and handle complex environments. Usually this works out; but there is a risk of loosing important information when reducing reality to a small scale model.
In order to avoid this in Kansei Engineering good validated Kansei procedures are required. The participants must be enabled to get an as complete affective impression of the product in question (Kansei) as possible; but not to provide more affective information than necessary. For example for evaluating the sound of closing car door a sound file might be sufficient. Presenting more information such as a picture or video of the car will make the study more complex for the participants at the same time increasing the risk for biasing the results. Not much research has been done in this area yet; or at least only few research results are published.
Keywords: Affective flow model; indirect design; honest products; proximity of interaction; proximity of presentation; optimization of Kansei Engineering study design