【正文】 ), magenta (320176。 and 220176。) also form a cluster. Although there are other clusters, such as the one containing green (80176。 ??Procedure??). This experimental site was created with Dreamweaver MX 2020 and developed in html and php. 7 Participants Eighty participants took part in this first study: – fifty Website “users” or visitors (25 men and 25 women, with a mean age of years), from a variety of backgrounds (university students, teachers, professionals carrying out administrative functions, etc.). – thirty Website designers (20 men and 10 women, with a mean age of years). This group was smaller than the user one, due to the lack of availability of professional designers. They, too, came from different backgrounds (., qualifications in puter science, design engineering or applied arts), but all of them had been in a job for between 2 and 35 years in which they were regularly called upon to create Websites. All these participants were French and none of them was colourblind. Procedure Participants were given a password to access the experimental site. When they logged onto it, they were all given the same instructions (see Appendix A), and the same 23 homepages were presented in random order. For each homepage, they had to indicate how much they liked it on a 7point Likert scale, ranging from “not at all” (point 1) to “a lot” (point 7), without intermediate labels associated to points going from 2 to 6. At this point, the process varied slightly, according to the participants status. They all had to answer four questions relating to general information: their gender, age, the average time they spent on the Inter each day, and their background. In addition, the Web designers had to answer supplementary questions about the colour they would have chosen for this homepage, the reasons for their choice, the role of colour in Websites and any training they had received about Website colours. Finally, all the participants read the last page of the Website, which signalled the end of the experimental task. This page provided information about contacting the experimenters and terminated their access to the experimental site, so that they could only perform the experiment once. 4 Results The data analysis was performed in two stages. The first analysis served to identify and select the colours that both the users and the designers preferred. The second analysis pared the users? and designers? preferences. First, the subgroups of colours were identified and selected on the basis of a hierarchical cluster analysis, using the squared Euclidean distance (Ward?s method). Linkage distance is 8 reported as Dlink/Dmax, which represents the quotient between the linkage distances for a particular case divided by the maximal linkage distance. The quotient was multiplied by 100 for standardize the linkage distance. Although the cluster boundaries differed according to the participants’ status, the hierarchical cluster analysis allowed us to identify clusters corresponding to the following colours: green, magenta, blue, orange and grey (see Fig. 4a and b). The inspection of Fig. 4a showed that, for the users, the most clearly discernible clusters are the ones containing the greens (100– 160176。 ), as well as between one aesthetic feature in particular (product colour) and users? preferences (167。 2 ). Emotions, cognition and product design The numerous papers on the subject of emotions offer divergent points of view about emotional phenomena ., [21,5,22]. Scherer[22] suggested distinguishing between various affective states and,in particular, contrasting utilitarian emotions and aesthetic two kinds of emotion result from an appraisal of environmental or proprioceptive information, but have different functions. Utilitarian emotions, such as anger and fear, allow us to adapt to events that may have major consequences for us as individuals. These adaptive functions may consist in the preparation of actions (such as confrontation or escape) or the recuperation and reorientation of work. In contrast, aesthetic emotions are unrelated to the need to satisfy vital and mandatory instance, a person can be impressed, admiring or diffuse sensations differ considerably from utilitarian emotions with regard to felt arousal and behaviour orientation. Although emotion is not understood as well as cognition, both of them can be regarded as information processing systems [23], but with different functions and operating parameters. Cognition allows us to interpret the world and make sense of it, whereas emotions are more judgmental, assigning positive and negative valences to the environment [6,7,5]. Advances in our understanding of cognition and emotion suggest that each impacts the , several studies have shown that emotions and affects have an impact on the cognitive functioning of individuals engaged in plex tasks [24,21,25]. According to Norman [7] and Csikszentmihalyi [26], emotions can even make us smart. For instance, positive affects broaden our thought processes and seem to enhance our creativity in finding design solutions [9]. Unlike negative affects, positive affects make difficult tasks easier to perform and make people more flexible and more tolerant of minor difficulties [6]. These approaches have implications for the design of products and interactive systems. In the field of HCI, a cognitive– affective model of organizational munication has been developed by Te?eni [27]. This author posits the notion that the munication process, which prises the munication medium (a Website in the case of our study) and the message form (., differences in the colours used for a Website), has an impact on the user and on whether the latter judges the munication to be appropriate. In line with this model, Norman [7] claims that the image which products present to the us