Emotion Assessment in Evaluation of Affective Interfaces

nature of emotions. A second factor that contributes to the difficulty of improving accuracy is the abstract nature of emotions. To apply emotion assessment techniques for evaluating affective interfaces, we need to measure and quantify the abstract construct of emotion. Measuring emotions is then a quest to map the abstract world of emotions onto a mathematical, quantitative measure. Without a quantitative measure, it is difficult to evaluate whether an emotion assessment technique is accurate. Various models of emotion representations are presented. However, none of them is agreed to be a universal, comprehensive representation of emotions. This shows that current knowledge about emotions is not concrete enough for researchers to model emotions. Perception of emotion. Adding to the difficulties, the basic assumption of emotion assessment follows the questionable informational model of affective interfaces. The informational model treats affect as a piece of information that can be transmitted from one person to another without error; emotion assessment also assumes that it is possible to measure one’s emotion accurately. However, this model is criticized by some researchers such as Boehner et al. [8], which indicates that the informational model, specifically accuracy of assessing one’s emotions, is questionable and still under debate in the research community. Since the basic assumption of emotion assessment is not solid, employing this technique for evaluation is vulnerable to criticism. Improving evaluation techniques. Developing good knowledge of the concept of accuracy in emotion assessment is essential for the evaluation of affective interfaces. However, due to the nature of emotions, it is difficult to determine the accuracy of emotion assessment. Applying emotion assessment techniques to affective interface evaluations is open to pitfalls. Evaluators of affective interfaces can follow a pessimistic approach, accepting these pitfall as limitations. Nevertheless, a more pro-active alternative is open to researchers: improve the evaluation techniques. CHAPTER 5. DISCUSSIONS 46