Emotion-Based Interaction Mode Selection Model for University Students

Abstract

User interaction is one of the biggest concerns for designers, programmers, and engineers since the user decides the acceptance of a technological solution, so the user must feel comfortable with the created solution. Since few studies adapt technology to users, it is necessary to conduct a study on user preferences to understand their bases. This work develops a validated survey to identify the demographic, technological, and expertise characteristics that influence a student's affective state according to a given interaction mode. This survey goes hand in hand with a model that predicts the emotional variables that a user will have when using one of these modes in order to obtain the emotional state of the person when using a specific device. A unique methodology was developed by combining an experimental process with CRISP DM and integrating its final steps to jointly create the survey and the model. The study included 64 university students who participated in a survey in which demographic, preference, and technological expertise data were collected. Participants' performance and affective states (valence, dominance, and arousal) were measured during their interactions with each device. Significant correlations between personal characteristics, performance, and affective states were identified for each device used. We also obtained the characteristics that influence the person's affective state and trained a model capable of predicting the affective variables with a sufficient amount of data.