Diagnosing user perception and acceptance using eye tracking in web-based end-user development

In the recent years in the End-User Development (EUD) research there is a shift from the study of tools that focus on desktop graphical applications, to the development of EUD for web environments. To this end mouse and eye tracking methodologies are used to implicitly monitor the end-user behavior in real time. In this paper we explore the potential correlation between eye movements and end-user perception and acceptance in modern web-based EUD environments. The aim is to find out whether end-users' perception and acceptance attributes can be reflected on their eye behavior when interacting with a web-based, database-driven EUD system. To check out our research hypotheses we have conducted a field test using a prototype EUD tool based on a natural language approach (named 'simple talking'), to assist end-users in creating database-driven mobile applications. The results of the field test show significant correlations between eye behavior and acceptance and perception. Self-Efficacy is correlated to fixations of any kind. Risk-Perception is correlated to the increment of the pupil size. Perceived Ease-of-Use is correlated to fixations that turned into clicks and to the increment of the pupil size. And Perceived Usefulness is correlated to fixation duration.