Supporting Dyadic Chat Communication with Eye Tracking Based Reading Awareness

Synchronous text-based communication has been used in CSCL scenarios and workplace learning quite extensively, e.g. in Virtual Math Teams (VMT). Yet, this type of communication suffers from inherent problems often referred to as chat confusion. Macro-level problems such as managing multiple chats, private messages, notifications and presence awareness have been solved. Micro-level tensions at a conversational level still exist, despite several potential solutions proposed by research and industry. The only features that are widely accepted are is-typing and a simulated has-been-read notification. Their adoption is presumably related to not requiring any user effort. However, incoherent chat logs and parallel utterance composition still lead to re-reading, revisions and withdrawal of incomplete messages. In a qualitative study, we evaluate the use of interactive eye tracking techniques to add a set of awareness features with the goal of exploring its impact on users' behavior and acceptance. Results show that awareness cues like message-seen, (re-)reading and read are supportive, as communication feels more direct and helps to predict context.