Abstract:

Interest in virtual reality (VR) as a disruptive learning tool has continuously raised along the last fifty years. Since its beginnings, lecturers, trainers and instructional designers have highlighted the VR potential to improve attention, retention and effectivity of learning, taking advantage of its immersion, engagement, measurement and feedback capabilities.

Trying to establish more detailed, non subjective evaluation methods of learning experience (LX) and thanks to new sensor and tracking technologies, researchers have added psychophysiological and behavioural metrics to their environments.
While electromyography (EMG), electrodermal activity (EDA), electroencephalography (EEG), functional magnetic resonance imaging (fMRI) are the most used and validated psychophysiological metrics, they are being combined or replaced by a very versatile, powerful metric: eye tracking.

Such physiological measures are found to be more effective than self-reported measures since
1) they are not intrusive (they do not disturb participants during the measurement),
2) they do no rely on participants’ own assessment of the emotional or cognitive experience, and
3) they can detect changes in participants in real time (that also allows to react in real-time).

Very first eye tracking systems had low accuracy, low usability and were not cost effective. Today, due to last technological innovations, we have cost effective, accurate, portable, easy to use HMD with integrated eye tracking systems, as HTC Vive (Most advanced comercial HMD system with roomscale tracking, http://www.vive.com) - Tobii (World leader company in eye tracking systems, http://www.tobii.com) bundle.

This scenario allows us to obtain a continuous log of user behaviour through his or her gaze while using a virtual reality based learning environment.

With such amazing raw data, researchers are able to assess learning experience (LX). Eye trackers measure the saccades (fast movements) and fixations (dwell times) of human gaze. Due to correlation shown between eye fixations and attentional focus, we are able to have some insights about cognitive and attentional processes in VR enviroments' users.

In LX, eXperience is used in this paper to create a four dimensional abstraction for emotions, flow, cognitive load and immersion experimented by users.
Eye tracking has showed an interesting potential in all four dimensions:
1) Assessing emotions: Pupillary changes have been associated with arousal. Level—pupil diameter increases when people react emotionally to either pleasant or unpleasant stimuli.
2) Assessing frustration/satisfaction: The lower the number of eye regressions (back tracks to already seen areas), the higher the level of satisfaction.
3) Assessing cognitive load: The greater the fixation duration the greater cognitive effort.
4) Assessing immersion: for an immersive learning environment, the users' number of fixations per second will decrease, as their attention becomes more focused.

In this paper, we are going to analyze how eye tracking is going to help us assessing the experience of learners using VR environments and how, if we are able to integrate and add real time support to this aspects of LX, we will be really close to design and develop adaptive environments that react to users behaviours and feelings.

Keywords:

Virtual reality, education, eye tracking, learning experience.