Not all eye-tracking methodology and data processing are equal. While the use of eye-tracking is intricate because of its grounding in visual physiology, traditional 2D eye-tracking methods are supported by software, tools, and reference studies. This is not so true for eye-tracking methods applied in virtual reality (imaginary 3D environments). Previous research regarded the domain of eye-tracking in 3D virtual reality as an untamed realm with unaddressed issues. The present paper explores these issues, discusses possible solutions at a theoretical level, and offers example implementations. The paper also proposes a workflow and software architecture that encompasses an entire experimental scenario, including virtual scene preparation and operationalization of visual stimuli, experimental data collection and considerations for ambiguous visual stimuli, post-hoc data correction, data aggregation, and visualization. The paper is accompanied by examples of eye-tracking data collection and evaluation based on ongoing research of indoor evacuation behavior.

During evacuation, individual navigation behavior is often dictated by the legibility of evacuation signs and the environmental setting itself. People tend to follow previously-used and known routes (to retrace) rather than follow evacuation signage. This has proven undesirable, even fatal, in emergencies and such behavior calls for a better understanding of the influencing factors. This contribution consists of a virtual reality experiment in which 72 participants evacuated from a hotel building in which the spatial configurations were altered. The tendency to retrace diminished when the evacuation route led through a wider, straight corridor, suggesting that the building's spatial configuration has an important influence on retracing and should be more carefully considered in the design of buildings in and agent-based simulations.

No abstract

The presented study aspires to utilize the gradually validated immense potential of collaborative immersive virtual environments (CIVEs) in higher education when designing and conducting geography lessons. These particular lessons focused on hypsography. A Research through Design approach and relevant qualitative methodology were used as we engaged two groups of domain experts (experienced geography teachers) to validate both the learning scenarios and the virtual environment we used. The lessons were administered via eDIVE—a novel platform for collaborative learning and teaching in virtual reality of our own design. The teachers underwent a hypsography virtual lesson and were randomly divided into two groups to be compared, which differed in the level of structure given to the lesson (one group received detailed instructions on what task they were to solve, while the other was given a free hand in exploring the environment and activities it afforded). The teachers’ experiences were then summarized in a post-lesson reflection and a subsequent focus group following the tasks. The participants’ expertise allowed insights to be gained into their first-hand experience as students, as well as their expert view of the lesson from an educational point of view. Virtual reality’s implementation into teaching practice was the key topic of the discussion.

The study evaluated the degree of comparability of human behaviour exhibited during an evacuation scenario conducted in two environments, one real and the other virtual. For this purpose, we created a precise 3D model (digital twin) of a real campus building. In both the experimental (virtual environment) and control (real environment) groups, the participants were all first-time visitors to the building. The groups were instructed to locate a target room where they would subsequently be required to complete a simple task, at which point an evacuation alarm would be set off. We evaluated three types of data collected during their quest to find a way out of the building: an eye-tracker logged gaze hits on various objects (e.g., navigation signs), recorded the locomotion and trajectory of the participants, and logged a combination of these parameters (e.g., points along the path where participants sighted certain objects). The sample was composed of 73 participants (35 for the real environment, 38 for the virtual environment). We analysed and confirmed the results and concluded that despite certain aspects of human behaviour differing during the simulated evacuation, the locomotion behaviour exhibited by the participants in both environments was generally comparable. To improve the potential use of these findings in practice, we selected easily accessible hardware and excluded expensive devices such as treadmills, etc.