Desktop versus immersive virtual environments: effects on spatial learning

ABSTRACT

Although immersive virtual reality is attractive to users, we know relatively little about whether higher immersion levels increase or decrease spatial learning outcomes. In addition, questions remain about how different approaches to travel within a virtual environment affect spatial learning. In this paper, we investigated the role of immersion (desktop computer versus HTC Vive) and teleportation in spatial learning. Results showed few differences between conditions, favoring, if anything, the desktop environment. There seems to be no advantage of using continuous travel over teleportation, or using the Vive with teleportation compared to a desktop computer. Discussing the results, we look critically at the experimental design, identify potentially confounding variables, and suggest avenues for future research.

KEYWORDS:

• Immersive virtual reality
• spatial learning
• immersion
• viewpoint transitions

Acknowledgments

We thank Yu Zhong for assistance with the data collection. We also thank Dr. Danielle Oprean for her suggestions on the data analysis, as well as all the participants in this research.

Declaration of interest

The authors declare that they have no competing interests.

Data availability

The virtual environment used in the current study and the datasets along with the analysis code are available on request from the corresponding author, Jiayan Zhao.

Notes

¹ https://lithodomosvr.com/

² http://contrastvr.com/

³ There is, though, anecdotal evidence to suggest the support of a 10 × 30 meter tracking area through the integration of multiple Vive base stations.

⁴ Such incompatibility may be enhanced by the large projected field of view of the immersive display (Moss & Muth, 2011). Empirical studies suggest that the simulator sickness can be mitigated by, for example, limiting the travel speed to a relatively low level, or narrowing down the visual field to a small circle in front of the user during travel but relaxing the field of view when travel stops (e.g., Google Earth VR – https://vr.google.com/earth/).

⁵ The physical field of view is set physically by the actual size of the display and viewing distance of the user.

⁶ Motion parallax is a type of visual depth cue in which objects that are closer appear to move faster than objects that are farther away.

⁷ Spatial updating is the strategy that people adopt to process sensory cues received during spatial learning (Hart & Moore, 1973).

⁸ In a triangle completion task, the participant traverses two outbound path legs before pointing to or directly returning to the unmarked path origin.

⁹ The standard Virtual Silcton paradigm is an open-access online product that was first launched in 2013 and administered via desktop computer, mouse, and keyboard (https://osf.io/6dhfz/). It integrates virtual navigation, learning assessments, and analytic tools for the study of human navigation behavior.

¹⁰ https://unity3d.com/

¹¹ These were randomly assigned and easy-to-spell animal names, which were different from the standard Virtual Silcton paradigm in which buildings were named after famous geographers.

¹² Geometric field of view refers to the visual angle encompassing the virtual scene, which is equivalent to the field of view of the virtual camera and is adjustable by software.

¹³ We initially introduced both continuous travel and teleportation in the Vive. However, a pilot study showed that continuous travel in the Vive caused users to experience moderate to serious simulator sickness.

¹⁴ https://studyfinder.psu.edu/

¹⁵ Pointing error was measured as the absolute angular difference between the judged pointing direction and the actual direction of the target, resulting in a maximum possible error of 180°.

¹⁶ We used a navigation log file that was created by the standard paradigm to find out those desktop continuous travel participants who followed the same sequence as Vive teleportation participants when traveling along the main routes.

¹⁷ The standard Silcton paradigm used the front door of the target building as reference location toward which the pointing direction was judged absolutely correct. In contrast, the newly developed Vive application used the geometric center of the building as reference location. To make the pointing data comparable between VR conditions, pointing errors for Vive teleportation participants were recalculated based on the front door of buildings.

¹⁸ The perceived size of target buildings depends on two factors: 1) the actual size of the building and 2) the distance the building is from the eye.

¹⁹ Participants in Experiment 2 were recruited from a psychology participation pool. It is conceivable that most of them did not have any background in geo-related areas.

²⁰ Path integration, as the basic form of spatial updating, refers to the process by which people use sensory cues to continuously track their position and orientation relative to the origin or destination within the environment in the absence of suitable positional cues of targets (i.e., position-informative information; see He & McNamara, 2017 for review).

Additional information

Funding

This work was partially supported by the National Science Foundation under Grants [1526520] and [1526448].