The reality of virtual reality

Abstract

Virtual reality (VR) has become a popular tool for investigating human behavior and brain functions. Nevertheless, it is unclear whether VR constitutes an actual form of reality or is more like an advanced simulation. Determining the nature of VR has been mostly achieved by self-reported presence measurements, defined as the feeling of being submerged in the experience. However, subjective measurements might be prone to bias and, most importantly, do not allow for a comparison with real-life experiences. Here, we show that real-life and VR height exposures using 3D-360° videos are mostly indistinguishable on a psychophysiological level (EEG and HRV), while both differ from a conventional 2D laboratory setting. Using a fire truck, three groups of participants experienced a real-life (N = 25), a virtual (N = 24), or a 2D laboratory (N = 25) height exposure. Behavioral and psychophysiological results suggest that identical exogenous and endogenous cognitive as well as emotional mechanisms are deployed to process the real-life and virtual experience. Specifically, alpha- and theta-band oscillations in line with heart rate variability, indexing vigilance, and anxiety were barely indistinguishable between those two conditions, while they differed significantly from the laboratory setup. Sensory processing, as reflected by beta-band oscillations, exhibits a different pattern for all conditions, indicating further room for improving VR on a haptic level. In conclusion, the study shows that contemporary photorealistic VR setups are technologically capable of mimicking reality, thus paving the way for the investigation of real-world cognitive and emotional processes under controlled laboratory conditions. For a video summary, see https://youtu.be/fPIrIajpfiA.

Keywords: EEG; HRV; anxiety; height; virtual reality.

FIGURE 1

Photographs from the real-life setting: (1) Basket on the ground at the starting point. (2) Basket with firefighter and participant declining. (3) Basket at a maximum height of 33 m/108 feet. (4) Basket on the ground with firefighter and Insta 360 Pro VR camera instead of a participant. Photos taken by BS.

FIGURE 2

Average topographic oscillatory power amplitude [ln(μV2)] distribution across participants by frequency band (rows) and conditions (columns) while staying at the highest point of the ride in the fire trucks basket. The pictograms in the last row depict the experimental setup per condition. See Supplementary Figure 11 for the topographical distribution separately per phase and frequency band power.

FIGURE 3

Median band power [ln(μV2)] per group, phase, and frequency band. Negative power values result from the logarithmic transformation during preprocessing: The logarithm of values greater than zero and smaller than one is negative. Hence, negative power values are to be read as smaller power compared to positive power values. The respective tables indicate the statistical characteristics per comparison in a reduced overview. Significant differences between groups are marked respectively (*p < 0.05, **p < 0.01; ***p ≤ 0.001), and the effect size r is interpreted as follows: a, small effect; b, medium effect; c, large effect. Find a detailed report of all statistics in the Supplementary material.

FIGURE 4

Changes in the HRV parameters SDRR (A) and rmSSD (B) for the separate phases of the ride in the fire truck’s basket compared to baseline per condition: The phases were corrected for the baseline (delta = HRV during the respective phase minus HRV during the baseline). The zero line thus marks the baseline, which was therefore not depicted separately. The error bars depict the standard deviation. Significant differences between groups are marked respectively (**p < 0.01, ***p < 0.001).