Article Figures & Data
Figures
- Figure 1.
Experimental design combining two physical BOs and the presence/absence of an avatar conditions. a, Participants performed the task in four different conditions obtained by the combination of the physical BO (i.e., Supine vs Standing) and the presence/absence of the avatar (i.e., Body vs Nobody). Note that the avatar (when shown) was always in a supine position. b, Participants performed the virtual navigation task, wearing a VR head-mounted display while either standing upright or lying supine on a bed. c, During the task, participants were navigating in a circular virtual arena, performing a spatial memory retrieval. In Encoding session, preceding each Retrieval session, a participant memorized locations of the target objects. At each trial of Retrieval session: (1) the image of the target object was shown; (2) a participant navigated to the retrieved location and responded; and (3) as feedback, the target object appeared at its correct location to be collected, vm: virtual meter.
- Figure 2.
Effect of physical BO on experienced BO and on navigational behavior in VR. a, When participants were physically supine (blue), they had a stronger experience of being supine in VR while they felt less as if they were standing, compared with when they were actually standing upright. The conflict score indicates how much their experienced BO in VR conflicted with their physical BO. The results showed that the conflicts were significantly larger in the physically supine BO condition than in the standing BO condition. Also, overall, when compared, regardless of their physical BO, participants reported a significantly stronger experience of being standing in VR, suggesting the standing BO as a presumed BO in VR. b, Participants’ reached location was farther away from the border when they were physically supine compared with when they were standing upright. c, A top-view schematic figure depicting the border effect. The changes in the distance from the border possibly reflected changes in self-location (or body boundaries) that are related to the change in the experienced BO in VR. Consequently, the distance from the border (calculated with respect to the viewpoint location) was greater when they were physically supine compared with when they were standing upright. **0.001 ≤ p < 0.01, ***p < 0.001.
- Figure 3.
Experienced BO affected by the BO-congruent avatar. a, When the participants were physically supine, presenting an avatar with supine posture (i.e., Body condition) strengthened the subjective experience of being supine and reduced the experience of standing upright compared with the Nobody condition, where no avatar was shown. The conflict score confirms that showing the BO-congruent avatar significantly reduced the conflict of the BO. When compared, regardless of the presence of the avatar, they felt more as if they were supine than standing, again highlighting the impact of the physical BO on the experienced BO in VR (Q_supine > Q_standing, overall). b, We also reproduced previous findings of a significant shift in the distance from the border between body and nobody conditions, suggesting that such drifts in self-location are associated with changes in the experienced BO in VR. **0.001 ≤ p < 0.01, ***p < 0.001.
- Figure 4.
Physical BO and the presence of the avatar affected navigation behaviors in VR. a, Distance from the border, which arguably reflects experienced BO in VR, was significantly larger when participants were physically supine rather than standing, and also when a supine avatar was presented in a first-person viewpoint position than no avatar was shown. Predicted mean value per condition in order: 6.86, 6.43, 7.62, and 7.27 vm. b, Distance Error, indexing spatial navigation precision, did not significantly differ among the four conditions. Predicted mean value per condition in order: 11.06, 10.60, 11.69, and 11.69 vm. c, Participants spent significantly less time in the retrieval phase when they were physically standing compared with when they were physically supine, regardless of the presence of the avatar. Predicted mean value per condition in order: 10.40, 9.34, 10.48, and 9.29 s. d, The navigated distances were significantly reduced in the Body conditions (i.e., Supine-Body and Standing-Body) compared with the Nobody conditions (i.e., Supine-Nobody and Standing-Nobody). Predicted mean value per condition in order: 79.18, 79.12, 76.29, and 76.07 vm. No significant interactions were found for all four parameters. A dedicated mixed-effects model was used for statistical assessments of each parameter. Single data points represent the measure per participant, and dot and whiskers on the left of each data cloud indicate predicted mean value and 95th percentile, respectively. *p < 0.05.
Tables
Questions Term Description Q1 Self-identification I felt as if what I saw in the middle of the scene was my body Q2 Threat I felt as if the threat (knife) was toward me Q3 Presence I felt as if I was located in the virtual environment Q4 Cyber-sickness I felt dizzy Q5 Control I felt as if I had 3 bodies Q6 Supine I felt as if I was supine in the virtual environment Q7 Standing I felt as if I was standing in the virtual environment At the end of each session, participants were provided seven questionnaire items and answered on a Likert scale ranging from 0 (strongly disagree) to 6 (strongly agree).
Figure Variables Data structure Type of test Power (a) Figure 2a Q_Supine: supine-nobody vs standing-nobody Ordinal data Wilcoxon signed-rank test r = 0.783 (b) Figure 2a Q_Standing: supine-nobody vs standing-nobody Ordinal data Wilcoxon signed-rank test r = 0.593 (c) Figure 2a Q_Standing vs Q_Supine in standing-nobody Ordinal data Wilcoxon signed-rank test r = 0.841 (d) Figure 2a Q_Standing vs Q_Supine in supine-nobody Ordinal data Wilcoxon signed-rank test r = 0.093 (e) Figure 2a Conflict score: supine-nobody vs standing-nobody Ordinal data Wilcoxon signed-rank test r = 0.762 (f) Figure 2b Distance from the border: supine-nobody vs standing-nobody γ distribution Mixed effects model (fixed effect: BO, random effect: Subject, random slope: BO, nested: object) df = 1, F = 11.08 (g) Figure 3a Q_Supine: supine-nobody vs supine-body Ordinal data Wilcoxon signed-rank test r = 0.720 (h) Figure 3a Q_Standing: supine-nobody vs supine-body Ordinal data Wilcoxon singed-rank test r = 0.595 (i) Figure 3a Q_Standing vs Q_Supine in supine-body Ordinal data Wilcoxon signed-rank test r = 0.759 (j) Figure 3a Conflict score: supine-nobody vs supine-body Ordinal data Wilcoxon signed-rank test r = 0.800 (k) Figure 3b Distance from the border: supine-nobody vs supine-body γ distribution Mixed effects model (fixed effect: BO, random effect: Subject, random slope: BO, nested: object) df = 1, F = 16.56 (l) Figure 4a Distance from the border γ distribution Mixed effects model (fixed effect: BO * avatar, random effect: Subject, nested: object) BO: df = 1, F = 10.51; avatar: df = 1; F = 44.65; interaction: F = 0.51 (m) Figure 4b Distance error γ distribution Mixed effects model (fixed effect: BO * avatar, random effect: Subject, nested: object) BO: df = 1, F = 0.57; avatar: df = 1; F = 7.78, interaction: F = 0.66 (n) Figure 4c Trial time Log-normal distribution Mixed effects model (fixed effect: BO * avatar, random effect: Subject, nested: object) BO: df = 1, F = 32.9; avatar: df = 1; F = 0.01, interaction: F = 0.09 (o) Figure 4d Navigated distance Log-normal distribution Mixed effects model (fixed effect: BO * avatar, random effect: Subject, nested: object) BO: df = 1, F = 0.06; avatar: df = 1, F = 25.9; interaction: F = 0.02
Extended Data Figure 2-1
Distance from the border serves as an objective behavioral measure of experienced BO (i.e., supine position) in VR. a, Distance from the border data were significantly correlated with ratings of Q_Supine (df = 1, F = 19.08, p < 0.001, n = 25). A mixed-effect model was used to assess their relationship. b, Their relationship at the within-subject level was further assessed through in-depth analysis. The distance from the border and Q_supine data were re-calculated and plotted with respect to the Supine-Nobody condition (i.e., scanner condition). We found that a change in the distance from the border of a subject in a condition was significantly associated with the change in the Q_supine rating of the subject in the condition. **: 0.001 <= p < 0.01, ***: p < 0.001. Download Figure 2-1, TIF file.
In this issue
