@Article{ijmhp.2025.066369,
AUTHOR = {Hao Fang, Hongyun Guo, Yinchao Chen, Hui Shi, Yihan Gan, Lin Li},
TITLE = {The Impact of Virtual Reality Environment Design on Emotional Recovery: Exploring Factors and Mechanisms},
JOURNAL = {International Journal of Mental Health Promotion},
VOLUME = {27},
YEAR = {2025},
NUMBER = {7},
PAGES = {1051--1069},
URL = {http://www.techscience.com/IJMHP/v27n7/63233},
ISSN = {2049-8543},
ABSTRACT = { <b>Objectives:</b> Emotional stress is a significant public health challenge. Virtual reality (VR) offers the potential for aiding emotional recovery. This study explores the impact of VR environment design factors on emotional recovery, examining underlying mechanisms through physiological indicators and behavioral responses. <b>Methods:</b> Two experiments were conducted. Experiment 1 employed a 4 [Scene Type: real environment (RE), virtual scenes that restore the RE (VR), virtual scenes that incorporate natural window view design (VR-W), and a no-scene control condition (CTL)] × 3 (Experimental Phase: baseline, emotion arousal, recovery) mixed design (N = 33). Participants viewed a 4-min anxiety-inducing video followed by a 3-min scene exposure. State-Trait Anxiety Inventory-State Form (STAI-S), galvanic skin response (GSR), and blood-volume pulse (BVP) frequency were analyzed with linear mixed-effects models. Experiment 2 used a 3 (Motion-control mode: Unnatural, Semi-natural, Natural) × 2 (Sound form: Spatial positioning, Surround) × 3 (Experimental Phase: baseline, emotion arousal, recovery) mixed design (N = 42). Presence was analyzed with the Scheirer–Ray–Hare test; phase efficacy was verified with Friedman tests. <b>Results:</b> Experiment 1 showed significant Scene Type × Experimental Phase interactions for GSR (F = 8.006, <i>p</i> < 0.001, η<sup>2</sup><sub>p</sub> = 0.624) and BVP frequency (F = 11.491, <i>p</i> < 0.001, η<sup>2</sup><sub>p</sub> = 0.704). VR-W produced the largest recovery (ΔGSR = –1.26; ΔBVP = –5.80; Hedges g ≥ 0.83) vs. RE and VR. STAI-S returned to baseline across all Scene Types. Experiment 2 revealed main effects of Motion-control mode (F = 8.55, <i>p</i> = 0.001, η<sup>2</sup><sub>p</sub> = 0.32) and Sound form (F = 4.35, <i>p</i> = 0.044, η<sup>2</sup><sub>p</sub> = 0.11) on Presence (Semi-natural + Spatial positioning highest). The greatest physiological recovery occurred with the Unnatural Motion-control mode (GSR H = 20.17, <i>p</i> < 0.001, ε<sup>2</sup> = 0.49; BVP H = 7.92, <i>p</i> = 0.019), amplified by Spatial positioning Sound form only in this mode. Design factors did not influence STAI-S change. <b>Conclusions:</b> VR scenes are as restorative as RE; embedding VR-W accelerates recovery. Maximal Presence is not essential: Unnatural Motion-control mode induced the largest physiological recovery, especially combined with Spatial positioning Sound form.},
DOI = {10.32604/ijmhp.2025.066369}
}