Abstract Background The civil engineering construction industry has long faced high accident rates and severe occupational psychological challenges. Research has shown that over 40% of construction workers report experiencing psychological distress such as anxiety and depression, and their level of psychological resilience is significantly negatively correlated with the incidence of safety accidents. Traditional mental health intervention models, such as lectures or group counseling, have limited effectiveness due to low participation and unrealistic scenarios. Therefore, exploring an immersive, quantifiable, and realistic high-pressure training method has become a key requirement for improving the psychological safety and operational efficiency of practitioners. Virtual reality (VR) technology provides an innovative path for this. Methods A targeted VR psychological resilience training system has been designed and developed, consisting of three progressive modules: a high-risk environment adaptation module (such as high-altitude and enclosed spaces), a sudden crisis stress module (such as simulating material slippage and equipment abnormalities), and an emotion regulation and cognitive reconstruction module. The study used a randomized controlled trial to randomly divide 162 frontline construction workers into an experimental group (n = 81, receiving standardized VR training twice a week for 6 weeks) and a control group (n = 81, receiving routine safety training of equal duration). Using the Psychological Resilience Scale, State Trait Anxiety Scale, Physiological Multichannel Instrument (to monitor Heart Rate Variability (HRV) and skin electrical response), and simulated task performance indicators, data collection was conducted three times before intervention, after intervention, and one month after intervention. The data analysis used repeated measures analysis of variance, independent sample t-test, and effect size calculation (Cohen's d). Results The data results indicate that the VR training system has a significant and sustained effect on improving psychological resilience. After intervention, the experimental group showed an average improvement of 38.4% in the total score of the psychological resilience scale compared to the control group (F = 25.73, p.001, η2 = 0.14), with the most significant improvement in the dimension of "resilience" (d = 1.21). Physiological data shows that the experimental group showed an average increase of 52.7% in high-frequency power of HRV (reflecting parasympathetic nervous system activity and recovery ability) and a reduction of about 45% in stress recovery time (p.01) when responding to sudden simulated emergencies. In simulating high-risk tasks, the experimental group's operational error rate decreased by 31.2% compared to the control group, and the decision response time was shortened by 22.5% (p.01). Tracking tests show that the positive effects mentioned above remain stable after one month (with a decay rate of8% for each indicator). Discussion The VR psychological training system effectively enhances the psychological resilience and stress management ability of construction workers through situational immersion, and data confirms its superiority in short-term interventions. The high interactivity of the system and the degree of scene simulation may be key driving factors. In the future, it is necessary to expand the sample and conduct long-term tracking to verify the sustainability of the effect; At the same time, it can integrate biofeedback and multi-user collaboration modules to further optimize system universality.
Yixiong et al. (Sun,) studied this question.