Abstract: This study simulated the stress responses of 50 virtual participants exposed to drip frequencies between 0.5 and 2 drops per second for durations of 30 to 120 minutes. Five predictive models LRSM, RFSM, SVM, ANN, and the proposed Hybrid Stress Prediction Model (HSPM) were evaluated across prediction accuracy, RMSE, physiological alignment, and computational efficiency. The HSPM demonstrated the strongest overall performance, achieving 85% prediction accuracy, surpassing ANN (80%), SVM (76%), RFSM (73%), and LRSM (70%). It also produced the lowest RMSE (0.25) and showed superior physiological alignment by accurately modelling both immediate and cumulative stress indicators, including heart rate variability and cortisol dynamics. While LRSM and RFSM underestimated delayed stress responses and ANN and SVM required greater computational resources, HSPM achieved a balanced trade-off between accuracy and efficiency, completing simulations in just 0.28 seconds. Sensitivity analysis revealed a 60% increase in cumulative stress under high-frequency conditions, confirming HSPM as the most reliable and realistic predictive framework.
Keywords: Water drip torture, psychological stress, Physiological response, Stress prediction and Hybrid modelling.
Title: Psychological and Physiological Responses to Prolonged Drip-Induced Stress Exposure
Author: Punam Hembram, Ashwini Raj Kumar
International Journal of Social Science and Humanities Research
ISSN 2348-3156 (Print), ISSN 2348-3164 (online)
Vol. 13, Issue 3, July 2025 - September 2025
Page No: 430-436
Research Publish Journals
Website: www.researchpublish.com
Published Date: 23-September-2025
