Empirical Comparison of Learning Effectiveness of Immersive Virtual Reality–Based Safety Training for Novice and Experienced Construction Workers

Abstract

Statistics show that novice and experienced construction workers face different risks of construction accidents. However, there is very little work on this topic. This study conducted empirical experiments to quantifying the effectiveness of immersive virtual-reality-based safety training (IVST) across 6 worker specialties in terms of facing 17 hazardous scenarios at real construction sites. An immersive virtual reality–based safety training system (IVSTS) is proposed to measure the effectiveness of IVST programs with regard to six specialties in 2 controlled groups: (1) novice workers (40 college students), and (2) experienced workers (40 randomly selected construction practitioners). The study results indicate that the proposed IVSTS improved the safety performance index of all 80 participants by 14.12% in terms of correct selection of personal protection equipment, and by 28.95% in terms of hazardous scenario identification. The improvement of safety learning performance of the novice workers was better than that of the experienced workers, especially for hazardous scenario identification, with a 15.90% higher improvement index. The improvement suggestions for the existing IVST program were summarized according to the influence of the individual characteristics of the trainees on the learning effect between the novice and the experienced IVST programs.