Enhancing the quality of safety training in the construction industry

Safety training is a critical aspect of construction safety programs, equipping workers with the skills and knowledge needed to operate safely in the high-risk and ever-changing environment of construction sites. Although many advanced training methods have been developed to improve safety outcomes, their widespread adoption has been limited due to high costs, time constraints, and scalability issues. Moreover, the long-term effectiveness of these interventions has not been thoroughly assessed. Therefore, a comprehensive understanding of what constitutes high-quality safety training—both quantitatively and qualitatively—while considering industry constraints and timelines and assessing the short-term and long-term effectiveness of such training, is essential for enhancing construction safety training programs for both practitioners and academics.This dissertation aims to: (1) identify the key attributes of high-quality safety training based on the perceptions of safety trainers and workers in the construction industry; (2) use these findings to define and develop several training delivery methods that consider time and cost constraints, focusing on enhancing knowledge-based learning outcomes and assessing their short-term effectiveness (3) evaluate effectiveness of these training delivery methods on the learning retention outcomes; and (4) build upon the findings from the previous studies to test a safety training approach, aimed at improving the behavioral aspects of safety learning outcomes.The analysis of trainers' and workers' perceptions of high-quality safety training showed no significant disagreements between the two groups, although certain attributes were noted only by one group. Both groups identified important elements such as well-balanced, clearly explained content that includes the reasoning behind safety rules, the use of storytelling, the presence of an experienced and engaging trainer, and a combination of theoretical and hands-on activities as indicators of high-quality training. Following these insights, five training delivery methods were developed: pre-recorded video, lecture, interactive lecture, flipped lecture, and interactive lecture with hands-on activity. A quasi-field experiment conducted among 591 construction workers at various jobsites, using experienced trainers, demonstrated that as the learner-centeredness of the training methods increased— from pre-recorded video to interactive lecture with hands-on activity— trainee engagement increased in a linear pattern. However, hazard recognition performance did not follow this trend; instead, it followed a quadratic trend, with the highest performance observed in both the least learner-centered (pre-recorded video) and the most learner-centered (interactive lecture with hands-on activity) groups. An analysis of learning retention one month later revealed a similar pattern, with the least change in hazard recognition seen in the interactive lecture group. Lastly, using previous findings, the impact of adding positive and negative story-based examples to pre-recorded video training was tested among civil engineering students. Results showed that while story-based examples did not affect risk tolerance, negative stories had a detrimental effect on risk perception, and positive stories had no significant effect.This study provides valuable guidance for construction industry trainers, highlighting that while increasing learner-centeredness during training may boost engagement, it may not necessarily improve learning outcomes. A consistent finding across all experiments was that energy-based hazard recognition training, which provided only the definition of energy-based hazards, significantly enhanced hazard recognition performance. This underscores the importance of well-defined content, even when limited to basic concepts. Future research should focus on developing clear and effective definitions for other learning objectives.