| description abstract | Effective safety management at construction sites requires accurate hazard recognition. However, previous research studies have revealed a significant proportion of hazards remain unrecognized. Poor hazard recognition may be attributed to the deficiency at one or more levels of situational awareness (SA). Understanding how different hazard types, particularly those with high fatality rates, influence workers’ SA is critical to improving safety outcomes. To address this research gap, three hazard types with different fatality hierarchies (F-I: fall, F-II: struck-by, and F-III: electrical hazards) were built in a complex and dynamic virtual reality (VR) construction environment. Participants’ hazard recognition behaviors regarding hazard perception, comprehension, and projection (Level 1, 2, and 3 SA) were measured by a well-recognized situational awareness global assessment technique (SAGAT). The findings revealed that the achievement of SA varied among different hazard types and was impacted by fatality hierarchy. There was a consistent decline in success rates from Level 1 to Level 3 SA across all hazards, and the potential to achieve a high-level SA demonstrated a positive correlation with the hazard fatality hierarchy. Specific reasons for failure at each SA level attributed to shortcomings included a deficiency in prior knowledge, narrowed attention, and the increased demand for cognitive resources. The findings enhance the understanding of workers’ SA in hazard recognition behaviors and provide a foundation for developing customized interventions. These interventions can be tailored to address SA deficiencies at different levels based on hazard types with distinct fatality hierarchies, ultimately improving safety performance at construction sites. | |
