科研之友
微信
新浪微博
Facebook
分享链接
摘要
The construction industry in China has relatively high fatality and injury rates, and traditional practices for construction safety are no longer sufficient to meet new challenges. Innovative strategies to reduce safety hazards and accidents are required to substantially improve current practices. An understanding of the influences of the construction safety climate and personal experience construction safety becomes critical before any effective measure can be established. By considering previous safety climate models, a Bayesian network (BN) based model is proposed, establishing a probabilistic relational network among causal factors, including safety climate factors and personal experience factors that have influences on human behavior pertinent to construction safety. It therefore provides a methodology to identify potential strategies for safety improvement. In this study, a survey involving more than 4700 employees at a large construction firm in China was applied to establish a BN. BN-based analysis demonstrated that the safety climate factors may have a more significant influence on an employee's safety behavior than personal experience factors. A method to find a strategy by controlling one individual factor (or simple strategy) to improve safety behavior was then investigated. It was found that the simple strategy could be more effective when safety climate factors were properly controlled. In addition, a strategy via controlling multiple factors (or joint strategies) may even better improve the safety behavior. The analysis suggested that a joint control of both safety climate factors and personal experience factors worked most effectively. Finally, the prediction of human safety behavior under a specific climate was tested with the BN.
- 出版日期2008-12
- 单位CSIRO; 清华大学
