LAUSR

Purpose Safety climate (SC) is considered a leading indicator of safety performance, but scholars suggest that a common SC assessment framework is yet to be developed. Following the debate between the importance of facet analysis and agent analysis, the purpose of this paper is to test a factor structure, developed by the authors in previous work and arising from their systematic literature review, highlighting the role of safety agents in a construction site setting. Design/methodology/approach Multi-level SC surveys were conducted at five construction sites in Sydney, Australia, collecting data from of 352 workers associated with a mega-construction project. While examining the factor analysis of different studies, data reliability and data validity of the survey findings were ensured and a goodness-of-fit of SC model was examined through structural equation modelling. Findings The systematic literature review of Newaz et al. (2018) suggested a five-factor model of: management commitment, safety system, role of the supervisor, workers’ involvement and group SC. However, empirical data indicated that the questionnaire used to measure “safety system” failed to pass scale reliability; thus, a four-factor model was proposed to develop an agent-specific SC factor structure in the construction industry. Originality/value The four-factor model indicates the role and level of influence of different safety agents to improve safety perceptions on construction sites. The findings of this study will encourage researchers in construction safety to use the simplified four-factor SC (agent-specific) model presented and test it to further develop a common factor structure for the construction industry. The fact that the model is comprised of four factors makes further implementation somewhat easier in the development of safety plans, and when considering the role of safety agents, therefore enhancing its potential value.