社会与工程心理学研究室知识库

员工不安全行为是工作场所超过三分之二安全事故的触发因素，安全氛围是影响员工安全行为的关键社会情境因素，探索安全氛围的改善对提高员工安全行为、预防企业安全事故具有重要意义。安全管理承诺是安全氛围的核心维度、对员工安全行为和安全结果预测力最强，也是安全氛围干预的首要目标。然而，鲜有研究探索这一知觉的来源并基于此设计和实施干预研究，这可能是现有研究干预效果不稳定的重要原因。此外，员工对组织不同层级管理者(主管和经理)安全管理承诺的知觉在不同水平上聚合，形成多水平安全氛围(团队安全氛围和组织安全氛围)。现有干预研究对多层级管理者同时干预但只验证其对某一水平安全氛围提升的作用，难以区分不同层级管理者对不同水平安全氛围的改善效应。更进一步，主管和经理日常大量互动过程中，经理关键安全管理行动通过主管知觉的组织安全氛围和主管关键安全管理行动对团队安全氛围的影响也值得关注。为了弥补现有研究的不足，本研究拟基于信号理论，探索能体现并传递主管和经理安全管理承诺的信号，即关键安全管理行动，并在此基础上探索主管和经理关键安全管理行动干预对团队安全氛围和组织安全氛围的提升及后续安全行为改善的效应。本研究将通过一个预研究和三个子研究分别探讨上述问题。

预研究通过行为事件访谈法，对某石油石化集团两个工厂67名员工进行访谈，探索主管和经理的关键安全管理行动。编码结果发现，主管和经理体现安全管理承诺的关键安全管理行动类型存在显著差异。具体而言，主管被提及的6个安全管理行动均为直接安全管理行动，其中“对安全行为认可和鼓励，对不安全行为提醒和纠正”和“表达对员工安全行为的期望”被提及的最多，是主管体现安全管理承诺的2个关键安全管理行动。与此不同，经理被提及的12个安全管理行动既包括直接安全管理行动也包括间接安全管理行动，其中“倾听、反馈安全意见和建议”、“拜访工作场所”、“关怀员工，对员工安全给予支持和政策指导”、“鼓励员工识别安全问题和风险”、“为安全投资必要的设备、物料等”以及“制定和优化安全制度和工作流程”分别在每个行动类型中被提及的最多，是经理体现安全管理承诺的6个关键安全管理行动。

研究一通过三阶段问卷调查法，探索主管和经理关键安全管理行动对团队及组织安全氛围的影响、经理关键安全管理行动通过主管知觉的组织安全氛围和主管关键安全管理行动的链式中介对团队安全氛围的影响。来自某建筑集团公司下属21个施工项目78名经理、90名主管和351名员工参与本研究。多水平路径分析的结果发现:主管和经理关键安全管理行动分别对团队和组织安全氛围有积极影响;经理关键安全管理行动通过主管知觉的组织安全氛围(一阶中介)和主管关键安全管理行动(二阶中介)对团队安全氛围有积极影响。

研究二通过实验组和对照组前后测现场准实验研究，探索主管关键安全管理行动干预对团队安全氛围和员工安全行为提升的效应。来自某石油石化集团3个工厂50个团队226名一线工人(实验组27个团队114人，对照组23个团队112人)参与本研究。实验组主管在实验4周内开展关键安全管理行动，而对照组主管不被要求开展这些行动。重复测量方差分析和多水平路径分析的结果发现:在实验结束后，与对照组相比，实验组主管开展关键安全管理行动的频率显著提高;主管关键安全管理行动干预可以提升团队安全氛围和员工安全行为;团队安全氛围的提升是实验干预和员工安全行为提升的中介。

研究三通过实验组和对照组前后测现场准实验研究，探索经理关键安全管理行动干预对组织安全氛围、团队安全氛围和员工安全行为提升的效应。来自某石油石化集团另外3个工厂30个生产单元108个团队379名一线工人(实验组10个工作单元39个团队129名员工，对照组20个生产单元69个团队250名员工)参与本研究。实验组经理在实验8周内开展关键安全管理行动，而对照组经理不被要求开展这些行动。重复测量方差分析和多水平路径分析的结果发现:在实验结束后，与对照组相比，实验组经理开展关键安全管理行动的频率显著提高;经理关键安全管理行动干预可以提升组织安全氛围、团队安全氛围和员工安全行为; 组织和团队安全氛围的提升是实验干预和员工安全行为提升的中介。

总体而言，本研究从安全氛围核心维度安全管理承诺的视角，基于信号理论设计并实施不同层级管理者关键安全管理行动的多水平安全氛围干预，为组织改善工作场所安全提供理论和实践价值。

More than two thirds of safety accidents in the workplace are the results of employee's unsafe behaviors. Safety climate is one of key social situational factors that influences how employees behave in terms of safety. Safety climate improvement, therefore, has become an important way to enhance employee's safety behavior and prevent safety negative events In fact, the primary target of safety climate promotion is to improve the employee's perceived management's commitment to safety, which is the core dimension of safety climate and the strongest predictor of safety performance among all aspects of safety climate. Few studies, however, have explored the source of this perception, and developed and implemented interventions based on it, which may be an important reason for the unstable intervention effects in current studies. In addition, employees perceive different levels of safety climate (group and organizational safety climate) through varied management's (supervisor's and manager's) commitment to safety. When multiple levels of management involve in the interventions at the same time, the intervention effects of these management on different levels of safety climate are not distinguished effectively. Furthermore, since there are a lot of daily interactions between supervisors and managers, the impact of managers' safety management actions on group safety climate through supervisors perceived organizational safety climate and supervisors' management actions is worth paying more attention. To fill the gap, by investigating the key safety management actions based on signaling theory, this study aims to explore the effect of supervisor's key safety management actions intervention on the enhancement of group safety climate and employee's safety behavior, as well as the effect of manager's key safety management actions intervention on the enhancement of organizational and group safety climate and employee's safety behavior. We explored the above issues through a preliminary study and three sub-studies respectively.

In the preliminary study, we used the behavioral event interview method to explore supervisor's and manager's key safety management actions. 67 management staff and front-line workers from a petrochemical enterprise were interviewed. The coding results showed that there were significant differences between the types of supervisor's and manager's key safety management actions. Specifically, 6 supervisors' direct safety management actions were mentioned in 3 action types. Among them, "recognition and encouragement to safe behaviors; reminding and correction to unsafe behaviors" and "expression of expected role behavior (safe behaviors)" were two key safety management actions to demonstrate supervisors' commitment to safety. On the other hand, 12 managers' direct and indirect safety management actions were mentioned in 6 action types. Among them, "listening to and feedback on safety opinions and suggestions", "visiting the workplace", "caring for employees, providing support to safety policies guidance for employee safety", "encourage employees to identify safety problems and risks", "investment in necessary equipment, materials, etc.", and "development and optimization of safety policies and procedures" were six key safety management actions to demonstrate managers' commitment to safety.

In the first study, we used three-stage questionnaire method to explore the impact of supervisor's and manager's key safety management actions on the group and organizational safety climate respectively, and the impact of the manager's key safety management actions on group safety climate through the mediation of supervisor perceived organizational safety climate and the supervisor's key safety management actions. 78 managers, 90 supervisors and 351 employees from 21 construction projects of a construction company participated this study. The results of multilevel path analysis showed that supervisors' and manager's key safety management actions had positive effects on the group and organizational safety climate respectively. The manager's key safety management actions had a positive impact on the group safety climate through a serial mediation of supervisor perceived organizational safety climate and supervisor's key safety management actions.

In the second study, we used a pre-post quasi-experiment to explore the effect of supervisor's key safety management actions intervention on group safety climate and employee's safety behavior. The study invited 226 front-line workers from 50 workgroups of 3 factories in a petroleum and petrochemical enterprise (114 workers from 27 workgroups in the experimental group and 112 workers from 23 workgroups in the control group) as subjects. The supervisors in the experimental group carried out key safety management actions during the experiment period of 4 weeks, while the supervisors in the control group were not required to carry out these management actions. The results of one-way repeated-measures multivariate analysis of variance and multilevel path analysis showed that, the frequency of key safety management actions carried out by the supervisors in the experimental group was significantly higher than control group at the end of the experiment; the intervention improved group safety climate and employee's safety behavior; the improvement of group safety climate was the mediator between the intervention and the enhancement of employee's safety behavior.

In the third study, we used a pre-post quasi-experiment to explore the effect of manager's key safety management actions intervention on organizational safety climate, group safety climate and employee's safety behavior. The study invited 379 front-line workers from 108 workgroups in 30 work units of other 3 factories in a petroleum and petrochemical enterprise (129 workers from 39 workgroups in 10 work units in the experimental group and 250 workers from 69 workgroups in 20 work units in the control group) as subjects. The managers of the experimental group carried out key safety management actions during the 8 weeks of the experiment, while control group managers were not required to carry out these management actions. The results of one-way repeated-measures multivariate analysis of variance and multilevel path analysis showed that the frequency of key safety management actions carried out by the managers in the experimental group was significantly higher than control group by the end of the experiment; the intervention improved the organizational and group safety climate, and employee's safety behavior simultaneously. The improvement of organizational safety climate and group safety climate were the mediators between the intervention and the enhancement of employee's safety behavior.

In general, from the perspective of management's commitment to safety, this study develops and implements safety climate interventions based on supervisor's and manager's key safety management actions, providing favorable theoretical basis and practical value for organizations to improve workplace safety.