The mediator role of safety motivation and knowledge between safety culture and safety performance: The results of a sociotechnical and macroergonomics approach

1 Introduction

Over the past three decades, researchers have paid considerable attention to safety culture due to its positive effect on improving safety outputs, such as injuries, deaths, and accident rates [1–3]. The concept of safety culture that emerged after the Chernobyl disaster is a subset of organizational culture, and implies how organizational management and human factors shape safety [4]. The improvement of safety culture creates an environment in which employees are aware of the risks and show a proactive response to them [5]. Therefore, safety culture is a managerial tool for controlling beliefs and behaviours regarding the safety of the workplace [6, 7].

Positive safety culture is important in improving employees’ trust in senior management decisions about safety and its ultimate goal is to enhance the organization’s safety performance. Safety performance refers to the criteria that reflect the overall performance of the management system to improve the organization’s safety [2, 8]. Each of these criteria has strengths and weaknesses and is measured based on the study objectives and available data. Generally, a criterion cannot be considered premier to another and the best criterion is chosen according to circumstances. For example, reactive measures are a retrospective approach and are used to compare the safety management effort at the end of a timeframe or project, while active indicators present the current status of safety management [9]. Reactive measures are more prevalent in early studies of safety assessment and are based on statistical data, such as the number of accidents or injuries. Nonetheless, alternative data such as the number of injuries collected by self-report questionnaires have become more common in the recent years [10, 11], indicating the greater credibility of these data [12, 13]. The number of injuries and accidents cannot be highly reliable due to the low frequency and incidence rate [8, 14]. Proactive measures are originated from behavioural models and many applications have been found in studies because of easy use and high confidence. The proactive measures of safety performance are defined as actions or behaviours that individuals display to improve their occupational health and safety [15]. Griffin and Neal introduced safety compliance and safety participation as the prominent proactive measures. Safety compliance refers to the essential behaviours that employees have to do to maintain the safety of the work environment. Indeed, safety participation refers to the behaviours that are conducted voluntarily and can indirectly improve safety in the workplace. However, these measures are important because of increased attention to safety (e.g., attending in safety courses) [16–19].

1.1 The relationship between safety culture and safety climate with safety performance

Numerous studies have examined the role of safety culture and safety climate in improving reactive safety measures [15, 20–24], proactive safety measures [25–28], or both [29, 30]. In these papers, meta-analysis and review studies have evaluated the results of empirical studies to better understand the findings [31–33]. In general, the trend towards proactive measures has increased in the recent years. Reviewing articles showed that the relationship between safety culture and safety performance criteria was evaluated independently or in the form of structural models. The review articles confirmed the relationship between safety culture and safety performance and indicated that safety culture factors were negatively and positively correlated to reactive and proactive measures, respectively. Hon, Chan [29] reported that the correlation coefficients between safety culture factors and immune function criteria were as follows: accidents rate (r = –0.25 to –0.11), safety compliance (r = 0.08 to 0.20), and safety participation (r = 0.26 to 0.48). Various approaches have been reported on the effect size of safety culture and proactive measures. Christian, Bradley [31] indicated that safety culture had a stronger relationship with safety partners than with safety acceptance. However, contradictory results were obtained by Hon, Chan [29]. Review studies have reported different amounts of explained variance between safety culture and safety performance. The results suggested that safety culture explained 5–15% of the variance of accident rate and 18–37% of the variance of employees’ safety behaviors [34]. Moreover, Liu, Huang exhibited a mediating approach for communicating safety, safety behaviors, and safety performance. The results demonstrated that safety climate predicted safety behavior, and safety behavior mediated the relationship between safety culture and the number of accidents [35]. Furthermore, Neal and Griffin proposed a chain approach to predict safety at the group level for safety motivation, which then affected safety behaviors. While safety behavior was a significant predictor of events, safety climate did not have this predictability. They argued that safety climate was a distal predictor of safety performance, while safety behavior predicted proximal safety performance [17].

1.2 The present study

According to review studies, the relationship be-tween safety culture and active safety measures has been confirmed. However, the effect paths are unknown and ambiguous. Attributes such as industry type, country, sample size, and statistical methods can be the reasons for this ambiguity. Another important reason is the extension of safety culture factors and the disagreement regarding the assessment factors. In fact, 3–22 factors and 11–300 items have been reported to assess safety culture [4, 33]. Griffin and Neal proposed a model based on job performance theories in which safety climate as the antecedent of safety performance, safety motivation and safety knowledge as mediators between safety climate and safety performance, and safety compliance and safety participation as the components of safety performance were introduced [19]. The present study was designed based on Neal and Griffins’ model to evaluate the effect of safety culture rather than safety climate on safety performance and to determine the role of safety motivation and safety knowledge as mediator variables. Although the concepts of safety culture and safety climate differ in definition and subjectivity, a common approach has been used in their evaluation. Researcher introduced safety climate as a snapshot of safety culture and highlighted that safety climate is the product of safety culture [36].

In this study, two main objectives were considered. The first objective of this study is to assess the relationship between safety culture and safety performance in which the safety culture assessment tool is developed based on organization’s sociotechnical characteristics related to an oil and gas industry. The second objective is to compare the direct and indirect paths in the presence of safety motivation and safety knowledge as mediators.

2 Methods

2.2.2 Safety performance

Eight items adopted from Neal and Griffin [17] were used to examine the components of safety performance, including safety compliance and safety participation. Safety compliance questions included: “I carry out my work in a safe manner ”, “I use all the necessary safety equipment to do my job ”, “I use the correct safety procedures for carrying out my job ”, and “I ensure the highest level of safety when I carry out my job ”. Safety participation questions included: “I promote the safety program within the organization”, “I put in extra effort to improve the safety of the workplace”, “I help my coworkers when they are working under risky or hazardous conditions”, and “I voluntarily carry out tasks or activities that help improve workplace safety” [18, 19]. The psychometric properties of the Persian version of Neil and Griffins’ Safety Performance Scale have been evaluated and its validity and reliability have been verified [38].

2.2.3 Safety motivation and safety knowledge

In the current study, safety motivation and safety knowledge were established as the determinants of safety performance as well as the mediator variables between safety culture and safety performance. Safety motivation refers to the willingness of employees to implement safety behaviors and the value of these behaviors. Safety knowledge refers to a certain extent to which employees are familiar with the safety regulations for working safely.

Eight items from Neal and Griffin [17] were selected to measure safety motivation and safety knowledge using a five-point Likert scale. The items were as follows: “I know how to perform my job in a safe manner”, “I know how to use safety equipment and standard work procedures”, “I know how to maintain or improve workplace health and safety”, “I know how to reduce the risk of accidents and incidents in the workplace”, “I believe that workplace health and safety is an important issue”, “I feel that it is worthwhile to put in effort to maintain or improve my personal safety”, “I feel that it is important to maintain safety at all times”, and “I believe that it is important to reduce the risk of accidents and incidents in the workplace”.

2.3 The research models

The direct and indirect paths between the study variables are presented in Fig. 1. Two indirect paths included the effects of safety motivation (P1 + P2) and safety knowledge (P3 + P4) as the mediator variables between safety culture and safety performance. In addition, P1 indicated the direct path between safety culture and safety performance. The effects of direct and indirect paths were examined using goodness of fit indices in three structural models. Fig. 2 (model 1) shows a direct path without the presence of indirect paths (mediator variables). Fig. 3 (model 2) shows a model with direct and indirect paths, which examines the effect of both paths simultaneously. Finally, Fig. 4 (model 3) shows a model with indirect paths.

OPEN IN VIEWER

Fig. 1

The research model and paths.

OPEN IN VIEWER

Fig. 2

The direct path between safety culture and safety performance (Model 1).

OPEN IN VIEWER

Fig. 3

Direct and indirect paths between safety culture and safety performance (Model 2).

OPEN IN VIEWER

Fig. 4

Indirect paths between safety culture and safety performance (Model 3).

Descriptive statistics and Structural Equation Modeling (SEM) were conducted using SPSS 23.0 (IBM Corp., Armonk, NY, USA) and AMOS 23.0, respectively. SEM was utilized to test the direct and indirect paths between safety culture and safety performance. SEM can examine distinctive but interdependent variables and quantifies these variables using multiple regression equations. A series of latent variables were used for investigating the observable variables that could not be measured directly. The associations between the variables and the role of the mediator variables were also explained using SEM.

In this study, safety culture, safety motivation, safety knowledge, and safety performance were considered as latent variables, which were measured using their factors and items as the observable variables. Goodness-of-fit indices were used to determine to what extent the model was consistent with the real data. Acceptable values for the indices were as follows: chi-square/degree of freedom (χ²/df) < 2 [39], Goodness of Fit Index (GFI) > 0.90, Adjusted Goodness of Fit Index (AGFI) > 0.90 [40], Comparative Fit Index (CFI) > 0.90, and Root Mean Square Error of Approximation (RMSEA) < 0.08 [41]. In addition to goodness of fit indices, loading factor was used to assess the relationship between the study variables. Factor loading is a kind of correlation coefficient that determines how much the latent variable explains the variance of the observable variables.

3 Results

As shown in Table 2, the highest and lowest coefficients were related to the correlation between work equipment and tools and safety motivation (r = 0.123) and between personal protective equipment and safety compliance (r = 0.701), respectively.

The hypothesis models for testing the direct effect of safety culture on safety performance are depicted in Figs. 1 and 2 . Fig. 1 presents the direct effect without the presence of the mediator variables. Fig. 2 presents the direct effect with the presence of the mediator variables. The standardized loading factors for the models are shown in Table 3, which indicated a significant positive relationship between safety culture and safety performance measures (P5). In all models, there were also acceptable values of the standardized loading factors between safety culture as the latent variable and its dimensions as the observable variables. As shown in Table 3, standardized loading factors of the three models were approximately equal, except for safety motivation ⟶ safety performance and safety knowledge ⟶ safety performance. This emphasized that the indirect paths in model 2 (P2 and P4) were not verified. The goodness of fit indices of the models are presented in Table 4. Accordingly, the indices confirmed all models.

The indirect effect of safety culture on safety performance through mediator variables (P1-P2 and P3-P4) is depicted in Figs. 3 and 4. In the presence of the direct path (Fig. 3), although safety culture was significantly related to safety motivation and safety knowledge, the indirect effects were not confirmed due to the non-significant relationship between safety motivation, and safety knowledge and safety performance. Based on Fig. 4, the indirect effects were confirmed in the absence of the direct path. Cumulative values of factor loadings for direct and indirect paths are shown in Table 5. The values indicated that the indirect paths had lower effects compared to the direct paths due to the presence of dual paths (due to the mediator variables).

The explained variances of the models are mentioned in Table 3. The results showed that the model without the direct path (Fig. 4) had the least variance. However, the highest value of variance for safety performance was explained in the model with the presence of mediator variables and the simultaneous effect of direct and indirect paths (Fig. 3). Accordingly, safety performance received the greatest impact through the direct path from safety culture, while mediator paths had the least effect on improving safety performance.

Considering the assumptions of SEM and the initial model of the study, safety culture and its dimensions as independent variables and safety performance (safety compliance and safety participation) as dependent variables of structural models were investigated with the presence and absence of mediator variables. The findings showed a significant and direct correlation between safety culture and safety performance. In the structural model, safety culture explained approximately 50% of the variance of safety performance. This suggested a robust link between the dimensions of safety culture to predict safety performance. This is a positive point for safety culture that predicted about 50% of safety performance and employees’ behaviors in the present study. Hon et al. showed that safety climate assessment using a 38-item questionnaire explained 43% of safety participation and 0.08% of safety compliance [29].

To reduce accidents in workplaces, safety culture and safety performance have been examined from a variety of perspectives. It has been emphasized that the examination of two technical and psychological views leads to promotion of a positive safety culture. Despite the key role that organizational culture plays in determining the success or failure of an organization, there is no consensus on how to properly describe the culture of an organization. In many studies, aspects of safety culture are often viewed from other characteristics of the organization, such as work plan, technology, work strategies, and separate economic decisions [42–44]. The current study findings indicated that safety performance was equally affected by safety participation and safety compliance. Christian and Bradley showed that safety culture had a stronger relationship with safety participation in comparison to safety compliance [31]. On the contrary, Hon and Chan reported that safety compliance had a stronger effect on safety performance [29]. This is due to the fact that safety compliance refers to the behaviors that employees are required to carry out at each level of their safety culture, and safety culture has a slight effect on this group of behaviors. In contrast, safety participation refers to voluntary behaviors. The level of voluntary behaviors to improve workplace safety is higher in an organization with a higher level of safety culture. The results of the present study showed equal correlation coefficients between the overall safety culture score and safety compliance and safety participation. The results of path analysis also showed that safety compliance and safety participation had approximately equal effects on safety culture. This can be due to individuals’ similar attitudes towards safety behaviors whether these behaviors are compulsory or voluntary.

The study findings in the model with the direct path indicated that although safety culture was significantly associated with safety knowledge and safety motivation, since safety performance was not significantly correlated to safety motivation and safety knowledge, these two paths of mediation were not confirmed. In fact, safety culture directly improved the individuals’ safety behaviors. Among the dimensions of the safety culture assessment tool, topics related to safety training and motivational activities were indirectly incorporated. Besides, two criteria of safety knowledge and safety motivation were assessed via general criteria. Accordingly, the safety culture tool predicted the employees’ safety behaviors with a more realistic approach.

The effectiveness of safety behaviors is important in terms of individual and organizational approaches. As discussed earlier, safety culture is a subset of organizational culture. Individual and organizational issues are simultaneously considered in the study of the effect of safety culture on employees’ attitudes and behaviors, while safety motivation and safety knowledge are individual topics. In this study, the mixed method with a macroergonomic approach provided a comprehensive view of the safety culture factors, so that the behavioral characteristics and commitment to safety and safety training domains adequately explained the employees’ motivation and knowledge. It was also emphasized that the use of credible tools had to be in accordance with the characteristics of the new community. Griffin and Neal and Vinodkumar and Bhasi examined the relationship between mediator variables (safety knowledge and safety motivation) and safety performance (safety compliance and safety participation) [19, 45]. In both studies, safety knowledge and safety motivation were considered as the determinants of safety performance, and safety compliance and safety participation as the components of safety performance. Indeed, the positive impact of safety culture on motivation and knowledge was confirmed. Safety motivation has been described as a person’s willingness to participate in safety and compliance with safe working practices. Safety knowledge also refers to an individual’s knowledge and skills regarding how to deal with safety regulations or to engage in safety activities to maintain safety. As many people have pointed out, safety awareness is not at the right level among workers. Besides, the content of the training classes does not meet the requirements of the job. However, it has been found that managers have a higher level of safety because they are responsible for the safety of their units.

It seems that the safety culture questionnaire items developed in this study had a more realistic approach to describing the employees’ motivation (punishment and reward in the workplace or higher priority to production relative to safety) and safety knowledge (individuals’ knowledge of safety regulations) for assessing safety knowledge and safety motivation. Safety knowledge and safety motivation are situated more intrinsically in the essence of the safety culture items. Hence, direct path is the main path in the impact of safety culture on safety performance. For example, encouragement and punishment or supervisors’ commands on more production are highly important to employees, which makes safety less important. Moreover, safety motivation and safety knowledge scores are quite high. However, the importance of the above factors causes motivation and knowledge not to have a significant effect on safety behaviors. Motivation is an intrinsic property for individuals, which leads to their desire for work. However, external requirements are also required to perform certain tasks, such as implementing safety regulations. Job pressure as an external requirement does not reduce motivation; rather, it changes the path to the person’s desire for a safe behavior. This was confirmed in the structural model [46]. Safety knowledge and safety motivation assessment tools were not appropriate because they were not focused on motivational factors in the target community.

The study had some limitations. First, the cross-sectional nature of the study design could not prove causal relationships. Second, only self-reported questionnaires without an observational confirmation or some kind of objective measurement used to consider the effects of safety culture on safety performance. Third, the study was performed only in an oil and gas refinery complex. Therefore, caution should be exercised when interpreting and generalizing the findings to other settings and organizations.

5 Conclusion

Nowadays, attention to safety culture in reducing accidents and improving proactive measures of safety performance has moved from theoretical debates to applications in nuclear energy, oil and gas, construction, and manufacturing industries. However, recognition of the appropriate factors to assess safety culture and determintion of the effective paths in improving safety outcomes are singnifican gaps in this field.

In the present study, a tool was used to evaluate safety culture that specifically addressed the organization’s sociotechnical characteristics of the industry. Moreover, the effective paths of safety culture in safety performance, which included direct and indirect pathways, were explored based on the studies performed by Toderi, Gaggia [18], Griffin and Neal [19], Bunner, Prem [46]. The results showed that the direct paths had a stronger effect due to the lack of interference of the mediator variables. Furthermore, development of the safety culture tool showed that most of the effective safety culture factors could be directly incorporated into the tools and have a greater impact on safety performance.