Reveal the safety behaviors of employees in the manufacturing sector through safety climate factors

Abstract

Background

One of the factors that causes industrial accidents is the unsafe behavior of people during work. When the importance of safety climate factors, which are thought to affect the safety behaviors of employees directly, is understood in terms of industrial accidents, it is obvious that the number of industrial accidents and related deaths will decrease. Understanding the factors that influence safety behavior is crucial for improving safety.

Objective

This study used structural equation modeling to determine the effects of safety climate factors on safety behavior factors.

Methods

401 sector employees in 61 enterprises were obtained through a survey conducted in the manufacturing sectors of Düzce Province.

Results

The study results indicated a positive relationship between safety information, safety awareness, and safety compliance while showing negative relationships between safety attitudes and safety compliance. Additionally, positive relationships were found between safety information, safety training, and safety participation; however, no significant effect was observed between safety awareness, safety attitude, and safety participation.

Conclusion

The study results guide how business managers should improve employees’ safety-related behaviors by focusing on safety climate factors.

Keywords

safety behaviors safety climate safety training safety participation safety compliance structural equation modeling

Introduction

Industrial accidents and diseases are major safety and health concerns in the manufacturing industry. In addition, unsafe behavior of employees is an important cause of work-related injuries and deaths. According to the Social Security Institution (SSI) data for 2022; 588.823 industrial accidents have occurred in Turkey, and 1.517 (0.26%) of these accidents resulted in death. In the same year, the number of people who had contracted occupational diseases was 953, while the number of deaths was 8 (0.84%). Workers between the ages of 20 and 35 years are most affected by industrial accidents.¹

According to 2022 data from the SSI, there were 258.570 industrial accidents in the manufacturing sector in Turkey, and 312 (0.12%) of these accidents resulted in death. It has also been reported that 507 people have occupational diseases. In the manufacturing industry sectors of Düzce Province, 4.573 industrial accidents occurred, and nine (0.20%) of these accidents resulted in death. It has also been reported that 17 patients had occupational diseases. The results obtained for the manufacturing sector of Düzce Province regarding industrial accidents constitute approximately 1.77% of the cases in the manufacturing sector in Turkey.¹

When the data for the last five years (2018–2022) were analyzed with the growth of the manufacturing sector in Düzce Province, it was determined that there has been an 82.70% increase in the number of industrial accidents in industrial enterprises and a 28.57% increase in the number of deaths due to these industrial accidents.¹ This shows that improving working conditions, providing employees with personal protective equipment suitable for their work, and providing employees with the necessary job-related training are key factors in reducing their risk-taking behaviors.²

Safety climate

The concept of safety climate has emerged as a result of studies on organizational culture and climate.³ Safety climate is the sum of employees’ shared perceptions of safety-related policies, procedures, and practices in their work environments.⁴ Although the constructs used to assess safety climate vary from study to study, the areas measured generally include safety awareness, safety attitude, safety training, safety policy, safety information, safety communication, and coworker support.^5,6

A safety climate can be defined as the assessment of safety criteria in the work environment.⁷ The safety climate is a useful tool for understanding employees’ and managers’ behavior regarding safety issues, perceptions of safety, the importance of safety in the work environment, and the participation of colleagues in safety. In addition to being necessary for safety performance, the safety climate has a significant impact on improving safety performance.^8,9

It has been confirmed that safety climate significantly influences safety behaviors in high-risk manufacturing sectors.¹⁰ When management prioritizes safety in the workplace environment, employees generally prefer safety behaviors, such as safety compliance and safety participation, when factors such as safety training, safety information, and safety communication are provided.¹¹ When safety is not prioritized in the workplace environment, employees exhibit negative attitudes and their motivation to do their work safely decreases. This situation may cause employees to engage in unsafe behaviors.¹² It has also been found that a safety climate can have profound effects on employees’ mental states and behaviors, leading to actual safety practices.¹³

A positive safety climate is an important part of a safe working environment. Many past studies have examined the benefits of a positive safety climate in an organization. Hofmann and Stetzer¹⁴; Hofmann and Stetzer¹⁵ examined the effects of safety climate using employees in a large chemical processing plant and a large utility company. The results showed that a positive safety climate is associated with a reduction in unsafe behavior among employees. Griffin and Neal¹⁶; Zacharatos et al.¹⁷ reported that a favorable safety climate is associated with safety compliance and motivation to stay safe in large manufacturing companies. Ansori et al.¹⁸ found that the safety climate is influenced by safety information, motivation, safety compliance, and safety participation. In their study in the mining sector, Yu and Li¹⁹ examined the relationship between psychosocial safety climates and unsafe behaviors. They found that a safety climate could reduce miners’ unsafe behaviors through the mediating roles of job stress and fatigue.

Safety training is considered the most important indicator of safety motivation for predicting safety behaviors under safety climate factors.²⁰ Christoffel and Gallagher²¹ defined safety training as the process of providing information to employees about occupational safety, developing desired attitudes towards hazards and risks, and reducing occupational accidents by encouraging safe behavior in the work environment. Simon and Piquard²² suggest that safety training improves workers’ information, attitudes, and behaviors through hazard anticipation training. Vredenburgh²³ emphasizes that increased perceived risk levels increase compliance with safety-related instructions, making training necessary for workers to recognize hazards in their work environment.

Ahn and Park²⁴ found a positive relationship between employees’ information of work procedures and general occupational safety, their attitudes towards safety, and their tendency to participate in safety-related activities. In this context, increased safety information paves the way for increased safety awareness and reduced occupational accidents.

Safety behavior

Safety behavior is defined as employees’ self-protective actions to avoid existing hazards in the work environment and the risks arising from these hazards.²⁵ Moreover, safety behavior includes participation in safety activities, helping coworkers, voicing safety concerns, and following safety rules.²⁶ Safety behavior traditionally consists of safety participation and compliance.²⁷ First, safety compliance refers to the appropriateness of the basic safety-related activities necessary to keep the workplace or work environment safe.^16,28,29 Safety compliance includes the behavior of following safety instructions in the workplace, such as the rule to use personal protective equipment. The behavior of complying with safety instructions becomes habitual and ensures that the employee starts working by taking the necessary precautions for his/her job before each work. Employees’ personal competence also affects their safety, which is critical for safety compliance.³⁰ Second, it refers to the voluntary behaviors of individuals that contribute to the development of a safety-supportive environment, rather than the direct provision of individual safety, such as helping coworkers.³¹

In manufacturing sector enterprises where safety factors are prioritized, it is obvious that there will be fewer occupational accidents and injuries with the increase in safety training, safety information, attitudes towards safety, and safety awareness of employees. Similarly, Okur³² emphasized that the factors leading to unsafe behaviors and situations should be eliminated in order to prevent workplace injuries. Karadal³³ and Damayanti et al.³⁴ stated that safety climate and safety culture affect safe behaviors and thus prevent work accidents and workplace injuries. Therefore, this study aims to determine the effects of safety climate factors on employee safety compliance and participation in the manufacturing sector. For this purpose, the impact of safety climate factors on safety compliance and safety participation was investigated through SEM.

In this study, similar previous scientific studies were reviewed to derive the safety climate factors affecting the safety behaviors of employees. Additionally, based on the literature review, an original research model and hypotheses were created for employees working in the manufacturing sector in Düzce Province. The validity of the hypotheses was determined based on the validation of the research hypotheses using the questionnaires obtained and the results of the analysis. The academic and practical implications of the research results are also discussed. Through this process, this study will provide industry managers with the necessary information to improve employees’ safety-related behaviors by explaining the relationship between safety climate and safety behavior factors.

Conceptual model

Within the scope of this study, models developed by Liang et al.³⁵; Shin et al.³⁶; Mohammadfam et al.³⁷ were adopted, and the research model shown in Figure 1 was created. The research model shows that safety climate factors (safety information, safety awareness, safety training, and safety attitude) can affect safety behavior factors (safety compliance and safety participation). With this research model, determining the safety climate factors that are thought to positively affect the safety behaviors of individuals working in manufacturing sectors and making the necessary improvements will provide a proactive contribution for future studies.

Figure 1.

Research model of the study.

This research model can be shown as a strong conceptual model to determine the hypothesized interactions between the safety climate factors and safety behavior factors of employees in all manufacturing sectors (Figure 1).

The hypothesized relationships between the factors specified in the research model are as follows;

H1; Safety information (SI) significantly affects employees’ safety compliance (SC).

H2; Safety information (SI) significantly affects employees’ safety participation (SP).

H3; Safety awareness (SA) significantly affects employees’ safety compliance (SC).

H4; Safety awareness (SA) significantly affects employees’ safety participation (SP).

H5; Safety training (ST) significantly affects employees’ safety compliance (SC).

H6; Safety training (ST) significantly affects employees’ safety participation (SP).

H7; Safety attitude (SAT) significantly affects employees’ safety compliance (SC).

H8; Safety attitude (SAT) significantly affects employees’ safety participation (SP).

Material and methods

Population and sample

This study, which aims to reveal the factors affecting the safety behaviors of employees in enterprises, was conducted on enterprises in the manufacturing sector operating in the central province of Düzce and its districts (Figure 2).

Figure 2.

Geographical distribution of Düzce (Central) Province and its districts.

Within the scope of this study, 157 different companies with 10 or more employees and 14.776 employees operating in these companies constituted the study population. The survey was conducted between January and March 2024. To identify the enterprises to be surveyed, the aim was to reach the entire main mass without sampling. The study was primarily planned according to a face-to-face survey and an on-site observation method. As a result of the survey application, 401 employees from 61 different enterprises were interviewed face-to-face. As Child³⁸ suggests, a sample size may be considered adequate if it is not less than five times the number of items in the survey. Thus, the sample size 401, which is more than five times the 29 items in the survey, was deemed sufficient for this study. Other authors have stated that the sample size should be at least five or ten times the number of variables in the model.³⁹ Jackson⁴⁰ stated that higher values of the parameter ratio per observation have a positive effect for some fit measures.

Data collection instruments

Within the scope of this study, a questionnaire with three different parts was created by examining studies in the literature on similar issues.^{2,12,16,35,36,41} The first part of the questionnaire included demographic characteristics (13 items), and the other parts are shown in Table 1.

Table 1.

Factors and subfactors for employees.

Factors	Subfactors
Safety climate factors	Safety information (5 items)
	Safety awareness (5 items)
	Safety training (5 items)
	Safety attitude (4 items)
Safety behavior factors	Safety compliance (5 items)
Safety behavior factors	Safety participation (5 items)

In the survey form, except for the demographic information, the sections were questioned using a 5-point likert scale (from 1 = strongly disagree to 5 = strongly agree). The data were analyzed using descriptive analyses, confirmatory factor analysis (CFA), and structural equation modelling (SEM) to assess the goodness of fit of the research model and test the hypotheses. CFA and structural model evaluations were performed using AMOS version 16.0 (Analysis of Moment Structures) software, and other analyses were performed using SPSS version 22.0 (Statistical Software for Social Sciences). In addition, a one-way analysis of variance (ANOVA) test was used to statistically evaluate whether the safety climate and safety behavior factors for the participating employees differed according to the company's field of operation.

Kaiser Meyer Olkin (KMO) and Bartlett values are expected to be significant in testing the construct validity of the data set used. While the KMO test determines whether the data are suitable for factor analysis, the Bartlett test of sphericity is used to test whether the variables are correlated. The KMO values vary between 0 and 1. If the KMO test results are less than 0.50, they cannot be accepted, 0.50–0.60 is bad, 0.61–0.70 is poor, 0.71–0.80 is moderate, 0.81–0.90 is good, and a value above 0.90 is excellent.⁴²

Reliability analysis measures the internal consistency between judgments in a measurement tool and provides information about the relationships between these judgments.⁴³ Cronbach's α was examined to determine the reliability of the dataset used. Cronbach's α varied between 0 and 1. It is stated that if the criterion values for Cronbach's α value are less than 0.40, the scale is not reliable, the value between 0.41–0.60 is low, the value between 0.61–0.80 is medium, and the value between 0.81–1.00 is high reliability.⁴⁴

Model testing and fit indices

Evaluating whether a particular research model fits the data is one of the most important steps in structural equation modelling.⁴⁵ In structural equation modelling, the model was tested using fit indices. Fit tests are generally based on comparing the covariance matrix of a theoretically specified model with the sample covariance matrix. A small difference between these two matrices indicates that the data fits the theoretical model.³⁹ If the fit indices obtained by testing the model show that there is a fit between the model and data, the structurally generated hypotheses are accepted.⁴⁶

In the evaluation of the fit indices, the meanings and basic properties of the indices are as follows: The chi-squared value (x²) is the traditional measure used to evaluate the overall model fit. It evaluates the magnitude of the discrepancy between the model and applied covariance matrices.⁴⁷

The goodness-of-fit statistic (GFI) and adjusted goodness-of-fit statistic (AGFI) show the extent to which the model measures the variance covariance matrix in the sample and is also considered as the sample variance explained by the model. AGFI is the GFI corrected for the sample size. GFI and AGFI values of 0.85 and above indicate acceptable fit, and 0.90 and above indicate good fit.⁴⁸

The comparative fit index (CFI) is a revised form of the normed fit index (NFI) that considers the sample size and performs well, even when the sample size is small.⁴⁸ It is one of the most widely used fit indices because it is one of the measures least affected by sample size.⁴⁹ A CFI value of 0.95 and above is an indicator of good fit.⁴⁷

A Tucker-Lewis index (TLI); when the degrees of freedom of the model are added to the normed fit index (NFI), the Tucker-Lewis index is obtained. TLI is also referred to as the Non-Normed Fit Index (NNFI) in the literature.³⁹ TLI value of 0.95 and above is an indicator of good fit.⁴⁷

Root mean square error of approximation (RMSEA): RMSEA should not be preferred in models with small samples (sample size less than 250) because it is excessively affected by sample size. This causes rejection of the model that should be accepted in reality. An RMSEA value of 0.08 or less indicates acceptable fit.³⁹

The standardized root mean square residual (SRMR) is the standardized difference between the observed and estimated covariances. An SRMR value close to zero indicated a perfect fit. Values of 0.08 and smaller indicate acceptable fit.^39,47 The parsimony normed fit index (PNFI) is based on the NFI adjusted for the loss of degrees of freedom.⁵⁰ The PNFI value varies between 0 and 1, with higher values indicating a more parsimonious fit.⁵¹

Results

Demographic information

The demographic information of the employees is shown in Table 2.

Table 2.

Demographic information of employees.

Demographic Information		Frequency	%	Demographic Information		Frequency	%
Age (years)	<20	4	1.0	Work experience in the same company (years)	<6	186	46.4
	20–30	107	26.7		6–10	112	27.9
	31–40	156	38.9		11–15	71	17.7
	41–50	111	27.7		16–20	21	5.2
	51–60	20	5.0		21–25	4	1.0
	>60	3	0.7		>25	7	1.8
	Total	401	100		Total	401	100
Gender	Female	60	15.0	Salary (TRY)	<14,000	8	2.0
	Male	341	85.0		14,000– 16,000	158	39.4
	Total	401	100		16,001– 18,000	194	48.4
Marital status	Married	295	73.6		18,001– 20,000	32	8.0
	Single	106	26.4		>20,000	9	2.2
	Total	401	100		Total	401	100
Educational status	Primary education	206	51.4	The field of operation of the company	Furniture	167	41.6
	High school	167	41.6		Textile	85	21.2
	Asst. degree	19	4.7		Machinery	79	19.7
	Undergraduate	8	2.0		Plastic	41	10.2
	Postgraduate	1	0.3		Construction	29	7.3
	Total	401	100		Total	401	100
Industrial accident experience	Yes	42	10.5	Working order	Shift worker	104	25.9
	No	359	89.5		Non-Shift worker	297	74.1
	Total	401	100		Total	401	100

Note: TRY: Turkish lira.

Table 2 shows that 38.9% of the employees were between the ages of 31–40. Most of the employees were male (85.0%), and 73.6% were married. The educational level of more than half of the employees (51.4%) ranged from primary to postgraduate education (0.3%). Of the respondents, 53.6% had worked for six or more years in the same company. When the monthly income of the employees was analyzed, it was observed that 85.4% of them earn between 16,001–18,000 TL. It can be seen that 41.6% of the companies in which the participants work operate in the furniture sector, followed by the textile sector at 21.2%.

Within the scope of this study, 89.5% of the employees had not experienced any industrial accidents in the past. More than half of the employees (74.1%) had shiftless working patterns.

A statistically significant difference was found between the sectors in which the participants operated and safety information and safety training (p < 0.05); however, no statistically significant difference was found between safety awareness, safety attitude, safety participation, and safety compliance (Table 3).

Table 3.

Comparison of safety climate factors and safety behavior factors for participants in terms of companies’ field of operation.

Factors	Companies	n	%	x̄*	σ	HG	p
Safety information	Furniture	167	41.6	3.82	1.093	a	0.004
	Textile	85	21.2	3.30	0.977	b
	Machinery	79	19.7	3.64	1.042	ab
	Plastic	41	10.2	3.62	1.132	ab
	Construction	29	7.3	3.96	1.051	a
Safety training	Furniture	167	41.6	4.21	1.021	a	0.012
	Textile	85	21.2	3.73	1.025	b
	Machinery	79	19.7	3.99	1.159	ab
	Plastic	41	10.2	3.82	0.934	ab
	Construction	29	7.3	4.03	0.913	ab
Safety attitude	Furniture	167	41.6	3.74	1.006	a	0.864
	Textile	85	21.2	3.67	0.995	a
	Machinery	79	19.7	3.78	1.054	a
	Plastic	41	10.2	3.66	0.936	a
	Construction	29	7.3	3.44	0.827	a
Safety awareness	Furniture	167	41.6	2.51	1.036	a	0.558
	Textile	85	21.2	2.41	1.049	a
	Machinery	79	19.7	2.52	1.055	a
	Plastic	41	10.2	2.42	1.127	a
	Construction	29	7.3	2.34	0.780	a
Safety compliance	Furniture	167	41.6	3.64	0.604	a	0.476
	Textile	85	21.2	3.65	0.624	a
	Machinery	79	19.7	3.56	0.803	a
	Plastic	41	10.2	3.76	0.682	a
	Construction	29	7.3	3.76	0.648	a
Safety participation	Furniture	167	41.6	3.70	1.090	a	0.375
	Textile	85	21.2	3.67	1.055	a
	Machinery	79	19.7	3.74	1.180	a
	Plastic	41	10.2	3.62	1.025	a
	Construction	29	7.3	3.50	1.112	a

from 1: strongly disagree to 5: strongly agree.

Note: n: sample size; x̄: mean; σ: standard deviation; HG: homogeneous group; p: significance value.

Reliability and validity analysis

The overall reliability was 0.828. The reliability of the subscales ranged from 0.767 to 0.955. The Kaiser-Mayer-Olkin (KMO) value for the validity of the scale has been determined to be 0.842. These results demonstrate that the scale has high reliability and does not pose an obstacle in terms of validity.^44,52 The results of the exploratory factor analysis (EFA) of the scale are shown in Table 4.

Table 4.

Validity and reliability analysis results.

Factors	Reliability	Validity
Factors	Cronbach's α	KMO	Barlett
Safety Information (5 items)	0.944	0.842	18.460.088
Safety Awareness (5 items)	0.775
Safety Training (5 items)	0.955
Safety Attitude (4 items)	0.784
Safety Compliance (5 items)	0.767
Safety Participation (5 items)	0.938
All Scale (29 items)	0.828

Note: KMO: Kaiser–Mayer–Olkin.

The skewness values of the parameter ranged from (−1.469) to (0.984) and kurtosis values ranged from (−1.347) to (+1.778). With these values, the data showed normal distribution.⁵³ As a result of the EFA, the analysis was completed by removing seven statements from the analysis (Table 5).

Table 5.

The items extracted from the scale.

No	Code	Items	Mean*
1	SC4	I ensure my safety during work	4.05
2	SA5	I pay attention to the use of personal protective equipment	3.81
3	SI2	I know how to use occupational health and safety equipment	3.69
4	SC5	I do not believe that a work can be done safely without following safety procedures	3.58
5	SP1	I support the safety program	3.57
6	SA1	I stop working when I realize that it is not safe to continue my work	3.45
7	SP3	I help my colleagues work under risky conditions	3.36

from 1: strongly disagree to 5: strongly agree.

Note: SA: safety awareness; SC: safety compliance; SI: safety information; SP: safety participation.

The items listed in Table 5 were extracted from the analysis because they did not provide significant results for any virtual factor, decreased factor reliability, or provided significant results for more than one virtual factor at the same time.

After these seven statements were removed from the analysis, a structure with six virtual factors was obtained, with the remaining 22 statements (Table 6).

Table 6.

Results of factor validity analysis.

No	Code	Variable	Factor Load	Announced Variance (%)	Cronbach's α
	Safety Training (ST)			19.828	0.955
1	ST1	Thanks to the training provided to me, I recognized hazards better	0.912
2	ST4	My company provided me with comprehensive training on safety issues.	0.890
3	ST3	Safety issues were prioritized in the training programs	0.889
4	ST2	My supervisor encourages participation in the training programs	0.886
5	ST5	I have sufficient training in safety issues in my workplace	0.875
	Safety Information (SI)			14.826	0.920
6	SI3	I know how to do my work safely	0.903
7	SI4	I know how to prevent potential hazards in my workplace	0.897
8	SI5	I know about the current hazards related to my job and how to take precautions	0.895
9	SI1	I have a lot of information of occupational safety rules	0.887
	Safety Awareness (SA)			12.346	0.966
10	SA2	When I have doubts about occupational safety issues, I act with caution	0.926
11	SA3	I do not work with unsafe machines and tools	0.922
12	SA4	I work with periodically maintained equipment	0.853
	Safety Participation (SP)			11.354	0.894
13	SP4	I willingly undertook the tasks given to me to develop my work safety	0.934
14	SP2	I suggest ideas to develop work safety	0.900
15	SP5	I helped my colleagues with occupational safety issues	0.867
	Safety Attitude (SAT)			11.192	0.784
16	SAT4	Management prioritizes productivity over safety	0.848
17	SAT3	I ignore occupational health and safety issues while working at my workplace	0.794
18	SAT1	People have accidents at work because they are unlucky	0.792
19	SAT2	I think it is a good idea to take risks to complete what I do at my workplace	0.664
	Safety Compliance (SC)			11.153	0.867
20	SC2	I carefully used all the safety equipment and tools during work	0.859
21	SC3	I adhere to current safety rules during work	0.849
22	SC1	I conduct my work safely	0.848
	Announced Total Variances				80.699
	Cronbach's α Value				0.800
	KMO Value				0.821
	Bartlett Value				0.678
	df				231
	sig.				0.000

Note: df: degree of freedom; KMO: Kaiser-Mayer-Olkin; sig: significance.

As a result of the EFA, the overall Cronbach's α value for the scale was 0.800. The Cronbach's α for the subscales varied between 0.784 and 0.966. In addition, the KMO sampling adequacy measure used to test the structural validity of the scale was 0.821. Within the six virtual factor structures, the first virtual factor alone explained 19.828% of the total explanatory power, while the total explanatory power was 80.699%.

Measurement model

The standard factor loadings, average variance explained (AVE), and composite reliability (CR) values of the variables are shown in Table 7.

Table 7.

Validity test results of the measurement model.

Latent Parameter		Code	Standard Factor Loadings (Estimate)	AVE ≥ 0.500	CR ≥ 0.700
Safety Climate Factors (SCF)	Safety Training (ST)	ST1	0.930	0.799	0.952
		ST3	0.913
		ST4	0.885
		ST5	0.880
		ST2	0.860
	Safety Information (SI)	SI5	0.905	0.723	0.912
		SI4	0.884
		SI3	0.820
		SI1	0.786
	Safety Attitude (SAT)	SAT4	0.830	0.502	0.799
		SAT3	0.713
		SAT1	0.679
		SAT2	0.590
	Safety Awareness (SA)	SA3	0.985	0.906	0.966
		SA2	0.970
		SA4	0.898
Safety Behavior Factors (SBF)	Safety Compliance (SC)	SC3	0.916	0.686	0.867
		SC2	0.810
		SC1	0.751
	Safety Participation (SP)	SP4	0.967	0.752	0.900
		SP2	0.871
		SP5	0.749

Note: AVE: average variance extracted; CR: composite reliability.

In the measurement model, the AVE coefficient should be evaluated for each latent parameter and the CR coefficient should be evaluated to test the loadings for each indicator. Convergent validity, assessed by AVE and CR values, shows whether the items are related to each other and whether they are in the same measurement.⁵⁴

As shown in Table 7, the standard factor loadings of all the measures were between 0.590 and 0.985. To ensure convergent validity, all CR values were expected to be greater than the AVE value (CR > AVE), and the AVE value was expected to be greater than 0.500.⁵⁵ Based on the data obtained, AVE values varied between 0.502 and 0.906, while CR values varied between 0.799–0.966 and these values were above acceptable levels. Therefore, convergent validity was ensured.

Table 8, demonstrates the correlation coefficient between the research variables. The correlation matrix between the variables ranges from −0.195 to 0.508. According to these results, the measurement model had discriminant validity.

Table 8.

Correlation matrix analysis between the research variables.

Variable	SI	ST	SAT	SA	SP	SC
SI	1.000
ST	0.033	1.000
SAT	−0.045	−0.127	1.000
SA	0.116	0.508	−0.122	1.000
SP	0.108	0.203	−0.093	0.124	1.000
SC	0.160	0.397	−0.195	0.423	0.159	1.000

Structural model

The goodness-of-fit values of the confirmatory factor analysis test results are listed in Table 9.

Table 9.

Goodness-of-fit values for CFA.

Statistic	Default Model Fit Value	Recommended Fit Value	Model Fit
x²	501.880	—	—
df	198	—	—
p-value	0.000	—	—
x²/df	2.535	0 ≤ x²/df ≤ 3	Good fit
GFI	0.896	0.850 ≤ GFI ≤ 0.900	Acceptable fit
AGFI	0.868	0.850 ≤ AGFI ≤ 0.900	Acceptable fit
CFI	0.959	0.950 ≤ CFI ≤ 1.000	Good fit
TLI	0.952	0.950 ≤ TLI ≤ 1.000	Good fit
RMSEA	0.062	0.050 ≤ RMSEA ≤ 0.080	Acceptable fit
SRMR	0.064	0.050 ≤ SRMR ≤ 0.080	Acceptable fit
PNFI	0.800	0.800 ≤ PNFI ≤ 0.950	Acceptable fit

Note: AGFI: adjusted goodness-of-fit index; CFI: comparative fit index; GFI: goodness-of-fit index; PNFI: parsimony normed fit index; RMSEA: root mean square error of approximation; SRMR: standardised root mean square residual; TLI: Tucker–Lewis index; x²/df: chi square/degrees of freedom.

When the goodness-of-fit values in Table 9 were compared with the reference fit values, it was observed that the structural model was within the acceptable limits.^39,56–58 A structural model of the scale is shown in Figure 3.

Figure 3.

Structural model of scale.

In Figure 4, the relationships and path coefficients between safety climate factors and safety behavior factors are depicted. A p-value of less than 0.05 for the path coefficient indicates that the hypothesis is supported and statistically significant.⁵⁹

Figure 4.

Hypothesis results of model (*p < 0.05; **p < 0.01).

In the study examining the effects of safety climate factors on safety behavior factors, six out of eight hypotheses were accepted, while two were rejected. First, it was determined that safety information positively influenced both safety compliance (estimate = 0.093; p < 0.05) and safety participation (estimate = 0.114; p < 0.05). Thus, hypotheses H1 and H2 are fully supported.

Second, safety awareness has a positive effect on safety compliance (estimate = 0.220; p < 0.01). However, safety awareness did not have a significant effect on safety participation (estimate = 0.013; p > 0.05). Therefore, while H3 is fully supported, H4 is not because it does not have a significant relationship.

Third, safety training had a significant effect on both safety compliance (estimate = 0.213; p < 0.01) and safety participation (estimate = 0.238; p < 0.01). Thus, Hypotheses H5 and H6 are fully supported.

Finally, safety attitude was found to have a negative effect on safety compliance (estimate = −0.110; p < 0.05). However, safety attitude did not have a significant effect on safety participation (estimate = −0.081; p > 0.05). Therefore, while hypothesis H7 was fully supported, hypothesis H8 was not because it did not have a significant relationship (Figure 4).

Discussion

The research model used in this study was compatible with the collected data, and statistically significant relationships were identified, except for two variables (safety awareness - safety participation and safety attitude - safety participation).

This study contributes to the relevant literature by considering safety climate factors (safety information, safety awareness, safety training, and safety attitude) to explain employees’ safety behavior factors.

Safety information

In this study, safety information was found to have a significant effect on safety participation and compliance. This means that the safety behaviors of workers working in manufacturing sectors can be changed by providing the necessary information about occupational safety rules and procedures, and by providing information about the existing hazards that may arise in the workplace environment and the precautions to be taken.

In similar studies, it has been stated that safety information has the strongest direct effect on safety behavior factors and safety behaviors will be strengthened by increasing the safety information of employees.^12,16,36,41 Ansori et al.¹⁸ determined that safety information and motivation mediate the relationship between the safety climate and safety behavior. In addition, they stated that safety information only affects safety compliance.

It was determined that there was a statistically significant difference between the safety information of the participants depending on the field of operation of the companies (p < 0.05). Accordingly, it was determined that the participants working in the furniture and construction sectors had the highest level of safety information, those working in the machinery and plastics sectors had a medium level, and those working in the textile sector had the lowest level. This result may be due to the fact that employees of different sectors have different levels of safety information according to their fields of operation and the variety of specific risks and safety measures in each sector. Because there are more physical risks and the possibility of occupational accidents in the construction and furniture sectors, employees in these sectors may be more trained in safety.

Safety awareness

Safety awareness has a direct positive effect on safety compliance. This result supports the literature.^35,36,41 However, safety awareness does not have a significant effect on safety participation. This suggests that increased safety awareness improves compliance with safety rules and work procedures, and excludes participation or participation.

To increase employees’ safety awareness, necessary training and seminars on occupational safety, information on how to use machinery and equipment, and a suitable environment in which employees can easily discuss occupational safety issues can be created. Thus, the safety behavior of sector employees can be improved.

There was no statistically significant difference in the security awareness of the participants depending on the field of operation of the companies (p > 0.05).

Safety training

Safety training has a positive effect on safety participation and compliance. It can be said that employees in the sector, thanks to the safety training they receive, are better able to recognize hazards, thus exhibiting positive behaviors related to safety. These findings support those in the existing literature.^7,16,60

Burke et al.⁶¹ stated that safety training designed for employees increases their safety information, reduces unsafe behaviors, and thus reduces industrial accidents that may occur.

Lim et al.⁶² stated that improving employees’ safety information through safety training has a positive effect on safety behavior. Therefore, companies should develop systematic and effective safety training programs to improve their employees’ safety information.

A statistically significant difference was found between the participants’ safety training depending on the field of operation of the company (p < 0.05). Accordingly, it was determined that the participants working in the furniture sector had the highest level of safety training, those working in the machinery, plastics, and construction sectors had the medium level, and those working in the textile sector had the lowest level.

Safety attitude

The positive effect of safety attitudes on safety behavior has been reported in many scientific studies.⁶³ It has been emphasized that promoting safety attitudes, safety information, motivation, and reducing job stress will lead to higher rates of safe behaviors among employees.^37,64

Within the scope of this study, it was determined that although safety attitude has a direct and negative effect on safety compliance, it does not have a significant effect on safety participation. However, this result does not support the findings of previous studies. An important reason for this is the differences (social and cultural) between the various cities where the scales were applied and the city of Düzce, in terms of employees. Zohar⁴ identified eight safety climate factors to measure the safety climate of production workers in 20 Israeli companies. However, Brown and Holmes⁶⁵ used the same questionnaire with an American sample of production workers and found only three safety climate factors, attributing this difference to cultural factors. McDonald and Ryan⁶⁶ argued that the factors that influence the safety climate in one industry may not be valid in other industries.

There was no statistically significant difference between the security attitudes of the participants depending on the field of operation of the company (p > 0.05).

Limitations of the study

Although promising findings have been obtained for all sectors in Düzce, this study has some limitations;

Firstly, although the scale improved to determine relationships between variables is expected to be applicable in Turkey, this study is limited to sectors in Düzce Province. For this reason, it is recommended that the developed scale be applied to other provinces or regions.

Secondly, since it is not possible to obtain employee data through qualitative research methods, the data for this study were obtained through a quantitative research method (questionnaire). Therefore, it is recommended that information be collected through qualitative research methods by interviewing people face-to-face to conduct in-depth research on safety climate factors and safety behavior factors in the future.

Thirdly, this study focused on safety climate factors that affect safety behavior factors to obtain reliable results. Future studies should focus on other factors, such as occupational stress and individual, organizational, social, economic, environmental, and workplace factors that are thought to have direct or indirect effects on safety behavior.

Conclusion

In this study, a model was created to predict the impact of safety climate factors (safety information, safety awareness, safety training, and safety attitude) on safety behaviors. This research model is the first study to provide insights into how employees in sectors operating in Düzce province perceive, understand, and participate in safety practices in the work environment.

According to the findings of the study, safety climate factors (safety information, safety awareness, safety training, and safety attitude) have a significant effect on safety compliance. Also, safety information and safety training have a positive effect on safety participation. However, safety awareness and safety attitude did not have any effect on safety participation.

According to the findings of the study, it was determined that participants working in the furniture sector had much more information about safety issues, and the level of safety training they received was much higher. However, participants working in the textile sector had the lowest level of safety information and safety training. In line with this result, it is recommended that managers of the textile sector should provide training and information to their employees to increase safety awareness of safety-related issues.

Since the research model developed within the scope of the study was applied to participants in the sectors in Düzce Province, the outputs of the study may not be generalizable to different regions due to social, cultural, and regional differences. This study may help sector representatives and employees to pay attention to safety climate factors and what kind of role should be followed in reducing industrial accidents due to safety behaviors. In addition, in future studies, researchers should apply multiple group analysis tests to determine whether the models confirming the effects of the relationships between variables vary according to demographic factors.

In future research, it is recommended to repeat the study in socially and culturally different regions to improve and validate the research findings.

Footnotes

ORCID iD

Muhammet Çil

Ethical considerations

Ethics Committee Permission Certificate dated 01/06/2023 and numbered 2023/197 was obtained from Duzce University Scientific Research and Publication Ethics Committee and it was stated that ‘there is no ethical and scientific drawback’ in conducting such a scientific study in enterprises operating in Duzce province.

Informed consent

Employees participated voluntarily and provided written informed consent.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Social Security Institution. Statistical yearbook, Ankara: Minister of Labor and Social Security. http://www.ssk.gov.tr (2022, accessed 15 April 2024).

Cil

Gedik

. Research on factors affecting the risk-taking behavior of personnel working in the forest products sector. Int J Occup Saf Ergon 2022; 28: 2315–2323.

Bensonch

Argyropoulos

Dimopoulos

, et al. Analysis of safety climate factors and safety compliance relationships in the oil and gas industry. Saf Sci 2022; 151: 105744.

Zohar

. Safety climate in industrial organizations: theoretical and applied implications. J Appl Psychol 1980; 65: 96.

Brondino

Silva

Pasini

. Multilevel approach to organizational and group safety climate and safety performance: co-workers as the missing link. Saf Sci 2012; 50: 1847–1856.

Huang

Verma

Chang

, et al. Management commitment to safety vs. employee perceived safety training and association with future injury. Accid Anal Prev 2012; 47: 94–101.

Mohamed

. Safety climate in construction site environments. J Constr Eng Manag 2002; 128: 375–384.

Clarke

. Contrasting perceptual, attitudinal, and dispositional approaches to accident involvement in the workplace. Saf Sci 2006; 44: 537–550.

Zohar

. Thirty years of safety climate research: reflections and future directions. Accid Anal Prev 2010; 42: 1517–1522.

10.

Zohar

Polachek

. Discourse-based intervention for modifying supervisory communication as leverage for safety climate and performance improvement: a randomized field study. J Appl Psychol 2014; 99: 113.

11.

Tao

Yang

Qiu

, et al. Analysis of safety leadership using a science mapping approach. Process Saf Environ Prot 2020; 140: 244–257.

12.

Guo

Yiu

González

. Predicting safety behavior in the construction industry: development and test of an integrative model. Saf Sci 2016; 84: 1–11.

13.

Tong

Wang

, et al. A dual perspective on work stress and its effect on unsafe behaviors: the mediating role of fatigue and the moderating role of safety climate. Process Saf Environ Prot 2022; 165: 929–940.

14.

Hofmann

Stetzer

. A cross-level investigation of factors influencing unsafe behaviors and accidents. Pers Psychol 1996; 49: 307–339.

15.

Hofmann

Stetzer

. The role of safety climate and communication in accident interpretation: implications for learning from negative events. Acad Manage J 1998; 41: 644–657.

16.

Griffin

Neal

. Perceptions of safety at work: a framework for linking safety climate to safety performance, knowledge, and motivation. J Occup Health Psychol 2000; 5: 347.

17.

Zacharatos

Barling

Iverson

. High-performance work systems and occupational safety. J Appl Psychol 2005; 90: 77.

18.

Ansori

Widyanti

Yassierli . The influence of safety climate, motivation, and knowledge on worker compliance and participation: an empirical study of Indonesian SMEs. Ing Inv 2021; 41: 1–9.

19.

. Psychosocial safety climate and unsafe behavior among miners in China: the mediating role of work stress and job burnout. Psychol Health Med 2020; 25: 793–801.

20.

Vinodkumar

Bhasi

. Safety management practices and safety behavior: assessing the mediating role of safety knowledge and motivation. Accid Anal Prev 2010; 42: 2082–2093.

21.

Christoffel

Gallagher

. Injury prevention and public health: practical knowledge, skills, and strategies. Massachusetts: Jones & Bartlett Learning, 2006.

22.

Simon

Piquard

. Contractor safety performance significantly improves. In SPE International Conference and Exhibition on Health, Safety, Environment, and Sustainability?

1991 Nov 11

(pp. SPE-23253). SPE.

23.

Vredenburgh

. Organizational safety: which management practices are most effective in reducing employee injury rates? J Saf Res 2002; 33: 259–276.

24.

Ahn

Park

. Applying Neal, Griffin, and Hart’s safety climate model to the field. J Korea Safety Manag Sci 2005; 7: 107–117.

25.

Seo

Lee

Kim

, et al. Analyzing safety behaviors of temporary construction workers using structural equation modeling. Saf Sci 2015; 77: 160–168.

26.

Mei

Wang

Liu

, et al. Effects of organizational safety on employees’ proactivity safety behaviors and occupational health and safety management systems in Chinese high-risk small-scale enterprises. Int J Occup Saf Ergon 2020; 26: 101–111.

27.

Neal

Griffin

. A study of the lagged relationships among safety climate, safety motivation, safety behavior, and accidents at the individual and group levels. J Appl Psychol 2006; 91: 946.

28.

Aryee

Hsiung

. Regulatory focus and safety outcomes: examining the mediating influence of safety behavior. Saf Sci 2016; 86: 27–35.

29.

Liu

Huang

, et al. Safety climate, safety behavior, and worker injuries in the Chinese manufacturing industry. Saf Sci 2015; 78: 173–178.

30.

Neal

Griffin

. Safety climate and safety behavior. Aust J Manag 2002; 27: 67–75.

31.

Neal

Griffin

Hart

. The impact of organizational climate on safety climate and individual behavior. Saf Sci 2000; 34: 99–109.

32.

Okur

. Perception of female employees on safety culture and occupational health and safety in textile sector: Diyarbakır province example . Master's Thesis, University of Hasan Kalyoncu, 2021.

33.

Karadal

Merdan

Abubakar

. The mediating role of safety behaviors in the effect of safety climate and safety culture on workplace injuries: a study in the foundry industry. Anemon Muş Alparslan Univ J Soc Sci 2019; 7: 341–351.

34.

Damayanti

Djakfar

Wisnumurti

AMN

. Safety behavior of individual construction workers: the role of big five personality, personal values and safety culture. J Southwest Jiaotong Univ 2022; 57: 575–583.

35.

Liang

Zhou

, et al. Unveiling the mechanism of construction workers’ unsafe behaviors from an occupational stress perspective: a qualitative and quantitative examination of a stress–cognition–safety model. Saf Sci 2022; 145: 105486.

36.

Shin

Gwak

Lee

. Modeling the predictors of safety behavior in construction workers. Int J Occup Saf Ergon 2015; 21: 298–311.

37.

Mohammadfam

Ghasemi

Kalatpour

, et al. Constructing a Bayesian network model for improving safety behavior of employees at workplaces. Appl Ergon 2017; 58: 35–47.

38.

Child

. The Essentials of factor analysis. London: Continuum, 2006.

39.

Karagöz

. SPSS–AMOS–META uygulamalı istatistiksel analizler (SPSS–AMOS–META applied statistical analysis). 2th ed. Ankara: Nobel Press, 2019.

40.

Jackson

. Revisiting sample size and number of parameter estimates: some support for the N: q hypothesis. Struct Equ Modeling 2003; 10: 128–141.

41.

Jung

Lim

Chi

. Impact of work environment and occupational stress on safety behavior of individual construction workers. Int J Environ Res Public Health 2020; 17: 8304.

42.

Büyüköztürk

. Sosyal bilimler için veri analizi el kitabı: İstatistik, araştırma deseni, SPSS uygulamaları ve yorum (Data analysis handbook for social sciences: Statistics, research design, SPSS applications and interpretation). 28th ed. Ankara: Pegem Press, 2020.

43.

Bayram

. Sosyal bilimlerde SPSS ile veri analizi (Data analysis with SPSS in social sciences). 6th ed. Bursa: Ezgi Press, 2017.

44.

Özdamar

. Paket programlar ile istatistiksel veri analizi II (çok değişkenli analizler) [Statistical data analysis with package programs II (multivariate analysis)]. 7th ed. Eskişehir: Kaan Press, 2010.

45.

Yuan

. Fit indices versus test statistics. Multivariate Behav Res 2005; 40: 115–148.

46.

Meydan

Şeşen

. Yapısal eşitlik modellemesi AMOS uygulamaları (Structural equation modeling AMOS applications). 2th ed. Ankara: Detay Press, 2015.

47.

Bentler

. Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Struct Equ Modeling 1999; 6: 1–55.

48.

Hooper

Coughlan

Mullen

. Structural equation modelling: guidelines for determining model fit. Electron J Bus Res Methods 2008; 6: 53–60.

49.

Fan

Thompson

Wang

. Effects of sample size, estimation methods, and model specification on structural equation modeling fit indexes. Struct Equ Model: Multidiscip J 1999; 6: 56–83.

50.

Mulaik

James

Van Alstine

, et al. Evaluation of goodness-of-fit indices for structural equation models. Psychol Bull 1989; 105: 430.

51.

Schermelleh-Engel

Moosbrugger

Müller

. Evaluating the fit of structural equation models: tests of significance and descriptive goodness-of-fit measures. Methods Psychol Res Online 2003; 8: 23–74.

52.

Kalaycı

. SPSS uygulamalı çok değişkenli istatistiksel teknikler (SPSS applied multivariate statistical techniques). 7th ed. Ankara: Asil Press, 2016.

53.

Cain

Zhang

Yuan

. Univariate and multivariate skewness and kurtosis for measuring nonnormality: prevalence, influence, and estimation. Behav Res Methods 2017; 49: 1716–1735.

54.

Cheung

Cooper-Thomas

Lau

, et al. Reporting reliability, convergent and discriminant validity with structural equation modeling: a review and best-practice recommendations. Asia Pac J Manage 2024; 41: 745–783.

55.

Hair

Black

Babin

, et al. Multivariate data analysis. London: Pearson Higher Ed, 2013.

56.

Al-Refaie

. Effects of human resource management on hotel performance using structural equation modeling. Comput Human Behav 2015; 43: 293–303.

57.

Bayram

. Yapısal eşitlik modellemesine giriş: AMOS uygulamaları (Structural equation modeling introduction: AMOS applications). 3th ed. Bursa: Ezgi Press, 2016.

58.

Kline

. Principles and practice of structural equation modeling. 5th ed. New York: Guilford Press, 2023.

59.

Low

BKL

Man

Chan

AHS

, et al. Construction worker risk-taking behavior model with individual and organizational factors. Int J Environ Res Public Health 2019; 16: 1335.

60.

Moon

. A study on the constructs of safety climate and safety performance. Seoul, Republic of Korea: Kyunghee University, 2014.

61.

Burke

Sarpy

Smith-Crowe

, et al. Relative effectiveness of worker safety and health training methods. Am J Public Health 2006; 96: 315–324.

62.

Lim

Park

Lee

. A study on the effects of safety management practices on safety behavior mediated by safety knowledge and safety motivation: focused on South Korea's manufacturing industries. J Intell Inform Syst 2024; 30: 253–284.

63.

Shin

Lee

Park

, et al. A system dynamics approach for modeling construction workers’ safety attitudes and behaviors. Accid Anal Prev 2014; 68: 95–105.

64.

Huang

Lee

McFadden

, et al. Beyond safety outcomes: an investigation of the impact of safety climate on job satisfaction, employee engagement and turnover using social exchange theory as the theoretical framework. Appl Ergon 2016; 55: 248–257.

65.

Brown

Holmes

. The use of a factor-analytic procedure for assessing the validity of an employee safety climate model. Accid Anal Prev 1986; 18: 455–470.

66.

McDonald

Ryan

. Constraints on the development of safety culture: a preliminary analysis. Ir J Psychol 1992; 13: 273–281.