Systematic Analysis of the Contributory Factors Related to Major Coach and Bus Accidents in China

Abstract

The purpose of this study was: (1) to propose a classification system for the contributory factors behind major coach and bus accidents with mass casualties based on the human factor analysis and classification system (HFACS); and (2) to identify the main contributory factors behind accidents and the main indicators of the causal factors. Based on 56 official investigation reports of major coach and bus accidents with more than 10 fatalities, a qualitative content analysis was conducted to develop a modified classification system for the contributory factors behind these accidents, and a gray correlation analysis was conducted to identify the main causative factors and indicators by calculating the correlation degrees. The results showed that the modified classification system for the contributory factors behind major coach and bus accidents can be divided into seven levels: government regulations, the organizational influence of passenger transportation enterprises, unsafe internal operational supervision, preconditions for drivers’ unsafe acts, drivers’ unsafe acts, proximate causes other than the driver’s act, and moderating factors affecting accident severity and probability. The organizational influence of passenger transportation enterprises is the most significant factor affecting the accidents. Thus, passenger transport enterprises must systematically strengthen their responsibility and safety management to prevent accidents. Accident investigations should begin with the accident process to determine the proximate cause as well as the factors that influence the likelihood and severity of the accident.

1. Introduction

1.1. Background

Vehicles such as buses and coaches—the main road transportation vehicles for passengers in developing and underdeveloped countries—account for the majority of fatal road transportation accidents. For example, an incident on 28 September 2019 along the Changchun–Shenzhen Expressway in Jiangsu Province, China, caused 36 deaths and 36 injuries, resulting in an immediate economic loss of CNY 71.00 million [1]. Reducing the number of serious injuries caused by coach and bus accidents is a concern for governments and administrators in developing and underdeveloped countries. However, although there are a number of studies on coach and bus accidents and the factors affecting their severity, few systematic studies have focused on the causal factors or prevention measures. This is due to a low rate of accidents in developed countries owing to the safety level of the coach and bus systems [2] and a lack of accident data in developing and underdeveloped countries.

1.2. Literature Review

According to existing research findings, the safety of road passenger transport is affected by various factors, such as the personnel, vehicles, and environment. Personnel factors mainly refer to drivers but also include passenger factors. Driver factors mainly refer to distracted driving [3,4], fatigued driving [5,6], driver error decision making [7], the use of psychoactive substances and/or alcohol [8], medical incapacity [9], and driver inexperience [10]. Passenger factors include both failure to wear seat belts [11] and a lack of emergency escape training [12].

Studies have revealed that dangerous driving is linked to organizational characteristics, and organizational learning support [13] and improving the organizational safety culture [14,15] can reduce dangerous driving and minimize accidents. Other organizational factors, such as driver scheduling [16], the salary system [17], and regulatory and penalty policies [18] also have strong correlations with accident risk. At the governmental level, the formulation and enforcement of traffic laws [19] and safety education of road traffic participants [20] have important roles in ensuring road safety.

In addition, the safety of the coach and bus, environmental factors, road conditions, and other non-human factors are important factors affecting transportation safety and accident severity. Furthermore, the safety factors of the bus, such as the height of the vehicle’s center of gravity [21]; cant rail and roof structure profile and size [22]; braking systems [23]; weather conditions, such as rain, cold, heat, fog, and wind [24,25]; and road conditions, such as the road geometric design, pavement conditions [26], road markings [27], and limited speed design [28] affect the occurrence and severity of accidents.

The majority of existing studies investigating the causes of accidents focus on a specific factor. Few studies have touched upon serious coach and bus accidents, especially the classification of the contributory factors and the main causal factors from a systematic perspective.

1.3. HFACS and Its Modification

The HFACS, which was originally developed for aviation accidents, is a framework for systematically analyzing the causes of accidents [29]. Due to its extensive diagnostic, sophistication, and reliability features, it has become one of the most extensively used and acknowledged models for accident causation investigations in a variety of industries [30,31]. It is frequently adapted to create models suited for accident analysis in fields other than aviation. Modified versions of the HFACS have been used in the railway (HFACS-Ras [32]), passenger vessel (HFACS-PV [33]), marine (HFACS-MA [34]), chemical (HFACS-HC [35]), oil and gas (HFACS-OGI [36]), and construction (I-HFACS [37]) industries.

Some studies have applied the HFACS to analyze the causes of major road traffic accident systems [38,39]. However, whether the framework can be applied directly to road traffic accident analysis requires further examination and verification. The HFACS is suitable for accidents caused by unsafe acts of front-line operators. The analytical method uses unsafe acts as the starting point for the analysis and then divides the causes of accidents into four levels from bottom to top: unsafe acts, unsafe preconditions for unsafe acts, unsafe supervision, and organizational influences. Pedestrians, non-motor vehicles, and motor vehicles use the road together due to the diversity of road functions, and accidents in which the driver of the passenger vehicle is not responsible or the driver is partially responsible often occur. Therefore, the HFACS framework based on unsafe acts makes it impossible to systematically analyze many road traffic accidents for which the drivers are not fully responsible. Thus, we incorporated three modifications to modify the original HFACS framework to optimize its adaptation to the road passenger transportation domain.

The modified HFACS framework includes seven levels. Levels 2–5 refer to the organizational influence of passenger transportation enterprises, unsafe internal operational supervision, preconditions for unsafe acts by drivers, and unsafe acts by drivers related to road passenger transportation accidents, corresponding to the original HFACS. Three levels named “proximate cause other than driver’s act” (L6), “moderating factors affecting accident severity and probability” (L7), and “government regulations” (L1) were added to create the modified HFACS framework. L6 refers to the factors that directly cause the accident but are not the responsibility of the drivers, such as the engine suddenly catching fire during normal driving. That is, both L6 and L5 (unsafe acts by drivers) are the direct causes of accidents. L7 refers to factors that do not directly contribute to accidents and are unrelated to the driver or passenger transport company. However, these factors can influence the severity and probability of an accident and are similar to moderating variables in statistical analysis. For example, a road crash guardrail of substandard quality will not effectively protect a bus that is out of control, resulting in an even worse accident. Furthermore, as road passenger transportation in China is regulated jointly by road transport authorities, road traffic safety management agencies, and local governments, we included “government regulations” (L1) in the framework.

1.4. The Purpose of This Study

Based on the Chinese regulations for the investigation and handling of production safety accidents [40], a production safety accident involving 10 or more deaths is classified as a major accident. The objective of this study is to investigate and formulate a modified classification system for the contributory factors behind major accidents with mass casualties caused by coaches and buses and to explore the relationships between these factors. Based on this, a more widely applicable method of accident analysis and more effective prevention and control measures for accidents are proposed. Our study findings may provide important guidance for risk control and accident prevention to policymakers and passenger transport enterprises.

2. Methods

2.1. Data Collection

We collected 56 official investigation reports involving major coach and bus accidents with at least 10 fatalities between 2010 and 2019. All the accidents are listed in Table A1 by year. These investigation reports, which contain details of the accident process and accident investigation results, were downloaded from the website of the Ministry of Emergency Management of the People’s Republic of China (https://www.mem.gov.cn/, accessed on 8 October 2020) and its subordinate agencies. It should be noted that the research object of this article is long-distance coach and bus passenger transport and does not include city buses.

2.2. Data Analysis Method

2.2.1. Qualitative Content Analysis

This study used qualitative content analysis to extract, analyze, and categorize the causal factors in accident investigation reports, using the inductive process to construct a classification system for accident causation. As a universal research method applicable to the humanities and social sciences, although qualitative content analysis has the disadvantages of high time consumption and labor intensity, and its conclusions are limited by the limitations of the definitions and classification framework used, compared with other research methods, this method has also its own advantages: firstly, unstructured data can be used as data sources and can offer possibilities for the quantification of categories [41]; secondly, the unobtrusive nature of the researchers’ data use can effectively avoid the influence of the situation on the data and research validity [42]. Thus, the qualitative content analysis of textual information is often used by researchers to form a category system or theory [43]. This indicates that the method is suitable for this study regardless of the data or research purpose.

The core of qualitative content analysis lies in coding, which essentially aims to generate necessary categories in the process of data analysis. For this reason, the coding program of grounded theory is often used in content analysis and has become an important part of qualitative content analysis. As a special method of theorizing from data [44], grounded theory subdivides coding into three stages. Open coding aims to examine and compare the textual materials to be analyzed and conceptualize and classify them. That is to say, the analytical task of open coding is to name concepts, define categories, and develop categories according to their attributes and dimensions. On the basis of open coding, axial coding aims to conduct a correlation analysis of the conceptual categories and to identify and establish various connections and relationships between the conceptual categories, so that the relationships between categories and categories and between categories and subordinate categories become increasingly specific. Selective coding aims to generalize the core categories that can cover all categories from the formed conceptual categories so as to establish a theoretical system [45].

The textual data used in this study were coded with Nvivo 11.0 (QSR International) using the three-level coding system of open, axial, and selective coding guided by the grounded theory methodology. First, we read the accident reports sentence by sentence and labeled and conceptualized the words and phrases that indicated the cause of the accident. Then, we combined and condensed the concepts with comparable meanings to form subcategories and completed the open coding. The subcategories were further combined to form generic categories, and the selective coding was performed on this basis. We further combined and abstracted the generic categories to form main categories and create the accident contributory factor classification system. Some coding examples are presented in Table 1. The coding process in Nvivo 11.0 is mainly performed by saving the free nodes and categorizing them into tree nodes and advanced tree nodes. The reports were coded by the second author and the third author to develop a modified classification system. Any disagreements were resolved through discussion. A sample report structured according to the defined levels and sublevels is presented in Table A2.

Shappell and Wiegmann’s (1997) HFACS and its derivatives in various safety fields were used to construct a modified causal system for major coach and bus accidents. A combination of methods, including the analytical network process [46], odds ratio (OR) [47], why–because analysis, and gray relational analysis (GRA) [34], were used to enhance the practical value of these studies. Therefore, we quantified the influence of each factor in the modified classification system using GRA.

2.2.2. Gray Relational Analysis

GRA is a multifactor statistical analysis method used to determine whether the sequences are closely related by comparing the similarity of their curve geometries. Closer curves indicate greater correlations between the reference sequence and corresponding sequences. GRA is often utilized to calculate uncertain correlations between factors and principal factors. As this method allows for the examination of known information about the system to achieve an understanding of the whole system, it is particularly suitable for the investigation of gray systems with incomplete information and has been successfully applied to the study of natural and social systems in many fields [48]. The main advantage of this method is that there are no requirements regarding the sample size and data characteristics [49]. Due to the small sample size in this study and the large number of analyzed contributory factors, the calculation of the correlation degrees was carried out using GRA.

The calculation process involved four steps:

• (i). Determining the reference sequence $x_0\left(k\right),\left(k=1,2,\cdots ,m\right)$ and comparison sequences $x_i\left(k\right),\left(i=1,2,\cdots ,10\right)$.

• (ii). Normalizing the reference and comparison sequences using the method of maximum values.

• (ii). Calculating the correlation coefficient ${\mathsf{\xi }}_i\left(k\right)$:

${\mathsf{\xi }}_i\left(k\right)=\frac{\underset{i}{\min }({\Delta }_i\left(min\right))+\rho \underset{i}{\max }({\Delta }_i\left(max\right))}{\left|x_0\left(k\right)-x_i\left(k\right)\right|+\underset{i}{\max }({\Delta }_i\left(max\right))}.$

In the equation above, $\rho$ is the resolution coefficient, and $\rho \in \left(0,1\right)$ usually takes the value $\rho$ = 0.5.

• (iv). Striving for the correlation degree $r_i$:

  $r_i=\frac{1}{10}{{\displaystyle \sum }}_{k=1}^{10}{\xi }_i\left(k\right).$

The larger the value of $r_i$ is, the closer the comparison sequence will be to the reference sequence. By ranking the correlation degrees, we can determine the main factors that affect the target value.

3. Results

3.1. The Modified Contributory Factor Classification System

As shown in Figure 1, the modified causative categories of major coach and bus accidents include the following seven levels: L1 (government regulations), L2 (organizational influence of passenger transportation enterprises), L3 (unsafe internal operational supervision), L4 (preconditions for unsafe acts by drivers), L5 (unsafe acts by drivers), L6 (proximate cause other than the driver’s act), and L7 (moderating factors affecting accident severity and probability). These seven layers are the main contributory factors behind coach and bus accidents, with each layer containing several primary and secondary causal indicators. For example, L1 contains three primary indicators: X1 (road transport regulations), X2 (road traffic safety management), and X3 (territorial supervision by the local government). X1 further contains five secondary indicators: X1a (insufficient basic guarantees), X1b (insufficient supervision), X1c (regulation violations), X1d (the ineffective implementation of policy), and X1e (insufficient organizational synergy). However, L6 contains only three primary causal indicators and no secondary causal indicators. Thus, there is no correlation analysis between the primary and secondary indicators. The description of the modified contributory factor classification system is presented in Table A3.

It is important to note that X16 (emergency situation related to vehicle parts) is different from X11 (substandard condition of vehicle parts) under L4 (preconditions for unsafe acts by drivers). X16 refers to the “sudden failure of vehicle parts”, which can directly lead to the accident, while X11 requires the intermediary factor of “unsafe acts by drivers”. X17 (illegal transportation of dangerous goods) refers to unsafe carrying such as overloading, illegal loading, or the illegal transport of inflammable and explosive materials, as required by the owner of the vehicle. In China, many bus drivers are employed by the owner of the bus. They only drive the bus and do not need to be responsible for anything else related to the operation of the vehicle.

3.2. Reliability and Validity of the System

To ensure the reliability of the system, Cohen’s κ formula was used to assess the consensus reached on the coding by the two coders. As shown in Table 2, six of the seven levels had a good reliability, while one was within the basic acceptable range. The validity of the system was ensured in several ways. Firstly, the 56 official accident investigation reports, which present valuable and detailed information about the accidents, were analyzed. Secondly, the coding was performed using terms from the public safety industry and road passenger transport-related management regulations. Finally, the results of the system were discussed carefully with experts and practitioners in related fields, who all shared a high level of agreement with the modified contributory factor classification system.

3.3. Correlation Analysis Results

Before we calculated the correlation degrees, we first extracted and summarized the statistics of all the contributory factors from the survey reports of the accidents based on the modified classification system (the results are shown in Table A4). The gray correlation degrees were calculated in three parts: first, the correlation degrees between each first-level factor (main category) and their summation; second, the correlation degrees between each primary indicator (generic category) and the first-level factor it belonged to; and third, the correlation degrees between each secondary indicator (subcategory) and the primary indicator to which it belonged. Owing to space constraints, we demonstrate only the specific steps of the calculation of the gray correlation degrees for the first part, as the other two parts follow a similar pattern.

3.3.1. Correlation Analysis between the First-Level Factors and Their Summation

There are four steps in calculating the gray correlation degrees between each contributory factor from L1 to L7 and their summation (L0). The numbers of L1 to L7 are obtained by adding up the contributing factors they contain. In the case of L1, L1 = X1 + X2 + X3. To save space, we presented the analytical data from step 1 and the result data from step 2 and step 3 in the Appendices (see Table A5, Table A6, Table A7 and Table A8).

• (i). We took L0 as the reference sequence and L1–L7 as the comparative sequences. In gray relational analysis, the reference sequence and comparison sequence are also called the parent sequence and subsequence, similar to a dependent variable and independent variable in statistical analysis. This part mainly aims to identify which layer of the causal factors in L1–L7 is more closely related to the accident’s cause by calculating the correlation degree. Therefore, L1–L7 were taken as the comparison sequences, and the data sequence L0 formed by the addition of L1–L7 was taken as the reference sequence. The data of the reference sequence and the comparative sequences, as shown in Table A5, were summarized based on the initial data in Table A4.

• (ii). The reference and comparison sequences were normalized using the maximum values method. We divided all the numbers in each row of L0–L7 by the maximum value of each row. Taking L0 as an example, since its maximum value is 224, we divide L0’s numbers by 224. The results are shown in Table A6.

• (iii). The absolute difference between the comparison sequences and the reference sequence was calculated. We subtracted each of the comparison sequence data from its corresponding reference sequence data and then took the absolute values. Accordingly, we calculated the correlation coefficients between the comparison sequences and reference sequence. The results are shown in Table A7 and Table A8. Thus, the groups of gray correlation coefficients were obtained, and each group contained 10 correlation coefficients.

• (iv). The correlation degrees between factors L0 and L1–L7 in the first layer were obtained. The seven gray correlation degrees between each first-level factor of L1–L7 and L0 (the summation of L1–L7) are as follows: $r_1=r$ (L0, L1) = 0.716, $r_2=r$ (L0, L2) = 0.891, $r_3=r$ (L0, L3) = 0.785, $r_4=r$ (L0, L4) = 0.806, $r_5=r$ (L0, L5) = 0.793, $r_6=r$ (L0, L6) = 0.676, $r_7=r$ (L0, L7) = 0.726. Thus, the ranking of the seven factors is as follows: L2 > L4 > L5 > L3 > L7 > L1 > L6. Based on the above analysis, the factors in the middle levels are the main causes of the accidents. Thus, we can state that the gray correlation degrees between each level of the classification system for the contributory factors and their summation appear to be weak at both ends but strong in the middle.

3.3.2. Correlation Analysis between Primary Indicators and Their First-Level Factors

We took each first-level factor as the reference sequence and the primary indicators under the first-level factors as the comparative sequences to calculate the correlation degrees between the primary indicators and the first-level factors. The calculated the correlation degrees between the primary indicators and the first-level factors are listed in Table 3.

Based on the analysis results in Table 3, the main influencing indicators affecting L1–L7 are X2 (road traffic safety management) (r (L1, X2) = 0.823), X5 (resource management) (r (L2, X5) = 0.778), X8 (insufficient supervision) (r (L3, X8) = 0.747), X10 (personnel factors) (r (L4, X10) = 0.803), X14 (errors) (r (L5, X14) = 0.743), X17 (illegal transportation of dangerous goods) (r (L6, X17) = 0.662), and X19 (individual-level factors) (r (L7, X19) = 0.690). Additionally, the correlation coefficients between X13 (road environment) and L4 (preconditions for unsafe acts by drivers) (r (L4, X13) = 0.801) as well as X10 (personnel factors of the driver) and L4 ($r$ (L4, X10) = 0.803) are comparatively high.

According to the gray correlation degrees in Table 3, with respect to the drivers, errors lead to a higher proportion of accidents, and the primary cause of these errors is the driver’s individual factors. With respect to passenger transportation enterprises, the primary causes include inadequate internal supervision and resource management. At the level of government regulations, ineffective road traffic safety management has a greater impact on accidents than insufficient road transport regulations and territorial supervision by the local government. The main proximate cause other than the driver’s act is the illegal transportation of dangerous goods, as required by the owner of the vehicle. Individual-level factors are the main moderating factors affecting the severity and probability of accidents.

3.3.3. Correlation Analysis between Secondary Indicators and Their Primary Indicators

Similar to the method described above, we took each primary indicator as the reference sequence and the secondary indicators under the primary indicator as the comparative sequences to calculate the correlation degrees between the secondary indicators and their primary indicators. The results of the correlation degrees are shown in Table 4.

Based on Table 4, X2a (lax enforcement of traffic laws) (r (X2, X2a) = 0.784), X5a (poor human resource management) (r (X5, X5a) = 0.757), X8c (dynamic monitoring in a virtual sense) (r (X8, X8c) = 0.767), X10a (poor safety awareness) (r (X10, X10a) = 0.840), X14b (decision-making errors) (r (X14, X14b) = 0.796), and X19a (passengers’ poor safety awareness) (r (X19, X19a) = 0.802) are the main indicators of X2 (road traffic safety management), X5 (resource management), X8 (insufficient supervision), X10 (personnel factors of the driver), X14 (errors), and X19 (individual-level factors), respectively. Table 4 shows that among the skill-based, decision-making, and perceptual errors, decision-making errors, such as slamming the steering wheel at a high speed, are the most common errors of drivers. A lack of safety awareness is the most important personal factor. In terms of the passenger transport enterprises and road traffic safety management authorities, dynamic monitoring and control, inadequate human resource management, and insufficient, lax enforcement are the major problems. In terms of the moderating factors at the individual level, the low level of passenger safety awareness is an urgent issue that needs to be addressed.

4. Discussion

The qualitative content and gray correlation analysis indicated that the majority of the coach and bus accidents with mass casualties were mostly caused by human factors, and the primary responsibility lay with the passenger transport enterprises. Furthermore, different levels of factors can have an impact on the occurrence of major passenger transportation accidents. The findings of this study have implications for traffic management practice in two ways. On the one hand, they provide new ideas for the conduct of accident investigations, and on the other hand, they provide policy recommendations for the improvement of related accident control strategies.

4.1. On Accident Investigations

4.1.1. Accident Causation Analysis Should Begin with the Accident Process

According to the statistical results of the proximate cause events of the 56 accidents in Table 5 (see Table A4 for the original data), the proportions of “unsafe acts by drivers” and “proximate cause other than driver’s act” as proximate events are 75.9% and 24.1%, respectively. Moreover, we counted the moderating factors that appeared in all the accident reports, and we found that the frequency of “moderating factors affecting accident severity and probability” is as high as 95 times, with 36 moderators at the individual level and 59 moderators at the organizational level. This indicates that the method of taking “unsafe acts by drivers” as the starting point of accident cause analysis is not applicable to the analysis of all passenger transport accidents, and the “accident cause other than driver’s act” and “modeling factors affecting accident severity and probability” can easily be ignored, which renders the results of the accident cause analysis as incomplete and unscientific.

This implies that the investigation process should begin with the accident process, identifying the causal chain of events based on the time series. Specifically, we should clarify the proximate causes in the analytical process. That is, we must determine whether the event was caused by the unsafe act of the driver or some other factor unrelated to the driver, or a combination of both. Furthermore, while investigating accidents, we need to consider the moderating factors and supervisory factors. The moderating factors may affect the severity and incidence of road passenger transportation accidents, while investigating supervisory and managerial causes can help us to identify safety management system defects. Accordingly, we constructed the AP–PM–OR model (accident process–proximate and moderating factors–organization and regulatory causes) for the investigation of road passenger transportation accident causation, as shown in Figure 2. A causal analysis of road passenger transportation accidents should begin with the accident process, as indicated by the AP–PM–OR model. Starting with a time series description of the event chain and working backward to the proximate and moderating factors, we can then examine the organizational and regulatory factors that contributed to the accident. In Figure 2, the factors on the right are caused by the events on the left, which can be understood by right-to-left reading.

4.1.2. The Analysis of the System Vulnerabilities Should Be Strengthened

Direct accident causes are the proximate causes; however, the root cause lies in the fact that certain risk factors are not effectively controlled [50]. Effective risk control relies on the establishment of an effective safety management system that monitors the risk of an accident constantly, rather than having a supervisor manage safety during working hours only [51]. Hence, in addition to the analysis of proximate causes and moderating factors, a detailed analysis of safety management systems is very important and necessary, especially with respect to the reasons why regulatory bodies and passenger transport enterprises fail to fulfill their safety responsibilities, enabling the gaps in institutional mechanisms to be identified and modified. Thus, we can reduce the vulnerability of the safety management system and lower the accident risk and probability.

4.2. On Accident Prevention and Control

4.2.1. Passenger Transport Enterprises Should Be Prohibited from Contracting Traffic Operation Rights to Individuals

One of the main reasons for the low level of safety management in passenger transport in China is the business model of contracts for operations, wherein transport companies hire individual contractors who manage their vehicles on specific routes and turn over the cost to the company as per the contract. The greatest problem with this business model is that these contractors often ignore safety measurements, including their failure to perform regular maintenance on the vehicles as required, provide formal training to the drivers they hire, and allow the drivers to undergo regular health checks, in order to maximize profits, which directly affects the safety of the transportation process. In addition, it is difficult to ascertain legal liability after an accident. As a result, it is imperative to prohibit the contracting of buses and coaches and implement the intensive and large-scale operation of passenger transport enterprises instead.

4.2.2. Construct a Legal Responsibility System for Passenger Transport Enterprises

According to the correlation analysis results above, L2 (organizational influence of passenger transportation enterprises), L3 (unsafe internal operational supervision), L4 (preconditions for unsafe acts by drivers), and L5 (unsafe acts by drivers) are important causative factors of an accident. This means that safety management by passenger transport enterprises has an important impact on accident prevention and control. However, Chinese laws pertaining to road traffic safety management contain few regulations related to corporate responsibilities at present. Although there are some terms related to drivers’ safety management of passenger transport companies in the Regulations on the Safety Management of Road Passenger Transport Companies [52], such as employment, pre-job training, safety education, and regular assessment, most of these are on the macro-level. Moreover, the specific implementation process and corresponding legal responsibilities are not clearly defined. Thus, these regulations have not been truly implemented by passenger transport companies because of the lack of penalties for non-compliance. Therefore, in view of the general lack of responsibility of Chinese passenger transport enterprises, laws and regulations should be modified and implemented to compel enterprises to fulfill their responsibilities.

4.2.3. Strengthen the Governance of Accident Risks Based on a System Perspective

According to Table A4, of the 56 accidents studied in this article, 12 accidents were caused by collisions with other vehicles, where the passenger vehicle had no legal responsibility. A total of 16 accidents were caused by coach drivers’ improper responses to other road traffic participants (such as motorcycles and tricycles), which means that these accidents were not entirely the responsibility of the coach drivers. In addition, one of the important reasons for the high number of accidents causing heavy casualties is that passengers did not ride in the prescribed manner, whether they stood in the carriage or sat without wearing seat belts. When the vehicle was out of control, they were squeezed against each other, and some passengers fell out of the window. This shows that there are significant differences between the causes of road traffic and other traffic accidents, such as railway, aviation, and navigation accidents.

To prevent and control road traffic accidents, managers must comprehensively evaluate the causes of accidents and factors influencing accident severity utilizing the system perspective as the only appropriate response to the complexity of this task [53] and considering road passenger transport as a complex system. In other words, managers should investigate all kinds of road passenger safety hazards in the management of road traffic accidents. On the one hand, they should pay attention to factors involved in the safety management of passenger transportation enterprises, while on the other hand, they should investigate other organizational and individual factors that affect road passenger transportation safety. Government regulators should actively introduce the ISO39001 standard to establish a systematic road traffic safety management mechanism so as to eliminate flaws in the fragmented prevention mechanisms and related management systems. Furthermore, traffic rules should be incorporated into the education syllabus, and the public should be made aware of the serious dangers of traffic accidents through various media channels [20]. By paying attention to all the aspects of the system factors, the public’s traffic safety risk literacy will be improved, and the concept of the whole society concerning traffic safety will be established and strengthened.

5. Conclusions

In this study, we analyzed 56 major coach and bus accidents by examining the official investigation reports of accidents with more than 10 fatalities using content analysis and gray correlation analysis. The study provided a systematic understanding of the reasons for which coach and bus accidents with more than 10 fatalities occur. According to the results of the qualitative content analysis, HFACS and modified HFACS versions in other fields are not applicable to road transport accident analysis. Due to the characteristics of the openness and mobility of road traffic, the proximate causes of road transport accidents include not only unsafe acts by drivers but also other proximate causes. Thus, accident causation analysis should begin with the accident process, identifying the proximate causes and moderating factors of the accident, and only then move outward to analyze the organizational and institutional causes of the accident. The gray correlation analysis results show that the safety management of passenger transport enterprises and preconditions for unsafe acts by drivers are the main potential sources of the risks of coach and bus accidents, but other causal factors should not be ignored. In particular, it is necessary to continuously improve the legal responsibility system. Clear national legislation and the enforcement of regulations should be developed on the macro-level. Moreover, passenger transport enterprises’ practices should be incorporated into national legislations and laws. However, our conclusions cannot be generalized for all countries because of their institutional differences. Accident cause analysis should be used to verify the scientific robustness and effectiveness of the modified contributory factor classification system and the AP–PM–OR model of accident causation analysis in the future.

Author Contributions

Conceptualization, Y.S. and J.H.; methodology, Y.S.; software, J.H.; validation, J.H., Q.Z. and C.W.; formal analysis, C.W.; investigation, J.H.; resources, J.H.; data curation, Q.Z. and C.W.; writing—original draft preparation, Y.S.; writing—review and editing, J.H. and Q.Z.; visualization, C.W.; supervision, J.H.; project administration, Y.S.; funding acquisition, Y.S. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest

The authors declare no conflict of interest.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figures and Tables

Generate image description

Figure 1. The modified contributory factor classification system.

Generate image description

Figure 2. AP–PM–OR model of accident causation analysis.

Table 1

Examples of accident report coding (partial).

Original Survey Report Statements	Open Coding			Axial Coding	Selective Coding
	Labeling	Conceptualizing	Sub-Category	GenericCategory	MainCategory
6 March 2014: The driver drove a Chuan R41518 bus on the road and slid downhill in neutral gear near Wuyi bridge.	Sliding downhill in neutral gear	Improper operation	Skill-based error	Errors	Drivers’ unsafe acts
10 November 2012: The driver drove the motor vehicle onto *** Road without maintaining a safe speed.	Failure to maintain safe speed	Unsafe vehicle speed	Decision-makingerrors
24 September 2013: The driver’s failure to carefully observe the intersection was the direct cause of the accident.	Failure to observe carefully	Pass without confirmation of safety
19 February 2013: EQ01466 bus drove to *** Town, and when it encountered a vehicle driving in the opposite direction, it was too close to the edge (cliff).	Too close to the edge when encountering oncoming traffic	Misjudged lateral distance	Perceptual error
12 April 2012: The accident was caused directly by ***, the driver of the Anhui L60670 bus, illegally occupying the other lane and colliding with oncoming traffic.	Illegally occupying the other lane	Driving in the opposite direction	Incidental violation	Violations
28 August 2011: At the time of the accident, the vehicle was traveling at a speed greater than 65 km/h.	Exceeding the speed limit	Exceeding the speed limit	Habitualviolations
10 June 2015: The large ordinary bus AL0240 overtook two trucks loaded with sand and gravel at 23 km + 300 m, scraped against AR0117 coming from the opposite direction, and fell over the steep cliff.	Illegal overtaking	Illegal overtaking

Note: *** represents some specific names, is a kind of anonymous processing of the relevant road name, city name and driver’s name.

Table 2

Cohen’s κ scores of the reliability test for selective coding.

Categories of the Selective Coding Stage	Cohen’s κ Scores
L1: Government regulations	0.78 b
L2: Organizational influence of passenger transportation enterprises	0.85 a
L3: Unsafe internal operational supervision	0.84 a
L4: Preconditions for unsafe acts by drivers	0.92 a
L5: Unsafe acts by drivers	0.95 a
L6: Proximate cause other than driver’s act	0.86 a
L7: Moderating factors affecting accident severity and probability	0.82 a

Note: The superscript ^a means the coefficient is adequate, and ^b means it is acceptable but not especially good.

Table 3

The gray correlation degrees between the primary indicators and their first-level factors.

$r$	L1	$r$	L2	$r$	L3	$r$	L4	$r$	L5
X1	0.711	X4	0.712	X8	0.747	X10	0.803	X14	0.743
X2	0.823	X5	0.778	X9	0.670	X11	0.630	X15	0.712
X3	0.692	X6	0.715			X12	0.633
		X7	0.644			X13	0.801
$r$	L6	$r$	L7
X16	0.479	X19	0.690
X17	0.662	X20	0.679
X18	0.577

Table 4

The gray correlation degrees between the secondary indicators and their primary indicators.

$r$	X1	$r$	X2	$r$	X3	$r$	X4	$r$	X5
X1a	0.576	X2a	0.784	X3a	0.851	X4a	0.743	X5a	0.757
X1b	0.794	X2b	0.582	X3b	0.807	X4b	0.797	X5b	0.734
X1c	0.755	X2c	0.725	X3c	0.580	X4c	0.636
X1d	0.683	X2d	0.755	X3d	0.604
X1e	0.775	X2e	0.636
		X2f	0.646
$r$	X6	$r$	X7	$r$	X8	$r$	X9	$r$	X10
X6a	0.767	X7a	0.655	X8a	0.734	X9a	0.617	X10a	0.840
X6b	0.767	X7b	0.810	X8b	0.751	X9b	0.642	X10b	0.793
X6c	0.800			X8c	0.767			X10c	0.641
								X10d	0.641
								X10e	0.579
$r$	X13	$r$	X14	$r$	X15	$r$	X19	$r$	X20
X13a	0.753	X14a	0.593	X15a	0.868	X19a	0.802	X20a	0.695
X13b	0.785	X14b	0.796	X15b	0.493	X19b	0.680	X20b	0.727
X13c	0.750	X14c	0.653					X20c	0.572
X13d	0.617							X20d	0.572
X13e	0.743							X20e	0.636
								X20f	0.547

Table 5

Descriptive statistics of the proximate causes of the accidents.

Proximate Causes			Frequency (Percentage)
Proximate causes	L5: Unsafe acts by drivers	X14 Errors	47 (42.0)
		X15 Violations	38 (33.9)
	L6: Proximate cause other than the driver’s act	X16 Emergency situation related to vehicle parts	4 (3.6)
		X17 Illegal transportation of dangerous goods	11 (9.8)
		X18 Other vehicle ramming	12 (10.7)
Total			112 (100)

Note: there is often more than one proximate cause of an accident. Thus, the number of proximate causes is not always equal to the number of accidents.

Appendix A

Table A1

The table of analyzed coach and bus accident cases (by year).

No.	Year	Full Name of the Accident Case
1	2010	“3/14” major accident on Datong Ring Expressway in Shanxi Province
2	2010	“5/23” major accident in Fuxin City, Liaoning Province
3	2010	“6/26” major accident in Haiyuan County, Ningxia Hui Autonomous Region
4	2010	“10/16” major accident in Haixi Mongolian and Tibetan Autonomous Prefecture, Qinghai Province
5	2011	“1/29” major accident in Guyuan City, Ningxia Hui Autonomous Region
6	2011	“2/4” drowning accident in Shaowu City, Fujian Province
7	2011	“3/16” major accident on Linfen Beltway in Shanxi Province
8	2011	“7/4” major accident on Xiantao Section of Suizhou–Yueyang Expressway in Hubei Province
9	2011	“7/22” fire accident on Xinyang Section of Beijing–Hong Kong–Macao Expressway in Henan Province
10	2011	“8/28” major accident in Shangyi County, Hebei Province
11	2011	“10/1” major accident in Xingshan County, Hubei Province
12	2011	“10/7” major accident on Binhai New Area–Baoding Expressway in Tianjin Municipality
13	2011	“10/7” major accident in Sheqi County, Henan Province
14	2011	“12/12” major accident in Feng County, Jiangsu Province
15	2012	“1/4” major accident on Shanghai–Kunming Expressway in Guizhou Province
16	2012	“2/25” major accident on Jincheng Section of National Highway 207 in Shanxi Province
17	2012	“3/13” major accident in Barkam County, Aba Tibetan and Qiang Autonomous Prefecture, Sichuan Province
18	2012	“4/7” major accident in Dalian Free Trade Zone, Liaoning Province
19	2012	“4/12” major accident on Xiaoxian Section of National Highway 311 in Anhui Province
20	2012	“4/22” major accident on Suzhou Section of Changshu–Hefei Expressway in Jiangsu Province
21	2012	“8/26” major accident on Yan’an Section of Baotou–Maoming Expressway in Shaanxi Province
22	2012	“8/31” major accident on Sanmenxia Section of Lianyungang–Horgos Expressway in Henan Province
23	2012	“11/10” major accident on Qiangao Highway in Xinjiang Uygur Autonomous Region
24	2012	“12/9” major accident in Minquan County, Henan Province
25	2013	“2/1” major accident in Gulin County, Sichuan Province
26	2013	“2/1” major accident in Ningxian County, Gansu Province
27	2013	“2/2” major accident in Congjiang County, Guizhou Province
28	2013	“2/19” Major accident in Jianshi County, Hubei Province
29	2013	“3/12” major accident on Jinzhou Yangtze River Bridge of Erenhot–Guangzhou Expressway in Hubei Province
30	2013	“6/18” major accident in Changji Hui Autonomous Prefecture, Xinjiang Uygur Autonomous Region
31	2013	“8/9” major accident on Hefei Section of Bengbu–Hefei Expressway in Anhui Province
32	2013	“9/24” major accident in Lixin County, Anhui Province
33	2014	“3/3” major accident in Hezuo City, Gansu Province
34	2014	“3/5” fire accident in Jilin city, Jilin Province
35	2014	“3/6” major accident in Yilong County, Sichuan Province
36	2014	“3/25” major accident on Qianjiang Section of Baotou–Maoming Expressway in Chongqing Municipality
37	2014	“7/19” major accident on Shanghai–Kunming Expressway in Hunan Province
38	2014	“8/9” major accident in Lhasa City, Tibet Autonomous Region
39	2014	“8/26” major accident on Guazhou Section of Lianyungang–Horgos Expressway in Gansu Province
40	2015	“1/16” major accident on Laizhou Section of Rongcheng–Wuhai Expressway in Shandong Province
41	2015	“4/4” major accident in Nayong County, Guizhou Province
42	2015	“5/15” major accident in Xianyang City, Shaanxi Province
43	2015	“6/10” major accident in Gongga County, Tibet Autonomous Region
44	2015	“6/26” major accident on Wuhu Section of Nanjing–Wuhu Expressway in Anhui Province
45	2015	“9/11” major accident on Xinxian Section of Daqing–Guangzhou Expressway in Henan Province
46	2016	“6/26” major accident on Chenzhou Section of Yizhang–Fengtouling Expressway in Hunan Province
47	2016	“7/1” major accident on Tianjin–Jizhou Expressway in Tianjin Municipality
48	2016	“8/20” major accident in Ziyun County, Guizhou Province
49	2017	“3/2” major accident in Yunxian County, Yunnan Province
50	2017	“4/29” major accident in Arun Banner, Inner Mongolia Autonomous Region
51	2017	“5/15” major accident in Yingtan City, Jiangxi Province
52	2017	“7/6” major accident on Longmen Section of Guangzhou–Heyuan Expressway in Guangdong Province
53	2017	“7/21” major accident on Yuxian Section of National Highway 109 in Hebei Province
54	2017	“8/10” major accident on Ankang Section of Beijing–Kunming Expressway in Shaanxi Province
55	2018	“2/20” major accident in Ganzhou City, Jiangxi Province
56	2019	“9/28” major accident on Wuxi Section of Changchun–Shenzhen Expressway in Jiangsu Province

Table A2

The coding process for the investigation report of the “2/20” major accident in Ganzhou City.

Defined Levels and SubLevels of Contributory Factors		Original Survey Report Statements
L1 Government regulations
X1 Road transport regulations	X1b Insufficient supervision	The Ruijin Municipal Road Transport Administrative Bureau did not strictly inspect and supervise passenger transport enterprises according to regulations.
	X1c Regulation violations	The Ruijin Municipal Road Transport Administrative Bureau found the illegal behavior of the Ruixiang Company and Lu Xiang Company four times, but never passed on the problem indicators to the public security traffic management department for further investigation. They did not issue corrective instructions to these two passenger transport companies, and their records regarding the questioning of the enterprises were falsified.
	X1e Insufficient organizational synergy	The Ruijin City Transport Bureau did not guide and supervise the Ruijin City Road Transport Authority in the regulation of passenger transport enterprises in the replacement of the material submitted for road transport operating permits.
X2 Road traffic safety management	X2d Inadequate inspection and supervision	There is an unreasonable distribution of duty points among the four traffic management brigades of the Ningdu County Public Security Bureau, thereby preventing the formation of a road inspection network that covers the entire county.
X3 Territorial supervision by the local government	X3a Poor leadership and organization	Multiple departments of the Ruijin Municipal Government do not co-operate effectively when enforcing traffic laws for passenger vehicles and passenger transport enterprises.
	X3b Inadequate inspection and supervision	The Ruijin Municipal Government has not investigated or managed the perennial chaos of the passenger transport market.
L2 Organizational influence of passenger transportation enterprises
X4 Basic safeguard	X4b Lack of safety managers	The number of safety managers at Ruixiang Transportation Enterprise was insufficient.
X6 Organizational culture	X6a Inadequate implementation of policies and systems	Ruixiang Transportation Enterprise did not hold safety meetings and conduct safety inspections as required.
	X6b Incomplete management of hidden hazards	Ruixiang Transportation Enterprise failed to take strong measures to prevent overcrowding in vehicles.
	X6c Failure to fulfill responsibility	Ruixiang Transportation Enterprise failed to carry out effective safety management of the vehicles operated jointly with Luxiang Transportation Enterprise and failed to implement their main responsibility for safety production.
X7 Operation method	X7b Improper vehicle operation	The Ruixiang company has a contract with the Luxiang company enabling it to operate and manage 20 buses of six passenger lines (including the accident bus).
L3 Unsafe internal operational supervision
X9 Supervision violation	X9a Non-compliant operation	In Jixiang Passenger Transportation vehicles, Luxiang Transportation Enterprise promotes overcrowding by linking driver and passenger earnings to business gains.
L4 Preconditions for unsafe acts by drivers
X10 Personnel factors of driver	X10a Poor safety awareness	Neither the driver nor passengers were wearing seat belts at the time of the accident.
X12 Weather conditions		There was precipitation at the time of the accident on the incident road.
X13 Road environment	X13a Bad road alignment	The incident section has a bend radius of 275 m and a longitudinal slope of 5.72%.
	X13b Bad road type	There is a deep ditch on the west side of the road, 6.8 m below the road surface
	X13c Lack of safety facility	The road section where the accident occurred is lacking in protection facilities.
L5 Unsafe acts by drivers
X14 Error	X14a Skill-based error	An improper operation of the driver caused the bus to swing widely before the accident.
X15 Violation	X15a Habitual violation	During the time of the accident, the vehicle’s speed increased from 58 km/h to 70 km/h from 10:12:23 to 10:12:45, despite the 40 km/h speed limit.
L7 Moderating factors affecting accident severity and probability
X19 Individual-level factor	X19a Poor safety awareness among passengers	Some passengers were standing in the carriage at the time of the accident.
X20 Organizational-level factor	X20b Inadequate road/bridge safety management	In May 2017, the Jiangxi Provincial Road Administration issued an additional security project plan for the incident section, requiring completion by 1 February 2018, but by the time of the accident, The Ningdu Branch of the Ganzhou City Road Administration Bureau still had not started construction.

Notes: 1. The full name of this accident is the “2/20” major accident in Ganzhou City, Jiangxi Province. It is No. 55 in Appendix A; 2. This accident caused 11 deaths and 20 injuries; 3. The owner of the bus was Ruixiang Transportation Enterprise, but the bus was contracted to Luxiang Transportation Enterprise. Jixiang Passenger Transportation was an independent operating entity of Luxiang Transportation Enterprise.

Table A3

Description of the modified classification system for contributory factors.

Contributory Factors	Description
◆ Government regulations
▲ Road transport regulations
• Insufficient basic guarantees	The system, methods, means, and staffing required for safety supervision are inadequate
• Insufficient supervision	The safety supervision of passenger transport enterprises, operating vehicles, automobile maintenance enterprises, or passenger stations is not in place
• Regulation violations	Violations in safety supervision, such as the violation of operational procedures, lower penalty standards, failure to correct any problems found, and failure to issue operating certificates as required
• Ineffective implementation of policy	Ineffective implementation of the relevant national laws, higher authorities’ regulations, local government policies, or bureau (institute) work plans
• Insufficient organizational synergy	Insufficient coordination with higher authorities, subordinate organizations, or other entities in handling-related matters
▲ Road traffic safety management
• Lax enforcement of traffic laws	Problems in road enforcement, such as inadequate supervision and ineffective investigation and punishment of violations
• Insufficient road access protection	Untimely problems with road traffic protection, including road hazard investigation, traffic order control, and accident rescue
• Inadequate safety education	Inadequate education and training of passenger transport companies and their drivers
• Inadequate inspection and supervision	Failure to detect the problem of unsafe operation by passenger transport companies, the inspection and supervision of the operating vehicles’ safety is not in place
• Ineffective implementation of policy	Failure to effectively implement the relevant laws, regulations, policies, or internal work programs
• Insufficient organizational synergy	Insufficient cooperation with relevant agencies in handling existing problems or problems identified during safety supervision
▲ Territorial supervision by the local government
• Poor leadership and organization	No effective organization and leadership of the safety supervision work of the relevant road transport safety management authorities
• Inadequate inspection and supervision	No effective supervision and inspection of road transport authorities, road traffic safety management agencies, and transport enterprises
• Regulation violations	Management beyond the authority of local safety regulations
• Ineffective implementation of policy	Inadequate implementation of the policies of the ministry (bureau) or higher government
◆ Organizational influence of passenger transportation enterprises
▲ Basic safeguards
• Inadequate production of a safety management system	Inadequate institutional system for safety management regulations, such as responsibilities for production safety, parking of non-operating vehicles, dispatching of operating vehicles, dynamic monitoring, and drivers’ safety education and training
• Lack of safety managers	Leaders in charge of production, full-time safety managers, and dynamic monitors are not equipped
• Insufficient safety budgets	Not equipped with special funds for production safety as required
▲ Resource management
• Poor human resource management	Inadequate management of the workers’ hiring, qualification review, education and training, remuneration, and assessment
• Inadequate facility resource management	Inadequate management of vehicles, dynamic monitoring equipment, or other facilities
▲ Organizational culture
• Inadequate implementation of policies and systems	Laws and regulations of road traffic or road transport production safety are often ignored
• Incomplete management of hidden hazards	Risks of production safety, such as the environment, management, and vehicle operation, are not investigated or not effectively investigated and managed
• Failure to fulfill responsibilities	People in charge of safety operations or safety management personnel do not perform their duties seriously
▲ Operation methods
• Improper business operation	Problems with the operation of passenger transport enterprises, such as contracting passenger transport enterprises to individuals, lack of guidance and supervision of branch offices, and lack of reasonable provision of competent leaders
• Improper vehicle operation	Passenger vehicles are often contracted or affiliated operations
◆ Unsafe internal operational supervision
▲ Insufficient supervision
• Poor management of drivers	Driver safety education, training, and assessment are not implemented as required by the operation management department; drivers’ violations are not strictly handled
• Poor vehicle management	Poor management of the vehicles’ technical safety condition, repair and maintenance, effectiveness of parts, private modification, or parking
• Dynamic monitoring in a virtual sense	Problems with dynamic monitoring, such as inadequate configuration, unreasonable settings, irregular operation, and lack of warning and correction of drivers’ unsafe acts
▲ Supervision violations
• Non-compliant operation	Irregularities in the operation, such as illegal authorization to operate, illegal organization operation, and forgery of operation materials
• Non-compliance with requirements	Noneffective implementation of the relevant laws of the state, regulations of road transport authorities, or correction notices of road traffic safety management agencies
◆ Preconditions for unsafe acts by drivers
▲ Personnel factors of driver
• Poor safety awareness	The safety concept has not been built up
• Lack of knowledge and skills	Lack of driving knowledge and skills
• Inexperience	Lack of experience and knowledge in dealing with unexpected situations
• Poor mental health	Unable to concentrate on safe driving due to fatigue, alcohol, or drugs
• Inattentiveness	Actions that distract the driver from concentrating on safe driving, such as answering the phone and chatting with passengers while driving
▲ Substandard condition of vehicle parts
• Not up to standard in terms of vehicle safety	Vehicle braking, driving, steering, and other components of the safety and technical condition are not up to standard, which will lead to the drivers’ errors
▲ Weather conditions
• Bad weather conditions	Bad weather that affects driving safety and road access conditions, such as rain, snow, and fog
▲ Road environment
• Bad road alignment	The road design characteristics of combined horizontal and vertical alignments that affect safe driving, such as the slope section, curve section, and curve and slope section
• Bad road type	From the cross-sectional view of traffic accident-prone road sections, such as near the cliff (ditch), near the water, intersections, bridges, tunnels, and width restrictions
• Lack of safety facilities	A defect in traffic signs, signals, lighting, or protective facilities, resulting in unsafe acts by drivers
• Poor road conditions	Slippery road caused by rain, snow, or ice, reducing the safety of driving
• Poor traffic conditions	Vehicles cannot pass normally due to construction or traffic accidents
◆ Unsafe acts by Drivers
▲ Errors
• Skill-based errors	Unsafe operating behavior of which the driver is rarely aware or unaware, usually related to inappropriate attention or memory loss, sometimes manifested as poor driving skills
• Decision-making errors	The operation is carried out as planned, but the plan is inadequate or did not apply to the situation
• Perceptual errors	Errors that occur when the driver’s sensory input does not match the real situation
▲ Violations
• Habitual violations	Intentional or habitual violation of the regulations governing road passenger transportation
• Incidental violations	Occasional violations of road passenger transport management regulations, non-habitual behavior
◆ Proximate cause other than the driver’s act
• Emergency situation related to vehicle parts	Partial system of the vehicle, such as braking, driving, or steering, breaks down in the process of driving
• Illegal transportation of dangerous goods	Overloading, illegal loading, or illegal transport of inflammable and explosive materials, as required by the owner of the vehicle
• Other vehicle ramming	The vehicle is hit by another out-of-control vehicle during normal travel
◆ Moderating factors affecting accident severity and probability
▲ Individual-level factors
• Poor safety awareness among passengers	Ignoring the rules of safe riding, disregarding the safety of vehicle loading, or lacking emergency self-rescue knowledge
• Violations by other road traffic participants	Other road traffic participants do not comply with traffic rules and affect the safe movement of the passenger vehicle
▲ Organizational-level factors
• Violations by other operational personnel	Vehicle investors violate road passenger transport business license or business management regulations; vehicle contract operators violate the provisions of road passenger transport regulations or contract contracts
• Inadequate road/bridge safety management	Road (bridge) access management, maintenance of the road (bridge), or the supervision of the road (bridge) construction project is not in place
• Insufficient safety management at passenger stations	Passenger station safety management, such as the inspection of vehicles entering and exiting the station or policy implementation, is not in place
• Inadequate safety measures in road construction	Road construction agency safety management, such as safety warning sign setting, greening changes, and hidden danger control, is not in place
• Errors/violations in vehicle maintenance	Vehicle maintenance errors and violations of the vehicle maintenance agencies’ regulations
• Violations in vehicle inspection	Irregularity in the passenger vehicle safety testing, such as reduced inspection, omission, and issuance of false inspection reports

Note: ◆/▲/• correspond to main, generic, and subcategories, respectively.

Table A4

Table of statistical results for all contributory factors.

Contributory Factors	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019
L1 Government regulations	14	22	32	43	42	30	28	41	12	11
X1 Road transport regulations	5	7	17	19	16	14	11	13	4	4
X1a Insufficient basic guarantees	0	0	0	1	0	1	1	0	0	0
X1b Insufficient supervision	3	4	10	8	7	6	3	6	1	1
X1c Regulation violations	1	0	3	3	1	2	2	3	1	1
X1d Ineffective implementation of policy	1	2	1	2	6	2	2	2	1	1
X1e Insufficient organizational synergy	0	1	3	5	2	3	3	2	1	1
X2 Road traffic safety management	7	10	12	16	18	13	11	19	5	5
X2a Lax enforcement of traffic laws	1	2	4	4	3	4	3	5	1	1
X2b Insufficient road access protection	1	2	1	2	2	1	0	0	0	0
X2c Inadequate safety education	2	1	3	1	2	3	3	4	1	1
X2d Inadequate inspection and supervision	3	3	4	3	5	4	3	4	1	1
X2e Ineffective implementation of policy	0	1	0	2	3	0	0	3	1	1
X2f Insufficient organizational synergy	0	1	0	4	3	1	2	3	1	1
X3 Territorial supervision by the local government	2	5	3	8	8	3	6	9	3	2
X3a Poor leadership and organization	1	2	2	4	4	1	2	4	1	1
X3b Inadequate inspection and supervision	1	3	1	4	4	1	2	3	1	1
X3c Regulation violations	0	0	0	1	1	0	0	0	0	0
X3d Ineffective implementation of policy	0	0	0	0	0	1	2	2	1	0
L2 Organizational influence of passenger transportation enterprises	24	11	44	61	48	33	24	37	8	10
X4 Basic safeguards	4	2	7	13	8	6	6	8	2	3
X4a Inadequate production safety management system	2	1	5	4	3	3	2	3	1	1
X4b Lack of safety managers	1	1	2	4	3	2	2	3	1	1
X4c Insufficient safety budgets	1	0	0	5	2	1	2	2	0	1
X5 Resource management	5	2	8	13	11	8	6	9	2	2
X5a Poor human resource management	3	1	3	7	5	4	3	5	1	1
X5b Inadequate facility resource management	2	1	5	6	6	4	3	4	1	1
X6 Organizational culture	11	7	24	24	21	14	9	17	3	3
X6a Inadequate implementation of policies and systems	4	2	8	8	7	4	3	5	1	1
X6b Incomplete management of hidden hazards	3	3	8	8	7	5	3	6	1	1
X6c Failure to fulfill responsibilities	4	2	8	8	7	5	3	6	1	1
X7 Operation methods	4	0	5	11	8	5	3	3	1	2
X7a Improper business operation	1	0	0	3	3	1	0	0	0	1
X7b Improper vehicle operation	3	0	5	8	5	4	3	3	1	1
L3 Unsafe internal operational supervision	11	6	32	39	29	18	15	22	5	4
X8 Insufficient supervision	10	5	22	24	19	13	9	14	3	3
X8a Poor management of drivers	4	2	7	8	6	5	3	6	1	1
X8b Poor vehicle management	3	1	7	8	7	4	3	4	1	1
X8c Dynamic monitoring in a virtual sense	3	2	8	8	6	4	3	4	1	1
X9 Supervision violations	1	1	10	15	10	5	6	8	2	1
X9a Non-compliant operation	1	0	3	7	3	1	3	3	1	1
X9b Non-compliance with requirements	0	1	7	8	7	4	3	5	1	0
L4 Preconditions for unsafe acts by drivers	15	18	32	44	29	30	8	19	3	3
X10 Personnel factors of driver	5	8	16	16	13	11	5	8	1	1
X10a Poor safety awareness	3	5	8	6	7	6	3	4	1	1
X10b Lack of knowledge and skills	2	3	5	5	3	4	0	2	0	0
X10c Inexperience	0	0	1	2	1	0	0	1	0	0
X10d Poor mental health	0	0	2	1	1	0	0	1	0	0
X10e Inattentiveness	0	0	0	2	1	0	1	0	0	0
X11 Substandard condition of vehicle parts	2	0	3	1	4	3	1	1	0	1
X12 Weather conditions	1	2	1	2	1	2	0	2	0	0
X13 Road environment	7	8	12	25	11	14	2	8	2	1
X13a Bad road alignment	2	1	3	4	3	2	0	2	0	0
X13b Bad road type	2	3	7	7	2	3	1	3	0	0
X13c Lack of safety facilities	2	2	0	6	1	3	0	0	1	1
X13d Poor road conditions	1	1	2	2	3	3	1	3	1	0
X13e Poor traffic conditions	0	1	0	6	2	3	0	3	0	0
L5 Unsafe acts by drivers	4	8	18	18	13	9	3	10	2	0
X14 Errors	2	5	10	10	9	4	0	6	1	0
X14a Skill-based errors	1	1	4	6	5	3	0	3	1	0
X14b Decision-making errors	0	4	5	2	4	1	0	3	0	0
X14c Perceptual errors	1	0	1	2	0	0	0	0	0	0
X15 Violations	2	3	8	8	4	5	3	4	1	0
X15a Habitual violations	2	1	7	7	4	5	3	4	1	0
X15b Incidental violations	0	2	1	1	0	0	0	0	0	0
L6 Proximate cause other than the driver’s act	3	6	3	3	3	2	2	3	1	1
X16 Emergency situation related to vehicle parts	0	1	1	0	0	1	0	0	0	1
X17 Illegal transportation of dangerous goods	2	1	1	2	1	1	2	0	1	0
X18 Other vehicle ramming	1	4	1	1	2	0	0	3	0	0
L7 Moderating factors affecting accident severity and probability	8	11	20	16	8	10	5	14	2	1
X19 Individual-level factors	2	6	10	4	4	3	1	4	1	1
X19a Poor safety awareness among passengers	1	4	7	2	1	2	1	0	1	1
X19b Violations by other road traffic participants	1	2	3	2	3	1	0	4	0	0
X20 Organizational-level factors	6	5	10	12	4	7	4	10	1	0
X20a Violations by other operational personnel	1	1	7	1	2	2	2	3	0	0
X20b Inadequate road/bridge safety management	2	3	1	6	0	2	0	4	1	0
X20c Insufficient safety management at passenger stations	2	0	0	0	0	0	1	1	0	0
X20d Inadequate safety measures in road construction	1	1	0	0	0	1	0	1	0	0
X20e Errors/violations in vehicle maintenance	0	0	1	5	0	0	1	1	0	0
X20f Violations in vehicle inspection	0	0	1	0	2	2	0	0	0	0
L0	79	82	181	224	172	127	79	138	31	29

Table A5

The analysis data for the reference sequence and comparative sequences.

	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019
$\mathrm{L}0$	79	82	181	224	172	132	85	146	33	30
$\mathrm{L}1$	14	22	32	43	42	30	28	41	12	11
$\mathrm{L}2$	24	11	44	61	48	33	24	37	8	10
$\mathrm{L}3$	11	6	32	39	29	18	15	22	5	4
$\mathrm{L}4$	15	18	32	44	29	30	8	19	3	3
$\mathrm{L}5$	4	8	18	18	13	9	3	10	2	0
$\mathrm{L}6$	3	6	3	3	3	2	2	3	1	1
$\mathrm{L}7$	8	11	20	16	8	10	5	14	2	1

Note: L0 equals the summation of L1–L7.

Table A6

Normalization results of the comparative sequences L1–L7 and reference sequence L0.

	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019
$\mathrm{L}{0}^{\prime }$	0.353	0.366	0.808	1.000	0.768	0.567	0.353	0.616	0.138	0.129
$\mathrm{L}{1}^{\prime }$	0.326	0.512	0.744	1.000	0.977	0.698	0.651	0.953	0.279	0.256
$\mathrm{L}{2}^{\prime }$	0.393	0.180	0.721	1.000	0.787	0.541	0.393	0.607	0.131	0.164
$\mathrm{L}{3}^{\prime }$	0.282	0.154	0.821	1.000	0.744	0.462	0.385	0.564	0.128	0.103
$\mathrm{L}{4}^{\prime }$	0.341	0.409	0.727	1.000	0.659	0.682	0.182	0.432	0.068	0.068
$\mathrm{L}{5}^{\prime }$	0.222	0.444	1.000	1.000	0.722	0.500	0.167	0.556	0.111	0.000
$\mathrm{L}{6}^{\prime }$	0.500	1.000	0.500	0.500	0.500	0.333	0.333	0.500	0.167	0.167
$\mathrm{L}{7}^{\prime }$	0.400	0.550	1.000	0.800	0.400	0.500	0.250	0.700	0.100	0.050

Table A7

The absolute differences between the comparison sequences and reference sequence.

AbsoluteDifference	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019
∆1	0.027	0.146	0.064	0.000	0.209	0.131	0.298	0.337	0.141	0.126
∆2	0.041	0.186	0.087	0.000	0.019	0.026	0.041	0.010	0.007	0.034
∆3	0.071	0.212	0.012	0.000	0.024	0.234	0.122	0.257	0.061	0.053
∆4	0.012	0.043	0.081	0.000	0.109	0.115	0.171	0.184	0.070	0.061
∆5	0.130	0.078	0.192	0.000	0.046	0.067	0.186	0.061	0.027	0.129
∆6	0.147	0.634	0.308	0.500	0.268	0.234	0.019	0.116	0.028	0.037
∆7	0.047	0.184	0.192	0.200	0.368	0.067	0.103	0.084	0.038	0.079

Note: According to the gray correlation coefficient calculation formula ${\mathsf{\xi }}_i\left(k\right)=\frac{\underset{i}{\min }({\Delta }_i\left(min\right))+\rho \underset{i}{\max }({\Delta }_i\left(max\right))}{\left|x_0\left(k\right)-x_i\left(k\right)\right|+\underset{i}{\max }({\Delta }_i\left(max\right))}$, we must first calculate $\left|x_0\left(k\right)-x_i\left(k\right)\right|$ and identify $\underset{i}{\min }\left({\Delta }_i\left(min\right)\right)\mathrm{and}\underset{i}{\max }\left({\Delta }_i\left(max\right)\right)$. From this table, we can see that $\underset{i}{\min }\left({\Delta }_i\left(min\right)\right)=0$, $\underset{i}{\max }\left({\Delta }_i\left(max\right)\right)$ = 0.634.

Table A8

The gray correlation coefficients between the comparison sequences and reference sequence.

$\mathit{r}$	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019
${\mathsf{\xi }}_1$	0.921	0.685	0.832	1.000	0.603	0.708	0.515	0.484	0.693	0.715
${\mathsf{\xi }}_2$	0.886	0.631	0.785	1.000	0.943	0.924	0.886	0.971	0.978	0.902
${\mathsf{\xi }}_3$	0.818	0.599	0.962	1.000	0.929	0.576	0.722	0.552	0.838	0.858
${\mathsf{\xi }}_4$	0.964	0.881	0.797	1.000	0.745	0.734	0.650	0.632	0.819	0.838
${\mathsf{\xi }}_5$	0.708	0.802	0.623	1.000	0.874	0.826	0.630	0.840	0.921	0.710
${\mathsf{\xi }}_6$	0.683	0.333	0.507	0.388	0.542	0.576	0.942	0.732	0.918	0.895
${\mathsf{\xi }}_7$	0.870	0.633	0.623	0.613	0.463	0.826	0.755	0.791	0.892	0.800

References

1. Ministry of Emergency Management of the People’s Republic of China. The Investigation Report of “9·28” Extraordinary Serious Road Traffic Accident on Changchun-Shenzhen Expressway in Wuxi, Jiangsu Province; Ministry of Emergency Management: Beijing, China, 2020; Available online: https://www.mem.gov.cn/gk/sgcc/tbzdsgdcbg/2020/202009/P020200911576236508884.pdf (accessed on 8 October 2020).

2. Cafiso, S.; Di Graziano, A.; Pappalardo, G. Using the Delphi method to evaluate opinions of public transport managers on bus safety. Saf. Sci.; 2013; 57, pp. 254-263. [DOI: https://dx.doi.org/10.1016/j.ssci.2013.03.001]

3. Bakhit, P.R.; Guo, B.B.; Ishak, S. Crash and Near-Crash Risk Assessment of Distracted Driving and Engagement in Secondary Tasks: A Naturalistic Driving Study. Transport. Res. Rec.; 2018; 2672, pp. 245-254. [DOI: https://dx.doi.org/10.1177/0361198118772703]

4. Sajid, F.; Javed, A.R.; Basharat, A.; Kryvinska, N.; Afzal, A.; Rizwan, M. An Efficient Deep Learning Framework for Distracted Driver Detection. IEEE Access; 2021; 9, pp. 169270-169280. [DOI: https://dx.doi.org/10.1109/ACCESS.2021.3138137]

5. Vennelle, M.; Engleman, H.M.; Douglas, N.J. Sleepiness and sleep-related accidents in commercial bus drivers. Sleep Breath.; 2010; 14, pp. 39-42. [DOI: https://dx.doi.org/10.1007/s11325-009-0277-z] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/19588178]

6. Kim, H.; Jang, T.W.; Kim, H.R.; Lee, S. Evaluation for Fatigue and Accident Risk of Korean Commercial Bus Drivers. Tohoku J. Exp. Med.; 2018; 246, pp. 191-197. [DOI: https://dx.doi.org/10.1620/tjem.246.191] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30487372]

7. Chung, W.Y.; Chong, T.W.; Lee, B.G. Methods to Detect and Reduce Driver Stress: A Review. Int. J. Automot. Technol.; 2019; 20, pp. 1051-1063. [DOI: https://dx.doi.org/10.1007/s12239-019-0099-3]

8. Papalimperi, A.H.; Athanaselis, S.A.; Papadodima, S.A. Incidence of fatalities of road traffic accidents associated with alcohol consumption and the use of psychoactive drugs: A 7-year survey (2011–2017). Exp. Ther. Med.; 2019; 18, pp. 2299-2306. [DOI: https://dx.doi.org/10.3892/etm.2019.7787]

9. McCollum, E. Effectiveness of an intervention to reduce accidents in bus drivers. Occup. Med.; 2022; 72, pp. 252-254. [DOI: https://dx.doi.org/10.1093/occmed/kqab185]

10. Tseng, C.M. Social-demographics, driving experience and yearly driving distance in relation to a tour bus driver’s at-fault accident risk. Tour. Manag.; 2012; 33, pp. 910-915. [DOI: https://dx.doi.org/10.1016/j.tourman.2011.09.011]

11. Klinjun, N.; Kelly, M.; Praditsathaporn, C.; Petsirasan, R. Identification of Factors Affecting Road Traffic Injuries Incidence and Severity in Southern Thailand Based on Accident Investigation Reports. Sustainability; 2021; 13, 12467. [DOI: https://dx.doi.org/10.3390/su132212467]

12. Zhou, C.Y.; Zhao, X.; Huang, R. Dynamic bayesian network-based escape probability estimation for coach fire accidents. Promet-TrafficTransportation; 2021; 33, pp. 193-204. [DOI: https://dx.doi.org/10.7307/ptt.v33i2.3537]

13. Gamero, N.; Silla, I.; Sainz-Gonzalez, R.; Sora, B. The Influence of Organizational Factors on Road Transport Safety. Int. J. Environ. Res. Public Health; 2018; 15, 1938. [DOI: https://dx.doi.org/10.3390/ijerph15091938] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30200560]

14. Naevestad, T.O.; Phillips, R.O.; Laiou, A.; Bjørnskau, T.; Yannis, G. Safety culture among bus drivers in Norway and Greece. Transp. Res. Part F; 2019; 64, pp. 323-341. [DOI: https://dx.doi.org/10.1016/j.trf.2019.05.006]

15. Mokarami, H.; Alizadeh, S.S.; Pordanjani, T.R.; Varmazyar, S. The relationship between organizational safety culture and unsafe behaviors, and accidents among public transport bus drivers using structural equation modeling. Transp. Res. Part F; 2019; 65, pp. 46-55. [DOI: https://dx.doi.org/10.1016/j.trf.2019.07.008]

16. Wang, S.Y.; Wu, K.F. Reducing intercity bus crashes through driver rescheduling. Accid. Anal. Prev.; 2019; 122, pp. 25-35. [DOI: https://dx.doi.org/10.1016/j.aap.2018.09.019]

17. Wu, X.L.; Zhang, H.M.; Xiao, W.X.; Ning, P.S.; Schwebel, D.C.; Hu, G.Q. Are bus company regulations associated with crash risk? Findings from a retrospective survey in four Chinese cities. Int. J. Environ. Res. Public Health; 2019; 16, 1342. [DOI: https://dx.doi.org/10.3390/ijerph16081342]

18. La, Q.N.; Duong, D.V.; Lee, A.H.; Meuleners, L.B. Factors underlying bus-related crashes in Hanoi, Vietnam. Transp. Res. Part F; 2017; 46, pp. 426-437. [DOI: https://dx.doi.org/10.1016/j.trf.2016.06.023]

19. Ali, Q.; Yaseen, M.R.; Khan, M.T.I. Road traffic fatalities and its determinants in high-income countries: A continent-wise comparison. Environ. Sci. Pollut. Res.; 2019; 26, pp. 19915-19929. [DOI: https://dx.doi.org/10.1007/s11356-019-05410-9]

20. Obregon-Biosca, S.A.; Betanzo-Quezada, E.; Romero-Navarrete, J.A.; Ríos-Nuñez, M. Rating road traffic education. Transp. Res. Part F; 2018; 56, pp. 33-45. [DOI: https://dx.doi.org/10.1016/j.trf.2018.03.033]

21. Chu, H.C. Assessing factors causing severe injuries in crashes of high-deck buses in long-distance driving on freeways. Accid. Anal. Prev.; 2014; 62, pp. 13-136. [DOI: https://dx.doi.org/10.1016/j.aap.2013.09.016]

22. Yusof, M.b.; bin Afripin, M.A.A. Effect of Beam Profile Size on Bus Superstructure Strength Having Rollover Crash. Appl. Mech. Mater.; 2013; 372, pp. 620-629. [DOI: https://dx.doi.org/10.4028/www.scientific.net/AMM.372.620]

23. Xu, A.; Qiao, X.M. Research on traffic accident and braking safe performances of inter-urban coach. Proceedings of the The International Symposium on Safety Science and Technology; Beijing, China, 24–27 September 2008.

24. Hammad, H.M.; Ashraf, M.; Abbas, F.; Bakhat, H.F.; Qaisrani, S.A.; Mubeen, M.; Fahad, S.; Awais, M. Environmental factors affecting the frequency of road traffic accidents: A case study of sub-urban area of Pakistan. Environ. Sci. Pollut. Res.; 2019; 26, pp. 5674-5685. [DOI: https://dx.doi.org/10.1007/s11356-019-04752-8] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30888616]

25. Huang, H.; Wang, J.; Hou, M.; Tang, X.; Cheng, J.C. Analysis of Wind as a Contributing Factor in Bus Accidents in Rolling Terrain. Proceedings of the 19th COTA International Conference of Transportation Professionals; Nanjing, China, 6–8 July 2019; [DOI: https://dx.doi.org/10.1061/9780784482292.311]

26. Karlaftis, M.G.; Golias, I. Effects of road geometry and traffic volumes on rural roadway accident rates. Accid. Anal. Prev.; 2002; 34, pp. 357-365. [DOI: https://dx.doi.org/10.1016/S0001-4575(01)00033-1]

27. Xu, X.C.; Saric, Z.; Zhu, F.; Babic, D. Accident severity levels and traffic signs interactions in state roads: A seemingly unrelated regression model in unbalanced panel data approach. Accid. Anal. Prev.; 2018; 120, pp. 122-129. [DOI: https://dx.doi.org/10.1016/j.aap.2018.07.037] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30107331]

28. Li, X.L.; Kuang, H.; Fan, Y.H.; Zhang, G.X. Traffic accidents on a single-lane road with multi-slowdown sections. Int. J. Mod. Phys. C.; 2014; 25, 1450036. [DOI: https://dx.doi.org/10.1142/S0129183114500363]

29. Shappell, S.A.; Wiegmann, D.A. A Human Error Approach to Accident Investigation: The Taxonomy of Unsafe Operations. Int. J. Aviat. Psychol.; 1997; 7, pp. 269-291. [DOI: https://dx.doi.org/10.1207/s15327108ijap0704_2]

30. Jiang, W.; Han, W. Analysis of “2·28” KEEPER Chemical Industries Hazardous Chemical Explosion Accident Based on FTA and HFACS. Int. J. Environ. Res. Public Health; 2018; 15, 2151. [DOI: https://dx.doi.org/10.3390/ijerph15102151]

31. Umut, Y.; Ersan, B.; Özkan, U. Assessment of collisions and grounding accidents with human factors analysis and classification system (HFACS) and statistical methods. Saf. Sci.; 2019; 119, pp. 412-425. [DOI: https://dx.doi.org/10.1016/j.ssci.2017.09.022]

32. Zhan, Q.; Zheng, W.; Zhao, B. A hybrid human and organizational analysis method for railway accidents based on HFACS-Railway Accidents (HFACS-RAs). Saf. Sci.; 2017; 91, pp. 232-250. [DOI: https://dx.doi.org/10.1016/j.ssci.2016.08.017]

33. Ugurlu, O.; Yildiz, S.; Loughney, S.; Wang, J. Modified human factor analysis and classification system for passenger vessel accidents (HFACS-PV). Ocean Eng.; 2018; 161, pp. 47-61. [DOI: https://dx.doi.org/10.1016/j.oceaneng.2018.04.086]

34. Chen, S.T. An approach of identifying the common human and organisational factors (HOFs) among a group of marine accidents using GRA and HFACS-MA. J. Transp. Saf. Secur.; 2020; 12, pp. 1252-1294. [DOI: https://dx.doi.org/10.1080/19439962.2019.1583297]

35. Lin, Z.; Fu, G.; Xue, Y.J.Y. Human and organizational factors in Chinese hazardous chemical accidents: A case study of ‘8.12’ Tianjin Port fire and explosion using the HFACS-HC. Int. J. Occup. Saf. Ergon.; 2018; 24, pp. 329-340. [DOI: https://dx.doi.org/10.1080/10803548.2017.1372943]

36. Nwankwo, C.D.; Arewa, A.O.; Theophilus, S.C.; Esenowo, V.N. Analysis of accidents caused by human factors in the oil and gas industry using the HFACS-OGI framework. Int. J. Occup. Saf. Ergon.; 2022; 28, pp. 1642-1654. [DOI: https://dx.doi.org/10.1080/10803548.2021.1916238]

37. Ye, G.; Tan, Q.; Gong, X.L.; Xiang, Q.T.; Wang, Y.H.; Liu, Q.J. Improved HFACS on Human Factors of Construction Accidents: A China Perspective. Adv. Civ. Eng.; 2018; 2018, 4398345. [DOI: https://dx.doi.org/10.1155/2018/4398345]

38. Zhang, Y.; Jing, L.; Sun, C.; Fang, J.L.; Feng, Y. Human factors related to major road traffic accidents in China. Traffic Inj. Prev.; 2019; 20, pp. 796-800. [DOI: https://dx.doi.org/10.1080/15389588.2019.1670817]

39. Zhang, X.Y.; Hu, W.W.; Zhou, J.L.; Zuo, Q.J.; Wu, R.L.; Tang, Z.X. The Failure Route between Active and Latent Error in Bus Accident. IEEE Access; 2019; 9, pp. 164941-164951. [DOI: https://dx.doi.org/10.1109/ACCESS.2019.2949858]

40. The State Council of the People’s Republic of China. Regulations on Investigation and Handling of Production Safety Accidents; State Council: Beijing, China, 2007; Available online: http://www.gov.cn/flfg/2007-04/19/content_589264.htm (accessed on 12 October 2020).

41. Schreier, M. Qualitative Content Analysis in Practice; Sage: London, UK, 2012; pp. 230-240.

42. Krippendorff, K. Content Analysis: An Introduction to Its Methodology; Sage: Thousand Oaks, CA, USA, 1980; pp. 29-31.

43. Elo, S.; Kyngäs, H. The qualitative content analysis process. J. Adv. Nurs.; 2010; 62, pp. 107-115. [DOI: https://dx.doi.org/10.1111/j.1365-2648.2007.04569.x] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/18352969]

44. Glaser, B.; Strauss, A. The Discovery of Grounded Theory: Strategies for Qualitative Research; Aldine: Chicago, IL, USA, 1967; pp. 56-70.

45. Strauss, A.; Corbin, J. Basics of Qualitative Research: Grounded Theory Techniques and Procedures; Sage: Newbury Parks, CA, USA, 1990; pp. 61-142.

46. Akyuz, E. A marine accident analysing model to evaluate potential operational causes in cargo ships. Saf. Sci.; 2017; 92, pp. 17-25. [DOI: https://dx.doi.org/10.1016/j.ssci.2016.09.010]

47. Jing, L.; Bai, Q.; Guo, W.; Yan Feng, Y.; Liu, L.; Zhang, Y.Y. Contributory factors interactions model: A new systems-based accident model. Syst. Res. Behav. Sci.; 2019; 37, pp. 255-276. [DOI: https://dx.doi.org/10.1002/sres.2618]

48. Wang, X.L.; Li, X.X.; Guan, W.L.; Peng, S.T. Study on Wharf Accidents Based on the Grey Cor-relation Analysis. Appl. Mech. Mater.; 2014; 505–506, pp. 683-688. [DOI: https://dx.doi.org/10.4028/www.scientific.net/AMM.505-506.683]

49. Guo, Y.; Sun, Y.; Chen, S. Research on Human-Error Factors of Civil Aircraft Pilots Based on Grey Relational Analysis. MATEC Web Conf.; 2018; 151, 05005. [DOI: https://dx.doi.org/10.1051/matecconf/201815105005]

50. Leveson, N. Engineering a Safer World: Systems Thinking Applied to Safety; MIT Press: Cambridge, MA, USA, 2011.

51. Choi, Y.; Park, J.H.; Jang, B. Developing safety checklists for predicting accidents. Proceedings of the International Conference on Information and Communication Technology Convergence (ICTC); Jeju Island, Republic of Korea, 17–19 October 2018; [DOI: https://dx.doi.org/10.1109/ICTC.2018.8539652]

52. Ministry of Transport of the People’s Republic of China. Regulations on the Safety Management of Road Passenger Transport Companies; Ministry of Transport: Beijing, China, 2018; Available online: https://xxgk.mot.gov.cn/2020/jigou/ysfws/202006/t20200623_3315491.html (accessed on 12 October 2020).

53. Jackson, M.C. Critical Systems Thinking and the Management of Complexity; John Wiley & Sons Ltd.: West Sussex, UK, 2019; pp. 261-289.