Plane Thermoelastic Waves in Infinite Half-Space Caused FACTA UNIVERSITATIS Series: Mechanical Engineering Vol. 22, No 1, 2024, pp. 139 - 152 https://doi.org/10.22190/FUME230726042X © 2024 by University of Niš, Serbia | Creative Commons License: CC BY-NC-ND Original scientific paper INTERACTIONS AND INFLUENCES ON COAL MINERS' SAFETY ATTENTION: AN EVALUATION USING IMPROVED DEMATEL-ISM De-Yu Xu1,2, Hong-Qing Cui2,3, Tian-Yu Wang1, Jiang-Wei Fu1, Guang-Bo Zhu4 1School of Energy & Environment Engineering, Zhongyuan University of Technology, China 2School of Safety Science and Engineering, Henan Polytechnic University, China 3Central Plains Economic Region Coalbed (Shale) Gas Innovation Center, China 4FAW-Volkswagen Co., Ltd. Qingdao Branch, China ORCID iDs: De-Yu Xu https://orcid.org/0009-0005-0826-6692 Hong-Qing Cui https://orcid.org/0000-0002-4282-2614 Tian-Yu Wang https://orcid.org/0009-0009-4508-2859 Jiang-Wei Fu https://orcid.org/0009-0003-4389-8121 Guang-Bo Zhu https://orcid.org/0009-0007-4480-056X Abstract. In coal mining, the myriad of factors influencing miners' attention to safety necessitates deeper exploration. Particularly, discerning the significance and interplay of these factors offers crucial insights into the actual disparities in miners' safety attentiveness. Yet, a limited number of comprehensive studies address this dimension. Thus, an advanced Decision Making Trial Evaluation Laboratory-Interpretive Structural Model (DEMATEL-ISM) has been employed to probe the determinants impacting coal miners' safety focus and the mechanisms underpinning these interactions. The objective is to provide strategies that could diminish the occurrence of minor accidents. Results revealed that there are 9 causative factors and 6 resultant factors shaping the coal miners' attention to safety. Within the structural model of these factors, three layers and seven levels were identified. Notably, the intricacy of relationships among these factors was found to be profound. Emphasis is recommended on the management of these intricate deep-level causative factors boasting high driving power, and mid-level resultant factors characterized by both substantial driving force and dependence. Key words: Minor injury accidents, Coal miners, Safety attention, Influencing determinants Received: July 26, 2023 / Accepted October 30, 2023 Corresponding author: De-Yu Xu School of Energy & Environment Engineering, Zhongyuan University of Technology, Zhengzhou 451191, China E-mail: 6675@zut.edu.cn https://orcid.org/0009-0005-0826-6692 https://orcid.org/0000-0002-4282-2614 https://orcid.org/0009-0009-4508-2859 https://orcid.org/0009-0003-4389-8121 https://orcid.org/0009-0007-4480-056X 140 D.Y. XU, H.Q. CUI, T.Y. WANG, J.W. FU, G.B. ZHU 1. INTRODUCTION In recent years, a steady improvement in coal mine safety has been observed in China, attributed largely to consistent investments and enhancements in safety technologies and equipment within coal mine enterprises. This enhancement has resulted in a notable decline in major accidents. Yet, minor injury accidents, accounting for >90% of all accidents [1], persist as a prevalent concern. Such accidents, apart from posing a significant threat to miner safety, also jeopardize the economic well-being and production of coal mines. Mainly attributed to miners’ unsafe behaviors, minor injury accidents have been found to predominantly stem from a decline in coal miners' attention to safety [2-4]. Emphasis on the factors that influence coal miners' attention to safety is underscored by its potential in mitigating these unsafe behaviors and consequently reducing minor injury accidents. Several scholars have diligently studied these influences. It has been frequently postulated that a decrease in coal miners' attention to safety contributes to human error, and a strong correlation exists between safety attention and unsafe behaviors [5,6]. Recent studies have highlighted relationships between individual characteristic factors and coal miners' safety attention [7-9]. Mathematical methods have also been employed to elucidate the determinants of miners’ safety attention [10,11]. The prevailing research predominantly focuses on the influence mechanism of various factors on coal miners' safety attention [12-14]. It is well-acknowledged that diverse factors impact miners' safety attention [15-17]. A comprehensive understanding of the significance and interplay of these factors provides insights into actual disparities in safety attention among miners. Nevertheless, a scarcity of in-depth studies on this crucial aspect limits the fulfillment of scientific requirements for coal mine safety management. The resulting challenge for practical management is the vast array of influencing factors and the evolving nature of safety information. These complexities make comprehensive control countermeasures elusive, potentially overlooking pivotal factors requiring attention in safety management. The combined usage of Decision Making Trial Evaluation Laboratory (DEMATEL) and Interpretive Structural Modeling (ISM) offers a framework to discern the importance and causal attributes of system factors, thereby shedding light on the intricate interactions between them [18-20]. Moreover, the Matrix Impacts Cross-reference Multiplication Applied to a Classification (MICMAC) delves deeper, analyzing the driving force and dependencies among system factors [21, 22]. Given their unique characteristics, an integration of the MICMAC method with DEMATEL-ISM has been proposed. Such a methodological convergence can effectively rank each influencing factor, elucidate the multi-layered interpretation structure of the system, and through rigorous analysis of driving forces and dependencies, unveil the interplay mechanism among these factors. Employing this improved approach can streamline the overwhelming workload in minor injury accident safety management and mitigate the potential oversight of key determinants. Ultimately, the aim is to enhance safety management efficiency, providing mining companies with a solid foundation to address strategic challenges in minor injury accidents by bolstering miners’ safety attention. Interactions and Influences on Coal Miners' Safety Attention: An Evaluation Using Improved... 141 2. IMPROVED DEMATEL-ISM METHOD 2.1. Value assignment of influencing factors The assignment of values to influencing factors predominantly relies on expert evaluations, fundamentally characterized by a systematic aggregation of expert opinions. Subsequent processing of their responses culminates in the consolidation of results. A questionnaire-based approach was adopted for this expert evaluation. Twenty experts, affiliated with three distinct research institutions in Henan province and specializing in the study of coal miners’ safety behavior, were solicited for their insights. Drawing from an array of studies and accident case analyses [7-11], these questionnaires distilled and highlighted 15 pivotal factors impacting miners' safety attention. These factors were compartmentalized into three distinct categories, as illustrated in Table 1: Individual factors, environmental factors, and management factors. Leveraging the index system for these factors, experts were prompted to designate values to the interaction intensity among these 15 determinants, rooted in their professional acumen and prior experiences. To counteract potential individual biases in the evaluations, mean values were computed for each factor. Table 1 The index system of influencing factors pertinent to coal miners' safety attention Classification Influencing factors Individual factors Safety cognition S1, Safety consciousness S2, Behavioral habit S3, Safety self-control S4, Risk perception S5, Educational level S6, Working years S7, Work pressure S8, Job burnout S9 Environmental factors Working environment S10, Working strength S11, Safety information stimulus S12 Management factors Safety supervision S13, Safety education and training S14, Safety atmosphere of the organization S15 The results from the value assignments were manifested as a matrix, denoted as the direct influence matrix A. This matrix, represented as A=(aij)n×n, is delineated in Eq. (1). Within this matrix, each constituent element aij signifies the potency of influence exerted by factor Si on factor Sj. The guiding metric for experts to designate these values is further elucidated in Eq. (1). Consequently, upon assimilating the experts' value designations, the direct influence matrix A was derived, as detailed in Eq. (2). 0 has no influence on 1 has a weak influence on 2 has a moderate influence on 3 has a strong influence on i j i j ij i j i j S S S S a S S S S    =     (1) where, when i = j, aij = 0. 142 D.Y. XU, H.Q. CUI, T.Y. WANG, J.W. FU, G.B. ZHU 0 3 3 3 3 0 0 0 1 0 0 0 0 0 0 0 0 3 3 3 0 0 0 2 0 0 0 0 0 0 0 2 0 2 2 0 0 0 0 0 0 0 0 0 0 1 2 2 0 2 0 0 0 2 0 0 0 0 0 0 0 2 2 2 0 0 0 0 2 0 0 0 0 0 0 3 3 3 3 3 0 0 0 2 0 0 0 0 0 0 3 3 3 3 3 0 0 0 2 0 0 0 0 0 0 3 3 3 3 3 0 0 0 3 0 0 0 0 0 0 3 3 3 3 3 0 0 0 0 0 0 0 0 0 0 2 2 2 2 2 0 0 3 3 0 0 0 0 0 0 2 2 2 2 2 0 0 3 3 0 0 0 0 0 0 3 3 3 3 3 0 0 2 3 0 0 0 0 0 0 3 3 3 3 3 0 0 2 3 0 0 3 0 0 0 3 3 3 A = 3 3 0 0 1 1 0 0 0 0 0 0 3 3 3 3 2 0 0 2 2 0 0 3 3 3 0                                 (2) 2.2. Enhanced DEMATEL-ISM analysis technique Post the expert-driven value assignment of influential determinants, a pressing need emerges to determine both the centrality and cause degrees of each constituent factor using the refined DEMATEL-ISM analysis procedure. In initial phases, the method seeks to elucidate the significance and causative characteristics of each factor. Successive stages demand the partitioning of system levels founded upon the centric and causative degrees of each variable. Consequently, a multi-layered, recursive interpretative structure model is crafted to decipher intricate interactions between said factors. Ultimately, the analytical procedure is harnessed to discern the drive and dependency values of each factor. An in-depth analysis of these values unravels the intricate interplay mechanisms among them. The ensuing subsections provide an exhaustive breakdown of the refined DEMATEL-ISM technique. 2.2.1. Determining centrality and cause degrees The pivotal initiation step in the refined DEMATEL-ISM analysis method hinges on the calculations of the centrality and cause degrees. These degrees offer preliminary insights into the relative significance of each influencing factor. The methodical steps are delineated below: (1) System components are identified, denoted as S1-Sn, wherein S1-Sn pertain to the factors enumerated in Table 1. (2) The direct influence matrix A is constructed. Guided by expert insight and acumen, the degree of influence exerted by factor Si on factor Sj is assessed, resulting in the formation of the direct influence matrix A=(aij)n×n, as given in Eq. (2). (3) The normalized direct influence matrix C, represented as C=(cij)n×n, is computed as follows: 1 max( ) ij ij n ij j a c a = =  (3) where, aij is an element of matrix A. Interactions and Influences on Coal Miners' Safety Attention: An Evaluation Using Improved... 143 (4) The comprehensive influence matrix T is deduced, with the matrix presented as T=(tij)n×n: 1( )T C I C −= − (4) Herein, C stands for the normalized direct influence matrix, and I symbolizes the unit matrix. The unit matrix I is characterized by diagonal elements equating to 1, with the remaining being 0. (5) For each factor, both the influential degree Fi and affected degree Ei are calculated. Fi, representing the aggregate value of row i in T, mirrors the encompassing value of the influence factor Si on its counterparts. Conversely, Ei, reflecting the summation of the value of column i in T, signifies the aggregate influence borne by factor Si due to other factors. Eqs. (5) and (6) represent Fi and Ei respectively: 1 , ( 1,2, , ) n i ij j F t i n = = = (5) 1 , ( 1,2, , ) n i ji j E t i n = = = (6) where, tij is an element of the matrix T. (6) Subsequently, for each factor, the centrality degree Mi and the cause degree Ni are computed. Mi, the cumulative value of both influential and affected degrees for each factor, pinpoints the prominence of factor Si within the system. Ni, the differential between the influential and affected degrees, sheds light on the causality dynamics among factors. Eqs. (7) and (8) give Mi and Ni, respectively: i i iM F E= + (7) i i iN F E= − (8) where, a positive Ni denotes Si as a causative factor, while a negative Ni classifies Si as a resultant factor. 2.2.2. Formulation of the multi-layer recursive interpretation structure model In the subsequent phase of the enhanced DEMATEL-ISM analysis method, the focus shifts to the conceptualization of a multi-layer recursive interpretation structure model. The processes entailed in this establishment are elucidated below. (1) The global influence matrix H is computed. This matrix, represented as H=(hij)n×n, is derived by incorporating the comprehensive influence matrix T with the unit matrix I. (2) The threshold value λ, essential for discerning the reachable matrix K, is derived from the comprehensive influence matrix T. Following this determination, elements within the global influence matrix H undergo processing, leading to the acquisition of the reachable matrix K=(kij)n×n. The component Kij is given in Eq. (9): 1 0 ij ij ij h k h   =   ＜ (9) 144 D.Y. XU, H.Q. CUI, T.Y. WANG, J.W. FU, G.B. ZHU In this context, the threshold value λ is expressed as the sum of α and β, where α and β respectively denote the mean and standard deviations of the elements encapsulated in T. (3) Stratification of the reachable matrix K is performed. From the matrix, both the reachable set R(Si) and the antecedent set A(Si) for factor Si are extracted. Here, R(Si) represents {Sj|Sj∈S, kij=1} and A(Si) denotes {Si|Si∈S, kji=1}. Subsequently, the stratification aligns with Eq. (10). The initial tier of factors is discerned, followed by the elimination of its corresponding row and column from K, aiding in the identification of the secondary tier of factors. This procedure is replicated iteratively until all system levels are demarcated. ( ) ( ) ( ),( 1,2, , )i i iR S A S R S i n= = (10) (4) Depiction of the interconnected framework of influencing factors across diverse tiers is executed. This is premised on the prior stratification of the system, culminating in the illustration of the multi-layer recursive interpretation structure model for the system. 2.2.3. Determination of driving force and dependence degrees In the progression of the advanced DEMATEL-ISM method, an essential step involves quantifying the driving force and dependence degrees of each factor. The subsequent delineation outlines the processes involved. The driving force degree, represented as Pi, and the dependence degree, denoted as Qj, for every factor are calculated. The quadrant classification diagram, a critical visual representation, is then constructed based on these values. Derived from the reachable matrix K, the expressions for Pi and Qj are given by Eqs. (11) and (12), respectively: 1 ( 1,2, , ) n i ij j P K i n = = = (11) 1 ( 1,2, , ) n j ij i Q K j n = = = (12) Within this context, Kij is recognized as an element of matrix K. The promotional degree of factor Si influencing other factors is signified by Si, whereas the degree to which factor Sj is influenced by other factors is represented by Qj. In the quadrant diagram, the positions of factor Si on the abscissa and ordinate are represented by Pi and Qj, respectively. The quadrant diagram, an essential analytical tool, is partitioned into four distinct quadrants. Each quadrant epitomizes a unique cluster: the autonomous cluster (Quadrant I), the independent cluster (Quadrant II), the connection cluster (Quadrant III), and the dependent cluster (Quadrant IV). 3. ANALYSIS OF RESULTS DERIVED FROM THE ENHANCED DEMATEL-ISM TECHNIQUE 3.1. Centrality and cause degrees Upon obtaining the direct influence matrix A, normalization was performed in accordance with Eq. (3). Subsequently, the comprehensive influence matrix T was deduced as per Eq. (4). With reference to the comprehensive influence matrix T, various degrees – namely the influential degree Fi, the affected degree Ei, the centrality degree Mi, and the cause degree Ni Interactions and Influences on Coal Miners' Safety Attention: An Evaluation Using Improved... 145 – associated with each influencing factor were computed, as detailed in Eqs. (7) and (8). Table 2 provides the calculated results concerning the centrality and cause degrees. Further elucidation on the causal relationships among the factors shaping miners’ safety attention can be discerned from Fig. 1. Notably, these results reveal a distinction between 9 causative factors and 6 resultant factors. Moreover, an elevated centrality degree is observed among the resultant factors. Table 2 Calculation results of centrality degree and cause degree Factors Fi Ei Mi Ni Sort by Mi Type S1 0.571 1.197 1.768 -0.627 6 Resultant factor S2 0.484 1.743 2.227 -1.259 1 Resultant factor S3 0.265 1.812 2.077 -1.546 4 Resultant factor S4 0.402 1.812 2.214 -1.409 2 Resultant factor S5 0.358 1.774 2.132 -1.417 3 Resultant factor S6 0.758 0.000 0.758 0.758 12 Causative factor S7 0.758 0.000 0.758 0.758 13 Causative factor S8 0.806 0.482 1.288 0.323 8 Causative factor S9 0.661 1.333 1.994 -0.672 5 Resultant factor S10 0.740 0.000 0.740 0.740 14 Causative factor S11 0.740 0.000 0.740 0.740 15 Causative factor S12 0.909 0.222 1.131 0.687 10 Causative factor S13 1.073 0.111 1.184 0.962 9 Causative factor S14 0.761 0.111 0.872 0.650 11 Causative factor S15 1.312 0.000 1.312 1.312 7 Causative factor Fig. 1 Causal relationship of influencing factors 146 D.Y. XU, H.Q. CUI, T.Y. WANG, J.W. FU, G.B. ZHU 3.2. Recursive structural interpretation of influencing factors The global influence matrix H was derived by integrating the unit matrix I with the comprehensive influence matrix T. With the introduction of the threshold value λ, elements within H were adjusted following Eq. (9). Consequently, the reachable matrix K, characterized in Eq. (13), was formulated. The threshold λ was discerned from the summation of the average value α and the standard deviation β stemming from the elements in matrix T, yielding a computed threshold value of 0.107. 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 1 1 1 1 1 0 1 0 0 0 0 0 0 0 0 1 1 1 1 1 0 0 1 1 0 0 0 0 0 0 1 1 1 1 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 1 0 0 0 0 1 1 1 1 1 0 0 0 1 0 0 1 0 0 0 1 1 1 1 1 0 0 0 1 0 0 1 1 0 0 1 1 1 K = 1 1 0 0 0 0 0 0 0 0 1 0 1 1 1 1 1 0 0 0 1 0 0 1 1 1 1                                 (13) With the aid of the reachable matrix K, both the reachable set R(Si) and the antecedent set A(Si) pertaining to factor Si were identified. Factors of the primary level were extracted based on Eq. (10). By eliminating corresponding rows and columns of level 1 factors within K, the level 2 factors were discerned. This methodological approach facilitated the classification of influencing factors into seven distinct levels, as cataloged in Table 3. Table 3 Results of division of levels of the influencing factors Level division Element set Level specification L1 S3, S4, S5 Surface-level direct factor L2 S2 Middle-level indirect factor L3 S1 Middle-level indirect factor L4 S6, S7, S9, S14 Middle-level indirect factor L5 S8, S12 Deep-level fundamental factor L6 S10, S11, S13 Deep-level fundamental factor L7 S15 Deep-level root factor Fig. 2 visually represents the multi-layer recursive interpretation structure delineating the factors influencing miners' attention to safety. An in-depth examination of the constructed model indicates a three-tiered structure comprising seven levels. Furthermore, the presence of 12 intermediate and profound factors is evident, accounting for a significant 80% of the total influential factors. The intricate internal relationships and dynamic mechanisms linking these deep-level factors underscore the complexity inherent to this domain. Interactions and Influences on Coal Miners' Safety Attention: An Evaluation Using Improved... 147 Fig. 2 Multi-layer recursive interpretation structure of influencing factors 3.3. Driving force and dependence degrees of influencing factors Drawing from the reachability matrix K, calculations for the driving force and dependence degrees of each influencing factor were meticulously executed, as delineated in Eqs. (11) and (12). Table 4 presents the computed values for both driving force and dependence degrees. The quadrant distribution map, showcased in Fig. 3, plots the influencing factors based on their calculated degrees. Table 4 Driving force values and dependence values of influencing factors Influencing factor Driving force Dependence S1 5 9 S2 4 10 S3 1 11 S4 1 11 S5 1 11 S6 6 1 S7 6 1 S8 7 3 S9 6 7 S10 3 1 S11 3 1 S12 7 3 S13 8 2 S14 6 2 S15 10 1 148 D.Y. XU, H.Q. CUI, T.Y. WANG, J.W. FU, G.B. ZHU Fig. 3 Classification of driving force and dependence of influencing factors The spatial distribution of influencing factors within the quadrant map conveys compelling insights. A striking 73% of the total influencing factors, amounting to 11 distinct factors, are situated within quadrants II and IV. Notably, these factors exhibit pronounced values in either driving force or dependence degrees. A significant observation concerns the positioning of safety cognition S1 and job burnout S9 within quadrant III, suggesting a pronounced concurrent driving force and dependence for these two factors. Further research could delve deeper into the underlying reasons for such classifications, offering a more nuanced understanding of the roles and interactions of these influencing factors within the mining safety landscape. This could pave the way for more targeted interventions and strategic policies to enhance miners' safety awareness and practices. 4. DISCUSSION 4.1. Assessment of influencing factors’ importance degrees In the system depicted in Table 2, the centrality degree serves as an indicator of the relative importance of the influencing factors. From the data, a specific hierarchical sequence regarding the factors' influence on miners' safety attention is discerned: S2 >S4 >S5 >S3 >S9 >S1 >S15 >S8 >S13 >S12 >S14 >S6=S7 >S10=S11. This hierarchy can be compartmentalized into three distinct levels. Notably, factors such as safety consciousness S2, safety self-control S4, risk perception S5, behavioral habits S3, job burnout S9, and safety cognition S1 are classified under the first tier. As per Fig. 1, the centrality degrees for these factors surpass 1.5, signifying their heightened importance within the system. These factors are thus inferred to exert a substantial impact on miners' attention to safety. The second echelon encompasses the safety atmosphere of the organization S15, work pressure S8, safety supervision S13, and safety information stimulation S12. These factors exhibit centrality degrees ranging between 1 and 1.5, denoting a medium level of systemic importance. Interactions and Influences on Coal Miners' Safety Attention: An Evaluation Using Improved... 149 Conversely, the education level S6, working years S7, working environment S10, and work intensity S11 are relegated to the third tier, characterized by their comparatively diminished importance. Their centrality degrees, all falling below 1, suggest a relatively minor influence on coal miners' safety attentiveness. 4.2. Examination of causative factors Within the system presented in Table 2, nine causative factors are identified. Their hierarchical significance unfolds as follows: S15>S8>S13>S12>S14>S6>S7>S10>S11. It is posited that these causative elements play an instrumental role in shaping resultant factors and miners' emphasis on safety. However, an intricate scrutiny reveals that factors such as the education level S6, working years S7, and safety education and training S14 are situated within the mid-tier of the system as depicted in Fig. 2. In Table 2, a lower degree of significance is attributed to these elements. Furthermore, as discerned from Fig. 3, these factors are categorized within quadrant II and are perceived as independent clusters with minimal dependency. A tepid response to the stimuli of profound root factors is observed. Collectively, it is deduced that these components maintain a marginal and somewhat isolated impact on the overarching system. Notably, while the working environment S10 and working intensity S11 are earmarked as profound factors in Fig. 2, they are positioned within quadrant I and are associated with the autonomous cluster as per Fig. 3. The driving forces underpinning these factors are delineated as being subpar, and their affiliations with mid-tier factors are not deemed robust. Addressing these particular elements with timely, targeted interventions may proffer enhanced outcomes in bolstering coal miners' attention to safety. Pivotal within the causative factors, the safety atmosphere of the organization S15 is discerned to reside at the L7 echelon. Concurrently, working pressure S8, safety information stimulation S12, and safety supervision S13 are situated either at the L5 or L6 strata, as visualized in Fig. 2. Recognized as the system's profound layers, these factors—due to their pronounced importance are placed within quadrant II in both Figs. 1 and 3. Being constituents of an independent cluster with a formidable driving force, their influence on both mid-tier and superficial elements is profound. Thus, these are perceived as the linchpins modulating coal miners' dedication to safety. 4.3. Exploration of resultant factors In Table 1, six resultant factors are delineated, possessing a sequence of significance as follows: S2>S4>S5>S3>S9>S1. Resultant from the cumulative impacts of other determinants, these factors are discerned to directly influence the coal miners' dedication to safety. An elevated level of importance is attributed to them within the system, as visualized in Fig. 1, suggesting a critical role in shaping miners' safety attentiveness. Upon examination, it was observed that factors such as behavioral habit S3, safety self- control S4, and risk perception S5 reside within the L1 echelon in Fig. 2, typifying the system's initial surface level. Positioned in quadrant IV, these factors are identified as belonging to the dependent cluster in Fig. 3. Characterized by heightened dependency and attenuated driving forces, their manifestation is predominantly influenced by other variables, rendering them contingent on the modulation of external factors. Conversely, safety cognition S1, safety consciousness S2, and job burnout S9 are mapped to the L2, L3, and L4 tiers, respectively, in Fig. 2. These mid-tier factors within the ISM serve as 150 D.Y. XU, H.Q. CUI, T.Y. WANG, J.W. FU, G.B. ZHU conduits, bridging antecedent and subsequent determinants. The driving forces intrinsic to these elements amplify the sway of superficial factors, while their dependencies bolster the imprint of profound determinants. However, an intriguing observation arises from the positioning of safety consciousness S2 within quadrant IV, aligning it with the dependent cluster. Despite its inherent strong dependency, a subdued driving force has been noted, attenuating its potential reinforcement of surface elements. Of particular note, safety cognition S1 and job burnout S9 are aligned with quadrant III, categorizing them within the connection cluster as illustrated in Fig. 3. Their simultaneous possession of robust driving forces and dependencies suggests that perturbations within these factors could readily ripple through the system, influencing upper and lower echelons. Such dynamism underscores the salience of these factors in determining coal miners' commitment to safety. The analysis infers that mid-level resultant factors, endowed with dominant driving forces and dependencies, significantly modulate coal miners’ focus on safety. 4.4. Analysis of practical implications from influencing factors Hierarchy is often employed as a critical technique in analyzing influencing factors and forms an integral foundation in safety management. Within this hierarchy, surface-level factors have been determined to directly influence safety attentiveness. The direct and targeted intervention of these factors has been acknowledged to yield optimal results in the contemporary domain. Deep-level factors, however, are perceived to be more pivotal, exerting their influence through mid and surface-level determinants. A vast array of deep- level factors has been observed to influence other levels, rendering the nature of their influence intricate. For these deep-level determinants, comprehensive emphasis and preventive measures are generally proposed in current strategies. Mid-level factors, traditionally, are not attributed significant influence over deep-level determinants but are discerned to primarily impact surface-level factors. The prevailing strategy recommends obstructing the connections between mid and surface levels. Nonetheless, challenges are evident in these hierarchical solutions to influencing factors. An overwhelming number of factors demand attention, resulting in significant workload. Given the dynamic nature of safety information, implementing these controls becomes challenging, thereby potentially overlooking certain mid-level factors over time. Strikingly, analyses from improved methodologies suggest that specific mid-level factors play a pivotal role in modulating coal miners' safety attention, raising concerns over safety management, especially in minor injury incidents. It has been inferred that the nuanced interrelations among various factors cannot be aptly captured solely via hierarchical structures. Instead, the implementation of refined methodologies is advocated to establish a hierarchical structure. Within this, both driving forces and dependencies of influencing factors should be meticulously analyzed, unearthing the intricate interplay among them. Such a comprehensive approach promises to yield insights of paramount practical relevance. Upon scrutiny of the enhanced DEMATEL-ISM method and the intricate interplay among all influencing factors, two focal points emerge for mining entities. Firstly, the deep-level causative factors manifesting pronounced driving forces demand attention. Secondly, mid-level resultant factors characterized by robust driving forces and dependencies necessitate scrutiny. Salient among these are safety atmosphere of the organization S15, work pressure S8, safety information stimulation S12, safety supervision S13, safety cognition S1, and job burnout S9. Addressing these six pivotal determinants in isolation poses challenges, given their intertwined Interactions and Influences on Coal Miners' Safety Attention: An Evaluation Using Improved... 151 nature and hierarchical interactions. Yet, devising unified, systematic countermeasures tailored to their characteristics appears promising for mining entities, potentially alleviating concerns over minor injury incidents by enhancing miners' safety vigilance. The aforementioned six influential factors and their main associated personnel can be systematically approached in two dimensions. The initial dimension targets the triad of safety information stimulation S12, safety supervision S13, and safety atmosphere of the organization S15. Integration of safety information stimulation S12 and safety supervision S13 into the overarching framework of organizational safety atmosphere S15 is recommended, elevating the significance of the former two within the latter. This amalgamation predominantly concerns the safety managers within mining firms, underscoring the pivotal role of establishing a scientifically rigorous safety organizational atmosphere to elevate the safety management caliber. The secondary dimension focuses on safety cognition S1, work pressure S8, and job burnout S9. Primarily pertinent to coal miners, these factors, however, necessitate vigilant oversight from safety managers. Proactive measures, both in terms of rewards and penalties, are suggested to foster miners' intrinsic motivation to hone their safety cognition and maintain optimal working conditions, thereby enhancing safety awareness. Concurrently, vigilant monitoring is essential to promptly identify concerns related to miners' work pressure and burnout. Decisive interventions should be swiftly deployed upon the identification of any detrimental behaviors jeopardizing miners' safety attention. The systematic management of these dimensions, centered on the highlighted six determinants, is postulated to hold profound practical and strategic implications in mitigating minor injury incidents within coal mining entities. 5. CONCLUSIONS In the presented research, factors influencing coal miners' attention to safety were meticulously examined through an enhanced DEMATEL-ISM method. From this comprehensive analysis, critical determinants closely interlinked were extracted from the broader factor group. Such extraction is postulated to diminish the substantial workload inherent in managing minor injury accidents and to counteract the potential oversight of pivotal factors. The findings elucidated that the coal miners' safety attention emanates from a composite interplay of factors, including their degree of importance, system level grade, driving force, and dependency characteristics. Notably, two salient dimensions emerged as particularly significant for mining entities. The first dimension underscored that deep-level causative factors, characterized by an intense driving force, profoundly influence both middle and surface-level determinants, thus serving as pivotal elements modulating miners' safety focus. The second dimension accentuated the importance of mid-level resultant factors. Endowed with a robust driving force and dependency, these factors are discerned to potentially sway deep and surface- level determinants, thereby critically affecting coal miners' safety attention. It is surmised that devising systematic response measures addressing the influencing factors from both dimensions holds potential for mining companies, offering strategic solutions to challenges posed by minor injury incidents and fostering enhanced safety attentiveness among miners. Anticipated avenues for future research encompass the exploration of barriers and conflicts inherent in the formulation of such systematic response measures, as these issues hold significant implications. Furthermore, a limitation identified pertains to the sample size and scope; subsequent investigations would benefit from expanded sample sizes and diverse organizational and cross-national contexts to bolster the generalizability and robustness of the findings. 152 D.Y. XU, H.Q. CUI, T.Y. WANG, J.W. FU, G.B. ZHU Acknowledgement: The paper is a part of the research done within the Key Scientific Research Project of Universities of Henan Province 21A440016 and Young Teacher Foundation of Zhongyuan University of Technology 2021XQG05. The authors would like to thank to the Henan Province Office of Education and Zhongyuan University of Technology for their support. REFERENCES 1. Niu, L., Liu, M., Li, N., Huang, M., 2016, Relationship among job burnout, safety attention and habitual violation behavior, China Safety Science Journal (CSSJ), 26(6), pp. 19-24. 2. Michael, M.A., Apurna, K.G., Janis, J., 2020, Risk factors for the number of sustained injuries in artisanal and small-scale mining operation, Safety and Health at Work, 11(1), pp. 50-60. 3. Ajith, M.M., Ghosh, A.K., 2019, Economic and social challenges faced by injured artisanal and small- scale gold miners in Kenya, Safety Science, 118, pp. 841-852. 4. Yin, Z., Shi, F., Li, N., Niu, L., 2017, Study on influencing factors of miners’ safety attention based on SEM, Journal of Safety Science and Technology, 13(3), pp. 113-118. 5. Kumar, P., Gupta, S., Gunda, Y.R., 2020, Estimation of human error rate in underground coal mines through retrospective analysis of mining accident reports and some error reduction strategies, Safety Science, 123, 104555. 6. Bayram, M., Arpat, B., Ozkan, Y., 2022, Safety priority, safety rules, safety participation and safety behaviour: the mediating role of safety training, International Journal of Occupational Safety and Ergonomics, 28(4), pp. 2138-2148. 7. Fariba, K., Mohammad, R.K., 2014, Preventing injuries in workers: the role of management practices in decreasing injuries reporting, International Journal of Health Policy and Management, 3(4), pp. 171-177. 8. Stojadinović, S., Svrkota, I., Petrović, D., Denić, M., Pantović, R., Milić, V., 2012, Mining injuries in Serbian underground coal mines – A 10-year study, Injury, 43(12), pp. 2001-2005. 9. Abikenova, S., Daumova, G., 2023, Provision of personal protection equipment, according to the risk of exposure to harmful industrial factors during copper-polymetallic ore mining, International Journal of Safety and Security Engineering, 13(1), pp. 89-95. 10. Sanmiquel, L., Rossell, J.M., Vintró, C., 2015, Study of Spanish mining accidents using data mining techniques, Safety Science, 75, pp. 49-55. 11. Shahani, N.M., Sajid, M.J., Jiskani, I.M., Ullah, B., Qureshi, A.R., 2020, Comparative analysis of coal miner’s fatalities by fuzzy logic, Journal of Mining and Environment, 11(4), pp, 967-976. 12. Bagherpour, R., Yarahmadi, R., Khademian, A., Almasi, S.N., 2017, Safety survey of Iran’s mines and comparison to some other countries, International Journal of Injury Control and Safety Promotion, 24(1), pp. 3-9. 13. Emily, J.T., Katie, A.G., Ann, L.P., Sandra, C.D., Vic P., Tammy, R.E., 2021, Safety culture: A retrospective analysis of occupational health and safety mining reports, Safety and Health at Work, 12(2), pp. 201-208. 14. Shahani, N.M., Sajid, M.J., Zheng, X., Jiskani, I.M., Brohi, M.A., Ali, M., Qureshi, A.R., 2019, Fault tree analysis and prevention strategies for gas explosion in underground coal mines of Pakistan, Mining of Mineral Deposits, 13(4), pp. 121-128. 15. Komljenovic, D., Loiselle, G., Kumral, M., 2017, Organization: a new focus on mine safety improvement in a complex operational and business environment, International Journal of Mining Science and Technology, 27(4), pp. 617-625. 16. Khosravi, Y., Asilian-Mahabadi, H., Hajizadeh, E., Hassanzadeh-Rangi, N., Bastani, H., Behzadan, A.H., 2014, Factors influencing unsafe behaviors and accidents on construction sites: A review, International Journal of Occupational Safety and Ergonomics, 20(1), pp. 111-125. 17. Oah, S., Na, R., Moon, K., 2018, The influence of safety climate, safety leadership, workload, and accident experiences on risk perception: A study of Korean manufacturing workers, Safety and Health at Work, 9(4), pp. 427-433. 18. Li, G., Yan, Y., Liu, W., Chen, Y., Wu, Z., 2021, Research on formation factors of miners’ unsafe emotions based on DEMATEL-ISM. China Safety Science Journal, 31(7), pp. 30-37. 19. Rukiye, K., Barbaros, Y., 2019, Building Bayesian networks based on DEMATEL for multiple criteria decision problems: A supplier selection case study, Expert Systems with Applications, 134(15), pp. 234-248. 20. Hedieh, S., Mohammad, K., 2020, Analysis of factors affecting project communications with a hybrid DEMATEL-ISM approach (A case study in Iran), Heliyon, 6(8), e04430. 21. Li, M., Wang, W., Zhang, Z., 2022, Study on construction risk factors based on ISM and MICMAC, Journal of Safety and Environment, 22(1), pp. 22-28. 22. Cui, Q., Wang, S., 2021, Influencing factors of unsafe behavior of construction workers based on ISM- MICMAC, Journal of Shenyang University (Natural Science), 33(3), pp. 272-278.