1. Introduction
Despite advancements in technologies and improvements in safety measures, construction remains one of the most hazardous industries in the world [1]. The construction industry, although employing only 7.5% of the world’s workforce, is disproportionately affected by occupational injuries and fatalities, accounting for 16.4% of the total incidents [2]. According to the latest data from the U.S. Bureau of Labor Statistics (BLS), there were 1061 fatal work injuries in the construction industry in the United States in 2020 [3]. In the United Kingdom, construction accounted for over a third of all worker fatalities in 2020, according to the Health and Safety Executive [4]. In Australia, the construction industry also has a higher rate of fatal and serious injuries compared to other industries. According to Safe Work Australia [5], the construction industry accounted for 20% of all worker fatalities in 2019. It is, therefore, crucial for construction workers and their employers to prioritize safety in order to prevent accidents, protect workers, and ensure the success of the project [6].
Construction workers’ safety behavior (CWSB) refers to the actions, habits, and practices that are implemented by construction workers to ensure their own safety, as well as the safety of their colleagues and others on the construction site [7]. Most construction accidents are highly correlated with worker-centered issues [8], among which workers’ unsafe behavior is a critical aspect, such as not following proper procedures, neglecting to use personal protective equipment, or engaging in horseplay [9]. CWSBs help to reduce the risk of fatal or serious injury while on the job. This protects the health and well-being of workers, their families, and the communities they live in. In addition, CWSB is also important for complying with legal requirements, improving productivity, and reducing costs.
Academics have also paid increasing attention to CWSB because of its significant role in construction workers’ occupational safety, resulting in a body of research that is growing exponentially. Numerous reviews have reported various aspects of CWSB. For example, Liu, et al. [10] systematically summarized the current state of research on the antecedents of construction workers’ safety cognition. Xiang, et al. [11] reviewed 53 studies on the subject of the cognitive mechanism of construction workers’ unsafe behavior and developed an integrated conceptual framework to illustrate how multiple types of cognitive failures interact to cause workers’ unsafe behavior. Fang, et al. [12] reviewed the development of computer vision studies that have been used to identify unsafe behavior from two-dimensional images that arise on construction sites. Cheng, et al. [13] conducted a systematic review of eye-tracking studies of construction workers’ safety behaviors from the perspective of neuroscience. Overall, the body of review research on CWSB highlights the importance of addressing this issue in the construction industry. Despite the prevalence of traditional review studies, the field still lacks a comprehensive understanding of the exact dimensions of knowledge that have been discovered, the potential interventions that may be utilized, and the areas of research that are emerging. Therefore, it is imperative to undertake a review study that aims to provide a clear direction for future research and serve as a resource for practitioners, policymakers, and the wider research community.
The science mapping approach, characterized as ‘a generic process of domain analysis and visualization’, seeks to graphically display the relationships between different disciplines, fields, and authors within a specific body of literature [14]. It allows us to measure the impact of research, analyze institutions and peer-reviewed journals, and gain a deeper understanding of scientific knowledge and citations [15]. As a result, traditional review studies in the domain of CWSB can be enhanced [16]. However, to date, no reviews have utilized the science mapping approach to examine the existing global body of knowledge on CWSB. This study, therefore, aimed to fill this gap by conducting a science mapping-based review of the scientific literature relating to CWSB, with the following specific objectives: (1) applying a science mapping approach to analyze the keywords, authors, countries/regions, and journal citations related to CWSB; (2) identifying existing key research topics in CWSB; and (3) highlighting limitations and gaps for future research studies in CWSB. The findings of this review will provide a comprehensive understanding of the body of knowledge on CWSB and identify crucial research topics and management measures to improve the occupational safety of construction workers.
2. Research Methodology
This paper presents a comprehensive scientometric analysis of CWSB research utilizing a science mapping approach. Science mapping is a well-established method for objectively mapping areas of scientific knowledge. It is also applied to visualize research topics in various fields in the construction industry, such as construction waste management [17], global mental health research in the construction industry [18], global green buildings [19], and the application of deep learning in the construction industry [20]. This method visualizes important patterns and trends in bibliographic data, facilitating the identification of key research themes and challenges. It is a valuable tool for researchers, policymakers, and other stakeholders interested in understanding the structure and dynamics of a scientific field. Figure 1 illustrates the research design.
2.1. Literature Search and Selection
The publications analyzed in this study were primarily obtained from the Web of Science Core Collection (WoS) database. The WoS database is a widely recognized and respected source of bibliographic information [1]. It offers comprehensive coverage of disciplines, advanced search functionalities, and standardized metadata, providing valuable context for review studies [21]. Additionally, the database’s citation indexing feature allows researchers to track the citation history of individual articles and identify influential papers, seminal works, and emerging trends within a field. Web of Science also ensures quality control by indexing only high-quality, peer-reviewed publications. Moreover, its compatibility with various bibliometric analysis tools and software facilitates efficient and effective literature retrieval, analysis, and visualization for review studies and science mapping [22]. Therefore, this study selected the WoS database for data collection.
This review focuses on the safety behaviors of construction workers. Therefore, the following search query was utilized in the WoS database: ((TS = (construction OR building)) AND TS = (worker OR employee OR labor OR labour workforce OR builder OR laborer) AND TS = (* safety OR * security)) AND TS = (behavior OR operation OR action OR act OR conduct). The “*” denotes a fuzzy search, and “TS” refers to the topic of the article, including the title, abstract, and author keywords. No date range limit was imposed on this study, so the date range was set to “all years published to date”. Only journal articles were selected for analysis, while proceedings papers, review articles, editorials, and conference papers were excluded. This is because journal papers typically provide more comprehensive and higher-quality information than other types of publications. During the manual review, irrelevant research directions, such as music, zoology, and marine freshwater biology, were removed. As of early February 2023, a total of 3280 publications were collected, and all literature information was exported from WoS to establish the dataset for this study.
2.2. Data Analysis
The data analysis is facilitated by the CiteSpace software in this study. CiteSpace is a Java application designed for visual analysis tasks in scientific mapping. The application can systematically create various accessible diagrams and has a significant advantage in mapping knowledge domains by providing insight into past research, discovering hidden meanings in vast amounts of information, and identifying emerging trends and critical points in specific domains by diagramming the direction of knowledge domains [22]. By adopting CiteSpace as a science mapping tool, four types of scientometric analyses and visualizations were conducted in this research [16]:
Co-word analysis: This segment encompasses keyword co-occurrence networks and the evolution of keyword networks.
Co-author analysis: This section encompasses co-occurrence networks of authors, countries, and institutions, and the analysis aims to reflect micro- and macro-level collaborations.
Co-citation analysis: This section identifies co-cited journals, co-cited authors, and co-cited literature.
Cluster analysis: This portion of the analysis seeks to categorize and label the cited literature into distinct clusters.
Two quantitative metrics, betweenness centrality and burst strength, correspond to key information in the diagram. Betweenness centrality detects the influence of nodes on the flow of information in the diagram and is typically used to find nodes that act as a bridge from one part of the diagram to another. Burst strength refers to the frequency surge strength of a particular publication concerning citation. Following the cluster analysis, a critical review was undertaken to discern research themes and the associated challenges.
3. Scientometric Analysis and Results
3.1. Wave of CWSB
Figure 2 illustrates the distribution of 3280 bibliographic records from 1991 to 2023. The first study related to CWSB, titled “Methane Studies for the Channel Tunnel”, was published in the Quarterly Journal of Engineering Geology and Hydrology. It investigated the potential for gas leaks caused by tunnel construction workers during the construction process [23]. The annual publication volume of CWSB research fluctuates slightly but is on an increasing trend, peaking in 2022.
In addition, more CWSB research publications emerged in the 21st century (2001–2023) than in the 20th century (1991–2000). The threshold of 100 articles per year was exceeded in 2013, followed by 300 and 400 articles per year in 2019 and 2021 (two years apart). This highlights the recent growth of CWSB research in line with the rapidly enhanced construction safety management in practice. As noted previously, the literature search was conducted on 8 February 2023, so publications that may appear in WoS after this date may not have been captured. Consequently, it is important to note that the number of publications in 2023 may rise towards the year’s end while restricting the review to publications retrievable on the literature search date.
3.2. Co-Word Analysis
Numerous research themes and topics in CWSB research have converged and evolved over more than three decades, epitomizing trends and frontiers in the field. Data from the WoS bibliographic records were assessed to construct a network of co-occurring keywords in the realm of CWSB. Keywords of research articles offer representative and succinct descriptions of the content, encapsulating the core substance of the articles and demonstrating the evolution of research topics over time. Through keyword co-occurrence networks, hot topics in the field of knowledge can be discerned within specific time intervals. Networks can elucidate developments in the domain of evolving research. In the WOS database, there are two types of keywords: (i) “author keywords” supplied by authors, and (ii) “keywords plus” identified by journals. These two types of keywords from 3280 bibliographic records were employed to construct a web co-occurrence keyword.
Figure 3 depicts a network of co-occurring keywords with 896 nodes and 3364 links. Node size signifies the frequency of keyword occurrences in the dataset. The top 10 most frequent keywords were ‘performance’ (frequency = 399), followed by ‘model’ (frequency = 398), ‘management’ (frequency = 383), ‘health’ (frequency = 371), ‘safety’ (frequency = 303), ‘climate’ (frequency = 284), ‘indoor’ (frequency = 284), ‘industry’ (frequency = 220), ‘construction safety’ (frequency = 217), ‘system’ (frequency = 217), and ‘safety climate’ (frequency = 245). The keywords ‘behavior’, ‘worker’, and ‘construction’ were excluded from the analysis. The aforementioned keywords are only a part of our research theme, and while they are essential foundational concepts, they do not determine the overall research direction. For instance, the keyword ‘behavior’ alone cannot specify the specific research focus on safety behavior. The term ‘performance’ emerged as the most frequently utilized keyword, as it can be employed to gauge the safety behavior of employees within the context of project management on construction sites.
Furthermore, several keywords attained relatively high median centrality scores, such as ‘industry’ (centrality = 0.11), ‘safety’ (centrality = 0.09), ‘attitude’ (centrality = 0.08), ‘occupational health’ (centrality = 0.18), ‘accident’ (centrality = 0.05), and ‘exposure’ (centrality = 0.05). These keywords interconnected various research themes and significantly impacted the development of CWSB research.
Moreover, a total of 55 keywords emerged as citation bursts through keyword analysis. Figure 4 highlights the five keywords with the highest burst strength, namely ‘labor and personnel issues’ (burst strength = 21.02, 2013–2017), ‘Hong Kong’ (burst strength = 11.94, 2008–2017), ‘culture’ (burst strength = 9.51, 2006–2015), ‘injury’ (burst strength = 9.04, 2007–2014), and ‘work’ (burst strength = 8.76, 2004–2015). It is worth noting that the keyword ‘Hong Kong’ has a high prominence in the research due to the Hong Kong Polytechnic University and the City University of Hong Kong being ranked first and second in terms of CWSB publications. During this period, there were numerous research articles published on CWSB related to Hong Kong, leading to the repeated mention of the keyword. These findings indicate that these themes represented hotspots in CWSB research during the corresponding years.
Additionally, ‘deep learning’ (burst strength = 6.39, 2021–2023), ‘simulation’ (burst strength = 4.61, 2021–2023), ‘personal protective equipment’ (burst strength = 4.41, 2020–2021), ‘safety compliance’ (burst strength = 4.21, 2021–2023), and ‘virtual reality’ (burst strength = 3.76, 2020–2023) are among the keywords experiencing a surge in citations post-2020, signifying their emergence as recent research hotspots in the CWSB domain.
3.3. Scientific Collaboration Networks in CWSB: Co-Author Analysis
Detailed information about authors can be obtained from the WoS database, and constructing a collaborative network facilitates the identification of pivotal researchers, institutions, and national collaborations in CWSB research. This offers valuable insights into the current state of research and influential authors in the field.
3.3.1. Co-Authorship Network
In this study, the top ten most productive authors in the field of CWSB were selected based on the number of journal publications, as presented in Table 1. Among them, Li H from Hong Kong Polytechnic University, Chan APC from Hong Kong Polytechnic University, and Lee S from the University of Michigan were identified as the top three authors.
A co-authorship network was created in Figure 5 to visualize the collaboration patterns among authors. This network consists of 645 nodes and 483 links, with each node representing an author and each link representing their collaboration. Node size indicates the number of publications, while link thickness indicates the level of collaboration. The different colors of the links correspond to different time periods. It can be found that three research communities were identified based on the collaboration and work in the field. The largest community was led by Albert Alex and included Gambatese John A, Hallowell Matthew R, Chan Alan Hoi Shou, Nnaji Chukwuma, and Lipscomb Hester Johnstone. Lee SangHyun led a research group consisting of Jebelli Houtan, Ahn Changbum Ryan, Han Sanguk, Choi Byungjoo, Ahn Seungjun, and others. Li Heng was identified as the central author of another research community that included other authors such as Anwer Shahnawaz, Skidmore Martin, and Antwl-afari Maxwell Fordjour.
In social networks, a researcher’s influence can be measured by their betweenness centrality, which reflects their connections to other nodes in the network [14]. Nodes with high betweenness centrality are important hubs that connect different groups of nodes and facilitate communication across the network. In CiteSpace, nodes with a betweenness centrality greater than 0.1 are identified with a purple ring. However, in this study, nodes with a betweenness centrality of zero were observed throughout the network, indicating the need for more collaboration between different research groups. The findings suggest that the development of the CWSB field requires more academic exchange activities.
3.3.2. Network of Countries/Regions
A network was constructed to examine the country distribution of articles in the CWSB domain, comprising 121 nodes and 494 links (Figure 6). Node size represents the total number of articles published in each country/region from 1991 to 2023. The analysis identified leading countries in CWSB research, with the USA (1027 articles), China (605 articles), the UK (248 articles), Australia (372 articles), Canada (167 articles), and South Korea (134 articles) making significant contributions to the development of CWSB research. The substantial volume of journal publications from these countries indicates progress in the field, with the USA contributing the most.
Centrality is defined as the ratio of the shortest path between two nodes to the sum of all such shortest paths. Five networks of purple rings formed by the USA (centrality = 0.59), UK (centrality = 0.24), Australia (centrality = 0.15), France (centrality = 0.11), and Italy (centrality = 0.11) were identified as key infrastructure nodes occupying a central position in the network of research activities.
Significant citation bursts were identified through the burst citation algorithm, with the top three countries in terms of burst citations being the USA (burst strength = 40.81, 1993–2011), China (burst strength = 13.58, 2022–2023), and Thailand (burst strength = 5.25, 2005–2011). It is noteworthy that China shows a higher citation growth in 2022–2023, indicating that the development of the CWSB field in China is gaining global attention.
3.3.3. Network of Institutions
In addition, this study constructed a network of institutional collaborations in the field of CWSB to pinpoint key research institutions and their collaborative efforts (Figure 7). The network comprises 656 nodes and 631 links. Several institutions, such as the Hong Kong Polytechnic University (101 articles), City University of Hong Kong (63 articles), University of Michigan (41 articles), RMIT (40 articles), and Queensland University of Technology (40 articles), have made substantial contributions to CWSB research. These institutions are regarded as global publication hubs for CWSB research, boasting unique research outputs in the CWSB domain. Moreover, nodes with high betweenness centrality, such as the University of Illinois at Urbana-Champaign (centrality = 0.13), Duke University (centrality = 0.11), West Virginia University (centrality = 0.10), and Curtin University (centrality = 0.10), have been identified and highlighted with purple rings in Figure 7, as they occupy key positions in the network and link research activities between different institutions.
Furthermore, the burst citation algorithm has identified institutions with significant citation bursts, including NIOSH (burst strength = 9.91, 1996–2013), Colorado University (burst strength = 9.42, 2009–2014), and Hangzhou University of Electronic Science and Technology (burst strength = 9.36, 2018–2019), suggesting that articles from these institutions have attracted significant attention in corresponding years. Figure 7 indicates that institutions from mainland China, Hong Kong, the USA and Australia have established successful partnerships for CWSB research. On the other hand, institutions from countries such as Singapore, the UK and Canada show limited institutional cross-linking, which may indicate a lack of cross-fertilization of CWSB research ideas at the institutional level. Therefore, fostering collaborative relationships within the CWSB network is essential to achieve the highest standards of scholarship and debate.
3.4. Co-Citation Analysis
Co-citation refers to the frequency with which two documents are cited together by other documents, and it is widely regarded as a valid and reliable measure of literature similarity. In this study, co-citation analysis has been employed to derive journal co-citation networks, author co-citation networks, and literature co-citation networks by analyzing bibliographic data from the WoS database.
3.4.1. Journal Co-Citation Network
This study utilized the journal co-citation networks to pinpoint significant journals within CWSB research and to visualize the intellectual structure of this field. By examining the network structure, researchers can identify clusters of related journals and track the evolution of the field over time. In this study, 221 cited journals received citation bursts, with 14 of these journals receiving citation bursts of 20.0 and above. Table 2 identifies the top 15 source journals for CWSB research. Safety Science topped the list with 213 (6.49%) published CWSB research papers, implying that the journal has greater authority in the field and is recognized by researchers. It is followed by the Journal of Construction Engineering and Management (164 articles) and the International Journal of Environmental Research and Public Health (117 articles).
The references cited in the 3280 records retrieved were analyzed, and then a journal co-citation network with 848 nodes and 2576 links was generated to detect the most important cited journals, as shown in Figure 8. Node size indicates the co-citation frequency of each source journal. In terms of co-citation frequency, the top five most influential journals were Safety Science (frequency = 1383), Journal of Construction Engineering and Management (frequency = 999), Journal of Safety Research (Frequency = 974), Accident Analysis and Prevention (Frequency = 934), and Journal of Applied Psychology (Frequency = 709). It is worth noting that the first four journals are also the top-ranked source journals in which articles on CWSB have been published. Thus, journals that contributed more to CWSB research also attracted more citations. In Figure 8, it is evident that some nodes exhibit high betweenness centrality and are highlighted by purple rings, such as the American Journal of Industrial Medicine (centrality = 0.12) and Safety Science (centrality = 0.11). These journals have a high degree of betweenness centrality, representing their high degree of interdisciplinarity. As major intellectual turning points, they link different journals at different stages of their careers. These journals are important centers of knowledge for academics, practitioners, and government agencies. Citation bursts were also found in the American Journal of Industrial Medicine (burst strength = 46.19, 1994–2016), Journal of Occupational And Environmental Medicine (burst strength = 41.25. 1996–2016), International Journal of Environmental Research and Public Health (burst strength = 30.84, 2021–2023), Construction Safety (burst strength = 29.47, 2003–2016), and American Journal of Public Health (burst strength = 29.22, 1997–2016). These findings imply that articles published in these journals have received strong citations in a short period of time, indicating their high level of interest and relevance.
3.4.2. Author Co-Citation Network
Author co-citation analysis is a valuable tool for visualizing the intellectual structure of academic disciplines. This is achieved by measuring the frequency of co-citations between authors in the cited literature, which helps identify relationships between authors whose works are cited in the same article. Additionally, author co-citation analysis can help analyze the evolution of research communities. In this study, we used author co-citation analysis to construct a network with 814 nodes and 2351 links (Figure 9). The size of each node represents the number of co-citations, and the links between nodes are indirect collaborations based on co-citation frequency.
The top ten most cited authors in this network were Zohar D (frequency = 375, Israel), Choudhry RM (frequency = 310, Saudi Arabia), Neal A (frequency = 252, Australia), Fang DP (frequency = 252, China), Hinze J (frequency = 250, USA), Clarke S (frequency = 211, UK), Lingard H (frequency = 211, Australia), Li H (frequency = 190, China), Mohamed S (frequency = 188, Australia), and Hofmann DA (frequency = 186, Germany). We also calculated the meso-centrality metric for each author in the network. The authors with the highest meso-centrality metric were Bur Lab Stat (centrality = 0.08), Cooper MD (centrality = 0.07), Mohamed S (centrality = 0.07), Zohar D (centrality = 0.07), Chi S (centrality = 0.07), Mitropoulos P (centrality = 0.07), and Goldenhar LM (centrality = 0.07). Authors with high citation frequency do not necessarily have high betweenness centrality, but those with both are likely to have a strong influence on the development of CWSB. These authors, identified as pivotal contributors within the network, occupy crucial positions that facilitate collaboration and information flow. Their scholarly output likely exerts a substantial impact on their respective fields, thus making their work indispensable for further investigation, analysis, or cooperative endeavors. In addition, these authors come from diverse locations, which suggests that research on CWSB has been conducted worldwide.
To further identify influential researchers, we used burst strength to measure rapid increases in citation frequency over short periods of time. The top five authors with the highest burst strength were Gillen M (burst strength = 22.17, 2005–2017), Lipscomb HJ (burst strength = 20.36, 2008–2016), Dedobbeleer N (burst strength = 20.32, 2002–2016), Guldenmund FW (burst strength = 20.01, 2006–2015), and Cooper MD (burst strength = 18.58, 2002–2015). These authors played a role in directing the development of CWSB during specific periods, indicating their strong influence in the field.
3.4.3. Document Co-Citation Analysis
Document co-citation analysis is used to examine a network of co-cited references, providing a knowledge base of the selected literature and revealing the number and authority with which it has been cited. In this study, co-cited references in the CWSB research area were analyzed to objectively explore the knowledge base of this field. Table 3 listed the top five cited CWSB researchers according to the WoS citation measure, including Haslam, et al. [24], Clarke [25], Choudhry and Fang [26], Mohamed [27], and Paterson, et al. [28]. The co-citation network was also constructed, as depicted in Figure 10, which comprises 937 nodes and 1839 links. Each node signifies a document labeled with the first author and publication year. The node size indicates the co-citation frequency, and the network is an exclusive compilation of 110,774 documents cited by the 3280 bibliographic records in this research, which may not be part of the indexed corpus. Guo, et al. [29] and Man, et al. [30] received 66 and 56 co-citations, respectively, and thus occupied the top two positions, followed by Li, et al. [31] (frequency = 51), Fang, et al. [32] (frequency = 46), and Zaira and Hadikusumo [33] (frequency = 41). These articles possess significant reference value and are widely acknowledged by their peers.
Documents with high betweenness centrality, as shown by the purple ring in Figure 10, are also noteworthy. Representative documents are Christian, et al. [7] (centrality = 0.18), Guo, et al. [29] (centrality = 0.12), Zohar [34] (centrality = 0.10), and they can be considered as major academic turning points in CWSB research, most of which are included in the top five highly cited articles, as shown in Table 3. In addition, strong citation bursts were found in the following five documents: Li, et al. [31] (burst strength = 18.23, 2017–2020), Choudhry [35] (burst strength = 16.53, 2016–2019), Shin, et al. [36] (burst strength = 14.39, 2016–2019), Seo, et al. [37] (burst strength = 13.83, 2018–2020), and Fang, et al. [38] (burst strength = 13.42, 2018–2020). The results show that the citations of these documents have increased significantly within a short period of time in the respective years. These documents may represent significant breakthroughs or innovations in the field, which generated a lot of interest and attention from other researchers.
4. Cluster Analysis
Cluster analysis is a widely-used exploratory data mining technique for statistical data analysis and knowledge discovery, capable of partitioning a large research dataset into distinct units, thereby aiding in identifying research themes, trends, and their interconnections within a research domain. The log-likelihood ratio (LLR) algorithm is known for producing high-quality clusters with high intra-class similarity and low inter-class similarity. Furthermore, it can select labels for each cluster based on the keywords of the documents cited in each cluster, reflecting the focus of that cluster with respect to uniqueness and coverage. In this study, the LLR algorithm was utilized as a clustering technique to identify 13 co-citation clusters.
Figure 11 represents an visualization of cluster analysis based on literature co-citation analysis, accounting for the effect of time on the results of cluster analysis and accurately reflecting the development of each cluster. Table 4 provides detailed information, including themes and size (i.e., number of members), with the silhouette reflecting the average homogeneity of the clusters. When the silhouette score surpasses 0.7, the clustering results have high reliability. In addition to the cluster labels shown in Figure 11, the LLR algorithm generated a series of alternative labels. Representative documents for each group are listed based on the frequency of co-citations within a group.
The largest cluster, Cluster #0 ‘safety behavior’, consisting of 105 members, was identified through this approach. Cluster #1, spanning the years 2009 to 2017, has 70 members and is named ‘safety risk perception’. Cluster #3, spanning the years 2013 to 2021, has 65 members and is named ‘safety behavior’. Cluster #4, spanning the years 2005 to 2014, has 59 members and is named ‘North Carolina’. Cluster #6, spanning the years 2007 to 2016, has 46 members and is named ‘repair maintenance’. Cluster #7, spanning the years 2001 to 2008, has 42 members and is named ‘meta-analytic review’. Cluster #9, spanning the years 2007 to 2014, has 28 members and is named ‘geo-referenced hazard area’. Cluster #10, spanning the years 2011 to 2022, has 17 members and is named ‘leadership practice’. Cluster #11, consisting of 11 members, is named ‘role’. The smallest cluster, Cluster #38 ‘identification’, spanning the years 2008 to 2009, contains only 4 members.
The earliest studies in this domain focused on investigating the relationship between various factors of unsafe behavior and occupational accidents through meta-analysis, analyzing the overall clustering trends from a timeline perspective [39]. Subsequent research has been centered around construction management systems and team-based safety training aimed at enhancing construction safety [31]. Further research has delved into the characteristics of worker behavior using computer vision and sensors [40], which are then combined with deep learning algorithms for object tracking and behavioral analysis [32]. At the same time, the potential of virtual reality in safety training is gradually being discovered and applied [41].
5. Critical Review
Three major research topics are identified from the CWSB research, including the conception and dimension of CWSB, critical influence factors of CWSB, and emerging technologies in CWSB research and management.
5.1. Conception and Dimension of CWSB
In general, CWSB refers to the macro abstraction of safety-related behaviors of construction workers during construction work [26,35]. Diverse scholars believe CWSB is a multidimensional concept and have approached the deconstruction of this construct from multiple vantage points. For example, Andriessen [42] divided CWSB into carefulness and safety initiatives. Marchand, et al. [43] then divided CWSB into safety compliance and safety initiative, arguing that carefulness and safety compliance characterize the same type of behavioral tendency. Similarly, Neal and Griffin [44] constructed a two-dimensional conceptual structure of safety compliance and safety participation, which is widely adopted by other scholars [29,45]. These scholars all deconstruction CWSB in terms of their submissive and active psychology, while there are also many other deconstruction methods. Larsson et al. [46,47] classified CWSB into structural safety behavior, interactional safety behavior, and personal safety behavior. classified CWSB into task performance and contextual performance. Hofmann and Morgeson [48] divided CWSB into individual levels and organizational levels. Furthermore, Meng, et al. [49], Zhang, et al. [50] argued that CWSB includes mutual help among workers, leader–subordinate relationships, participation in suggested decisions, and self-control. Overall, CWSB is a complex and multidimensional concept that has been approached from various perspectives by different scholars. The different approaches to deconstructing the concept of CWSB provide a deeper understanding of the various factors that contribute to CWSB.
5.2. Critical Influence Factors of CWSB
Factors that influence CWSB are diverse and complicated. Factors impacting CWSB may be from multiple levels of individual, organization, and workplace. Construction workers’ individual physiological and psychological characteristics significantly impact their safety behavior. For example, Fang, et al. [51] indicated that physical fatigue contributes to construction workers’ unsafe behaviors. Kao, et al. [52] construction workers with insomnia are more likely to perform unsafe behavior. Leung, et al. [53] highlighted the impacts of construction workers’ emotional stress on their safety behaviors. Zhang and Fang [54] identified cognitive failure as the primary cause of construction workers’ unsafe behavior. Additionally, individual factors such as risk perception and acceptance levels, safety knowledge, safety competence, and safety intentions have been shown to exert direct effects on CWSB [55,56,57].
Organizational attributes can influence worker safety behavior. Leaders’ management affects workers’ behaviors. For example, Kapp [58] found that safety climate positively moderates leadership management behavior and worker behavior. He, et al. [59] demonstrated that high-quality leader-member exchanges within the construction team can enhance CWSB. Kaufman, et al. [60] also indicated that safety support from team leaders can motivate worker safety behavior. In addition, colleagues’ behavior also impacts CWSB. Choudhry, et al. [61] illustrated that some workers’ unsafe behaviors originate from peer pressure. Liang, et al. [62] also found construction workers’ safety violations have social contagion effects.
The workplace environment plays a critical role in influencing CWSB. A positive and safe work environment can significantly reduce the risks of accidents and injuries [63]. Factors such as effective communication, proper training, and access to appropriate safety equipment contribute to promoting responsible behavior among workers [54]. Keeping the workplace clean also encourages CWSB [64]. Conversely, a stressful or chaotic workplace can lead to carelessness, inadequate adherence to safety protocols [13], and increased potential for accidents [65]. Yang, et al. [66] also indicated construction workers tend to work unsafely in noisy workplace environments.
These various factors interact and contribute to the overall CWSB of construction workers, highlighting the importance of considering a holistic approach to promoting safety in the construction industry.
5.3. Emerging Technologies in CWSB
In recent years, as technology continues to advance and expand in applications, an increasing number of technologies are being employed for the research and management of CWSB. Notably, computer vision, wearable devices, and virtual reality (VR) serve as prominent examples.
The application of computer vision can be used to enhance the safety behavior of construction workers. Researchers can develop a computer-based monitoring system that tracks workers’ safety behaviors in real time. The integration of action recognition methods, including deep learning algorithms, can automatically detect non-compliance with safety rules and regulations, such as the failure to wear safety helmets [67], walking through concrete/steel supports [68], or not wearing safety harnesses at heights [69]. For example, Li, et al. [31] employed novel information technologies such as Proactive Behavioural Safety (PBBS) and Proactive Construction Management Systems (PCMS) to automatically monitor and record workers’ behavior on-site, leading to improved construction safety. Similarly, Han and Lee [40] extracted motion data from three-dimensional human skeletal motion models to identify unsafe worker behavior with the aid of computer vision techniques. Ding, et al. [70] enhanced the accuracy of detecting image points of interest (unsafe behavior) by combining convolutional neural networks and long-term, short-term memory neural networks. Computer version-based methods offer a variety of potential directions for researching construction workers’ safety behaviors.
Wearable devices are increasingly being used to promote CWSB [71]. These devices can provide real-time feedback and monitoring of workers’ movements and actions, alerting them to potential hazards and reminding them of safety protocols. For example, Cheng, et al. [1] identified the electroencephalogram’s (EEG’s) potential in measuring and computing construction workers’ recognized states and thus revealing the mechanism of CWSB from the perspective of neuroscience. Other CWSB research such as Ke, et al. [72] and Chen, et al. [73], also used EEG technologies. In addition, Hasanzadeh, et al. [74], Noghabaei, et al. [75], and Xu, et al. [76] then used wearable eye-tracking devices to record construction workers’ visual attention distribution to investigate their safety behaviors. Furthermore, various other wearable devices were also used in CWSB research, such as posture sensors [77], GPS trackers [78], activity trackers [79], etc. Overall, wearable devices offer a wide range of possibilities for researching CWSB. By leveraging these devices, researchers can collect data on workers’ behaviors and movements in real-time, identify potential safety risks, and develop interventions to improve safety and prevent accidents and injuries on construction sites.
VR technologies are increasingly being used to contribute to research on CWSB. VR provides a safe and controlled environment where workers can experience and practice different hazardous scenarios; without the risk of injury [79]. This technology can be used to simulate various construction scenarios, allowing researchers to study workers’ behavior in response to different safety hazards and situations [80,81,82]. VR can be used as a training tool to teach workers about safety protocols and procedures [83]. For example, Joshi, et al. [84] developed a VR module to train safety protocols in the prestressed/precast concrete industry. VR can be used to create virtual simulations of different construction sites and scenarios, which researchers can use to identify potential hazards and assess workers’ reactions to them [41,85]. Noghabaei, et al. [75] combined EEG and eye tracking in an immersive virtual environment to predict when safety hazards will be successfully recognized during hazard recognition efforts using machine learning techniques. Another usage of VR in CWSB research is to evaluate the effectiveness of different safety interventions. Researchers can use VR to simulate a hazardous scenario and test the effectiveness of different safety protocols; such as providing workers with additional training; modifying work procedures; or introducing new safety equipment. For instance; Kim, et al. [85] investigated the use of VR as a behavioral intervention tool to prevent workplace accidents caused by worker habituation to risks associated with hazards in road construction work zones. In summary, VR technology offers a powerful tool for researching CWSB. By using VR simulations, researchers can create a safe and controlled environment where they can study workers’ behavior and develop new approaches for improving safety and preventing accidents and injuries on construction sites.
5.4. Research Gaps and Recommendations
CWSB research has made significant progress in recent years, but there are still some gaps that need to be addressed, including (1) a comprehensive understanding of the influence mechanism of CWSB, (2) the long-term effectiveness of safety interventions, and (3) the integration of new technologies into safety programs.
5.4.1. Comprehensive Understanding of Influence Mechanism of CWSB
Despite significant progress in identifying factors that affect CWSB, there is still a need for a more comprehensive and systematic understanding of the underlying influence mechanisms. This includes identifying how individual, organizational, and cultural factors interact to influence safety behavior in construction. In order to achieve this, future research could adopt a more systematic approach to identify, classify, and analyze the various factors that influence safety behavior. This may involve using multi-level models to examine the interaction between individual factors, such as personality traits and experience, organizational factors, such as safety culture and leadership, and cultural factors, such as societal norms and values.
5.4.2. Long-Term Effectiveness of Safety Interventions
Despite significant efforts to improve safety behavior among construction workers through various interventions and programs, there is a gap in research regarding the long-term effectiveness of these initiatives. Existing research has mainly focused on short-term improvements in safety behavior, such as the use of personal protective equipment or participation in safety training programs. However, it is essential to investigate whether these short-term changes translate into sustained improvements in safety behavior over time. In order to address this gap in research, future studies could focus on evaluating the long-term sustainability of safety interventions and their effects on safety behavior outcomes. These studies could involve long-term follow-up assessments that track safety behavior outcomes over extended periods beyond immediate post-intervention periods. Furthermore, the research could identify the key factors that contribute to the long-term effectiveness of safety interventions, such as the role of leadership, safety culture, and worker involvement.
5.4.3. Integration of New Technologies into Safety Programs
The construction industry is increasingly adopting new technologies, such as wearable sensors and VR training, to enhance safety in the workplace. However, there is a gap in research regarding the integration of these technologies into existing safety programs and their impact on safety behavior. Further research is needed to identify the most effective ways to integrate these technologies into safety programs and maximize their impact on safety behavior. This includes identifying the types of technologies that are most effective for different types of safety interventions and developing strategies to train workers on how to use these technologies effectively. In order to address this research gap, future studies could explore how these technologies can be integrated into existing safety programs to improve safety behavior among construction workers. Research could also investigate the potential barriers to the adoption of these technologies and ways to overcome them. Additionally, studies could evaluate the impact of these technologies on safety outcomes and their cost-effectiveness compared to traditional safety interventions.
In summary, there are several research gaps in CWSB that need to be addressed to further our understanding of how to promote safer working conditions and reduce the risk of accidents and injuries on construction sites. By addressing these gaps, researchers can develop new approaches for improving safety behavior and preventing accidents and injuries in the construction industry.
6. Conclusions
This study adopted a science mapping-based method to review research related to CWSB published in WoS-index journals through co-word, co-author, citation analysis, and cluster analysis. In order to complete this task, a three-step holistic review approach consisting of bibliometric search, scientometric analysis, and in-depth qualitative discussion was used to review related research in the domain of CWSB.
In the co-author analysis, this paper discovered that the main researchers’ contributions and influence were as follows: Li H, Chan APC, and Lee S emerged as the most prolific authors, while Zohar D, Choudhry RM, and Neal A gained the top three co-citation positions. However, when comparing the most productive authors with the most influential authors, it was evident that not all highly productive researchers had an equally significant impact on CWSB research. Some researchers (e.g., Zohar D) obtained a substantial number of co-citations and a high citation burst despite having published relatively few papers. Regarding the distribution of journal papers on CWSB, the majority originated from the USA, China, and Australia. Moreover, the Hong Kong Polytechnic University, the City University of Hong Kong, and the University of Michigan were the most productive institutions in the CWSB domain. These countries and institutions also facilitated research collaboration between different countries and institutions.
Based on co-citation analysis results, some of the most notable findings in CWSB research were published in core journals such as Safety Science, Journal of Construction Engineering and Management, Journal of Safety Research, and Accident Analysis and Prevention. These journals have also maintained high co-citation frequencies and citation bursts over the past decade, signifying their robust and enduring influence on CWSB research. The majority of the top 25 highly cited articles, according to the WoS citation metric, were published in these journals, with Haslam, et al. [24] receiving the highest number of citations. Two CWSB articles, Guo, et al. [29] and Man, et al. [30], garnered the most co-citations. In the past five years, Li, et al. [31], Choudhry [35], Shin, et al. [36], Seo, et al. [37], and Fang, et al. [38] have high citation bursts, indicating that computer vision based on deep learning algorithms is a recent research trend.
Regarding keyword analysis and cluster analysis, ‘performance’, ‘model’, and ‘management’ exhibited the highest co-occurrence frequency, while ‘deep learning’ and ‘simulation’ emerged as keywords in recent years, signifying their widespread study within the CWSB domain. Furthermore, the LLR algorithm identified 13 co-citation clusters based on keywords associated with the analyzed literature’s references. The largest cluster is #0, ‘safety behavior’ with 105 documents, while the smallest cluster is ‘identification’ with four documents. Based on cluster label analysis, alternative labels, and cluster timeline development, several hot topics related to CWSB research can be summarized: computer vision, deep learning, VR, human-computer interaction, simulation, optimization, and object detection.
The critical review mainly focused on three key research areas, namely; (1) the conception and dimension of CWSB, (2) the critical influence factors of CWSB, and (3) the emerging technologies in CWSB research and management. Lastly, the current review proposed three research gaps and directions for future studies that could benefit both researchers and industry practitioners in mitigating workers’ unsafe behaviors in construction. They include (1) a comprehensive understanding of the influence mechanisms of CWSB, (2) the long-term effectiveness of safety interventions, and (3) the integration of new technologies into safety programs. By addressing the aforementioned research gaps, researchers can further enhance their understanding of the factors influencing CWSB and devise effective strategies to mitigate risks and improve occupational safety in the construction industry.
The present study, like other reviews in the field of CWSB, is subject to certain limitations. Firstly, the scope of the relevant articles included was restricted to peer-reviewed journal articles, and the search strategy was confined to the literature available on the WoS. Secondly, only journal articles published in English were taken into consideration. Further research is necessary to broaden the inclusion criteria by incorporating other types of publications, such as conference papers, books, and articles published in other languages, as well as other databases such as PubMed, Scopus, and Google Scholar. Despite these limitations, the study’s contribution to the field is expected to be valuable to researchers and practitioners in the construction industry and beyond.
Conceptualization, B.C. and H.C.; Data curation, Y.W.; Formal analysis, B.C., Y.W. and H.L.; Funding acquisition, B.C. and H.C.; Investigation, B.C. and H.L.; Methodology, B.C., Y.W., H.L. and H.C.; Project administration, J.H. and H.C.; Resources, H.C.; Supervision, J.H. and H.C.; Validation, B.C. and Y.W.; Writing—original draft, B.C., Y.W. and H.L.; Writing—review & editing, B.C., Y.W., J.H. and H.C. All authors have read and agreed to the published version of the manuscript.
Not applicable.
Not applicable.
The data presented in this study are available in the manuscript.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Footnotes
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Enlarge this image.Figure 4. Top 25 keywords with the strongest citation bursts in the CWSB literature (1991–2023).
The top 10 most productive authors in CWSB research, 1991–2023.
| Author | Institution | Country | Count | Percent |
|---|---|---|---|---|
| Li H | Hong Kong Polytechnic University | China | 38 | 1.16% |
| Chan APC | Hong Kong Polytechnic University | China | 27 | 0.82% |
| Lee S | University of Michigan | United States | 27 | 0.82% |
| Lingard H | Royal Melbourne Institute of Technology | Australia | 26 | 0.79% |
| Han S | Kangwon National University | South Korea | 21 | 0.64% |
| Teizer J | Aarhus University | Denmark | 20 | 0.61% |
| Chan AHS | City University of Hong Kong | China | 19 | 0.58% |
| Hon CKH | Queensland University of Technology (QUT) | Australia | 19 | 0.58% |
| Albert A | North Carolina State University | United States | 18 | 0.55% |
| Fang DP | Kunshan Integrated TCM & Western Med Hosp | China | 18 | 0.55% |
Top 15 source journals in the research corpus.
| Journal Name | Number of Articles |
Percent |
|---|---|---|
| Safety Science | 213 | 6.49% |
| Journal of Construction Engineering and Management | 164 | 5.00% |
| International Journal of Environmental Research and Public Health | 117 | 3.57% |
| Automation In Construction | 98 | 2.99% |
| Engineering Construction and Architectural Management | 62 | 1.89% |
| Journal of Safety Research | 59 | 1.80% |
| Sustainability | 57 | 1.74% |
| International Journal of Occupational Safety and Ergonomics | 53 | 1.62% |
| American Journal of Industrial Medicine | 49 | 1.49% |
| Construction Management and Economics | 41 | 1.25% |
| Journal of Management in Engineering | 32 | 0.98% |
| Work: A Journal of Prevention Assessment Rehabilitation | 31 | 0.95% |
| Accident Analysis and Prevention | 30 | 0.91% |
| Journal of Construction Engineering and Management Asce | 29 | 0.88% |
| Journal of Computing in Civil Engineering | 28 | 0.85% |
The top 5 cited articles among the 3280 articles, 1991–2023.
| No. | Total Citations | Article | Year | First Author |
|---|---|---|---|---|
| 1 | 566 | Contributing factors in construction accidents | 2005 | Haslam, et al. |
| 2 | 555 | The relationship between safety climate and safety performance: A meta-analytic review | 2006 | Clarke, et al. |
| 3 | 418 | Why operatives engage in unsafe work behavior: Investigating factors on construction sites | 2008 | Choudhry, et al. |
| 4 | 376 | Safety climate in construction site environments | 2002 | Mohamed, et al. |
| 5 | 369 | Vaccine hesitancy and healthcare providers | 2006 | Paterson, et al. |
Co-citation clusters of CWSB research 1991–2023.
| Cluster ID | Size | Silhouette | Cluster Label (LLR) | Alternative Label | Representative Document | Mean Year |
|---|---|---|---|---|---|---|
| #0 | 105 | 0.928 | safety behavior | mediating role; coping behavior; risk-taking behavior; construction worker safety behavior | Yang, J (2021) | 2017 |
| #1 | 70 | 0.85 | safety risk perception | hazard recognition; training transfer; safety training outcome; workers perspective | Namian, M (2016) | 2013 |
| #2 | 66 | 0.937 | computer vision | using computer vision; unsafe behaviour; ubiquitous site photo; learning-based risk analysis | Pham, HTTL (2021) | 2017 |
| #3 | 65 | 0.881 | safety behavior | virtual reality; wearable sensor; struck-by hazard; construction task | Lee, BG (2021) | 2017 |
| #4 | 59 | 0.943 | north carolina | hispanic construction worker; latina manual worker; work organization; occupational risk | Marin, LS (2015) | 2009 |
| #5 | 55 | 0.863 | occupational health | new tool; multiple source; cosmopolitan construction project; mixed method | Lingard, HC (2010) | 2007 |
| #6 | 46 | 0.871 | repair maintenance | minor alteration; neural network model | Gurcanli, GE (2013) | 2011 |
| #7 | 42 | 0.975 | meta-analytic review | dispositional approaches; accident involvement; contrasting perceptual attitudinal; engaging employees safety participation | Clarke, S (2006) | 2003 |
| #8 | 30 | 0.986 | human-robot interaction safety risk | assessment tool; evaluative safety training; workforce development; bests practice strategies | Okpala, I (2023) | 2019 |
| #9 | 28 | 0.911 | geo-referenced hazard area | building information modeling; rule checking; using range point cloud data; protective equipment planning | Teizer, J (2015) | 2010 |
| #10 | 17 | 0.974 | leadership practice | construction workgroup; communication practice; missing link; training intervention | Newaz, MT (2019) | 2016 |
| #11 | 13 | 0.997 | role | production; empirical case study; teamwork practice; cognitive model | Mitropoulos, P’Takis’ (2009) | 2006 |
| #38 | 4 | 0.997 | identification | high-performance sustainable construction project; safety risk | Fortunato, BR (2012) | 2008 |
References
1. Cheng, B.; Fan, C.; Fu, H.; Huang, J.; Chen, H.; Luo, X. Measuring and Computing Cognitive Statuses of Construction Workers Based on Electroencephalogram: A Critical Review. IEEE Trans. Comput. Soc. Syst.; 2022; 9, pp. 1644-1659. [DOI: https://dx.doi.org/10.1109/TCSS.2022.3158585]
2. Zhang, Z.; Guo, H.; Gao, P.; Wang, Y.; Fang, Y. Impact of owners’ safety management behavior on construction workers’ unsafe behavior. Saf. Sci.; 2023; 158, 105944. [DOI: https://dx.doi.org/10.1016/j.ssci.2022.105944]
3. OSHA. Census of Fatal Occupational Injuries (CFOI)—Current and Revised Data, 2019; OSHA: Washington, DC, USA, 2021.
4. HSE. Fatal Injuries in the Workplace in Great Britain; HSE: Moscow, Russia, 2020.
5. SWA. Work-Related Traumatic Injury Fatalities; SWA: Canberra, Australia, 2019.
6. Zhou, Z.; Goh, Y.M.; Li, Q. Overview and analysis of safety management studies in the construction industry. Saf. Sci.; 2015; 72, pp. 337-350. [DOI: https://dx.doi.org/10.1016/j.ssci.2014.10.006]
7. Christian, M.S.; Bradley, J.C.; Wallace, J.C.; Burke, M.J. Workplace safety: A meta-analysis of the roles of person and situation factors. J. Appl. Psychol.; 2009; 94, 1103. [DOI: https://dx.doi.org/10.1037/a0016172]
8. Ding, Y.; Liu, M.; Luo, X. Safety compliance checking of construction behaviors using visual question answering. Autom. Constr.; 2022; 144, 104580. [DOI: https://dx.doi.org/10.1016/j.autcon.2022.104580]
9. Xia, N.; Xie, Q.; Griffin, M.A.; Ye, G.; Yuan, J. Antecedents of safety behavior in construction: A literature review and an integrated conceptual framework. Accid. Anal. Prev.; 2020; 148, 105834. [DOI: https://dx.doi.org/10.1016/j.aap.2020.105834]
10. Liu, Y.; Ye, G.; Xiang, Q.; Yang, J.; Miang Goh, Y.; Gan, L. Antecedents of construction workers’ safety cognition: A systematic review. Saf. Sci.; 2023; 157, 105923. [DOI: https://dx.doi.org/10.1016/j.ssci.2022.105923]
11. Xiang, Q.; Ye, G.; Liu, Y.; Miang Goh, Y.; Wang, D.; He, T. Cognitive mechanism of construction workers’ unsafe behavior: A systematic review. Saf. Sci.; 2023; 159, 106037. [DOI: https://dx.doi.org/10.1016/j.ssci.2022.106037]
12. Fang, W.; Love, P.E.D.; Luo, H.; Ding, L. Computer vision for behaviour-based safety in construction: A review and future directions. Adv. Eng. Inform.; 2020; 43, 100980. [DOI: https://dx.doi.org/10.1016/j.aei.2019.100980]
13. Cheng, B.; Luo, X.; Mei, X.; Chen, H.; Huang, J. A Systematic Review of Eye-Tracking Studies of Construction Safety. Front. Neurosci.; 2022; 16, 891725. [DOI: https://dx.doi.org/10.3389/fnins.2022.891725]
14. Chen, C. Science Mapping: A Systematic Review of the Literature. J. Data Inf. Sci.; 2017; 2, pp. 1-40. [DOI: https://dx.doi.org/10.1515/jdis-2017-0006]
15. Gómez-Domínguez, V.; Navarro-Mateu, D.; Gómez-Domínguez, T.; Giménez-Espert, M.d.C. How much do we care about teacher job insecurity during the pandemic? A bibliometric review. Front. Public Health; 2023; 11, 1098013. [DOI: https://dx.doi.org/10.3389/fpubh.2023.1098013] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36844865]
16. Antwi-Afari, M.F.; Li, H.; Chan, A.H.S.; Seo, J.; Anwer, S.; Mi, H.-Y.; Wu, Z.; Wong, A.Y.L. A science mapping-based review of work-related musculoskeletal disorders among construction workers. J. Saf. Res.; 2023.in press [DOI: https://dx.doi.org/10.1016/j.jsr.2023.01.011]
17. Li, Y.N.; Li, M.; Sang, P.D. A bibliometric review of studies on construction and demolition waste management by using CiteSpace. Energy Build.; 2022; 258, 14. [DOI: https://dx.doi.org/10.1016/j.enbuild.2021.111822]
18. Nwaogu, J.M.; Chan, A.P.C.; Hon, C.K.H.; Darko, A. Review of global mental health research in the construction industry A science mapping approach. Eng. Constr. Archit. Manag.; 2020; 27, pp. 385-410. [DOI: https://dx.doi.org/10.1108/ECAM-02-2019-0114]
19. Li, Q.W.; Long, R.Y.; Chen, H.; Chen, F.Y.; Wang, J.Q. Visualized analysis of global green buildings: Development, barriers and future directions. J. Clean. Prod.; 2020; 245, 15. [DOI: https://dx.doi.org/10.1016/j.jclepro.2019.118775]
20. Hou, L.; Chen, H.S.; Zhang, G.M.; Wang, X.Y. Deep Learning-Based Applications for Safety Management in the AEC Industry: A Review. Appl. Sci.; 2021; 11, 821. [DOI: https://dx.doi.org/10.3390/app11020821]
21. Lu, K.; Deng, X.; Jiang, X.; Cheng, B.; Tam, V.W. A review on life cycle cost analysis of buildings based on building information modeling. J. Civ. Eng. Manag.; 2023; 29, pp. 268-288. [DOI: https://dx.doi.org/10.3846/jcem.2023.18473]
22. Tang, M.; Mu, F.; Cui, C.; Zhao, J.-Y.; Lin, R.; Sun, K.; Guan, Y.; Wang, J. Research frontiers and trends in the application of artificial intelligence to sepsis: A bibliometric analysis. Front. Med.; 2022; 9, 3968. [DOI: https://dx.doi.org/10.3389/fmed.2022.1043589]
23. Warren, C.; Birch, G.; Bennett, A.; Varley, P. Methane studies for the Channel Tunnel. Q. J. Eng. Geol. Hydrogeol.; 1991; 24, pp. 291-309. [DOI: https://dx.doi.org/10.1144/GSL.QJEG.1991.024.03.04]
24. Haslam, R.A.; Hide, S.A.; Gibb, A.G.F.; Gyi, D.E.; Pavitt, T.; Atkinson, S.; Duff, A.R. Contributing factors in construction accidents. Appl. Ergon.; 2005; 36, pp. 401-415. [DOI: https://dx.doi.org/10.1016/j.apergo.2004.12.002] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/15892935]
25. Clarke, S. The relationship between safety climate and safety performance: A meta-analytic review. J. Occup. Health Psychol.; 2006; 11, pp. 315-327. [DOI: https://dx.doi.org/10.1037/1076-8998.11.4.315] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17059296]
26. Choudhry, R.M.; Fang, D.P. Why operatives engage in unsafe work behavior: Investigating factors on construction sites. Saf. Sci.; 2008; 46, pp. 566-584. [DOI: https://dx.doi.org/10.1016/j.ssci.2007.06.027]
27. Mohamed, S. Safety climate in construction site environments. J. Constr. Eng. Manag.; 2002; 128, pp. 375-384. [DOI: https://dx.doi.org/10.1061/(ASCE)0733-9364(2002)128:5(375)]
28. Paterson, P.; Meurice, F.; Stanberry, L.R.; Glismann, S.; Rosenthal, S.L.; Larson, H.J. Vaccine hesitancy and healthcare providers. Vaccine; 2016; 34, pp. 6700-6706. [DOI: https://dx.doi.org/10.1016/j.vaccine.2016.10.042]
29. Guo, B.H.W.; Yiu, T.W.; Gonzalez, V.A. Predicting safety behavior in the construction industry: Development and test of an integrative model. Saf. Sci.; 2016; 84, pp. 1-11. [DOI: https://dx.doi.org/10.1016/j.ssci.2015.11.020]
30. Man, S.S.; Chan, A.H.S.; Wong, H.M. Risk-taking behaviors of Hong Kong construction workers—A thematic study. Saf. Sci.; 2017; 98, pp. 25-36. [DOI: https://dx.doi.org/10.1016/j.ssci.2017.05.004]
31. Li, H.; Lu, M.J.; Hsu, S.C.; Gray, M.; Huang, T. Proactive behavior-based safety management for construction safety improvement. Saf. Sci.; 2015; 75, pp. 107-117. [DOI: https://dx.doi.org/10.1016/j.ssci.2015.01.013]
32. Fang, W.L.; Ding, L.Y.; Luo, H.B.; Love, P.E.D. Falls from heights: A computer vision-based approach for safety harness detection. Autom. Constr.; 2018; 91, pp. 53-61. [DOI: https://dx.doi.org/10.1016/j.autcon.2018.02.018]
33. Zaira, M.M.; Hadikusumo, B.H.W. Structural equation model of integrated safety intervention practices affecting the safety behaviour of workers in the construction industry. Saf. Sci.; 2017; 98, pp. 124-135. [DOI: https://dx.doi.org/10.1016/j.ssci.2017.06.007]
34. Zohar, D. Thirty years of safety climate research: Reflections and future directions. Accid. Anal. Prev.; 2010; 42, pp. 1517-1522. [DOI: https://dx.doi.org/10.1016/j.aap.2009.12.019] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20538108]
35. Choudhry, R.M. Behavior-based safety on construction sites: A case study. Accid. Anal. Prev.; 2014; 70, pp. 14-23. [DOI: https://dx.doi.org/10.1016/j.aap.2014.03.007] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24686162]
36. Shin, M.; Lee, H.S.; Park, M.; Moon, M.; Han, S. A system dynamics approach for modeling construction workers’ safety attitudes and behaviors. Accid. Anal. Prev.; 2014; 68, pp. 95-105. [DOI: https://dx.doi.org/10.1016/j.aap.2013.09.019] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24268437]
37. Seo, J.; Han, S.; Lee, S.; Kim, H. Computer vision techniques for construction safety and health monitoring. Adv. Eng. Inform.; 2015; 29, pp. 239-251. [DOI: https://dx.doi.org/10.1016/j.aei.2015.02.001]
38. Fang, D.P.; Wu, C.L.; Wu, H.J. Impact of the Supervisor on Worker Safety Behavior in Construction Projects. J. Manag. Eng.; 2015; 31, 12. [DOI: https://dx.doi.org/10.1061/(ASCE)ME.1943-5479.0000355]
39. Glendon, A.I.; Litherland, D.K. Safety climate factors, group differences and safety behaviour in road construction. Saf. Sci.; 2001; 39, pp. 157-188. [DOI: https://dx.doi.org/10.1016/S0925-7535(01)00006-6]
40. Han, S.; Lee, S. A vision-based motion capture and recognition framework for behavior-based safety management. Autom. Constr.; 2013; 35, pp. 131-141. [DOI: https://dx.doi.org/10.1016/j.autcon.2013.05.001]
41. Shi, Y.M.; Du, J.; Ahn, C.R.; Ragan, E. Impact assessment of reinforced learning methods on construction workers’ fall risk behavior using virtual reality. Autom. Constr.; 2019; 104, pp. 197-214. [DOI: https://dx.doi.org/10.1016/j.autcon.2019.04.015]
42. Andriessen, J. Safe Behavior And Safety Motivation. J. Occup. Accid.; 1978; 1, pp. 363-376. [DOI: https://dx.doi.org/10.1016/0376-6349(78)90006-8]
43. Marchand, A.; Simard, M.; Carpentier-Roy, M.C.; Ouellet, F. From a unidimensional to a bidimensional concept and measurement of workers’ safety behavior. Scand. J. Work Environ. Health; 1998; 24, pp. 293-299. [DOI: https://dx.doi.org/10.5271/sjweh.323]
44. Neal, A.; Griffin, M.A. A study of the lagged relationships among safety climate, safety motivation, safety behavior, and accidents at the individual and group levels. J. Appl. Psychol.; 2006; 91, pp. 946-953. [DOI: https://dx.doi.org/10.1037/0021-9010.91.4.946] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/16834517]
45. Xia, N.N.; Xie, Q.H.; Hu, X.W.; Wang, X.Q.; Meng, H. A dual perspective on risk perception and its effect on safety behavior: A moderated mediation model of safety motivation, and supervisor’s and coworkers’ safety climate. Accid. Anal. Prev.; 2020; 134, 12. [DOI: https://dx.doi.org/10.1016/j.aap.2019.105350]
46. Larsson, S.; Pousette, A.; Torner, M. Psychological climate and safety in the construction industry-mediated influence on safety behaviour. Saf. Sci.; 2008; 46, pp. 405-412. [DOI: https://dx.doi.org/10.1016/j.ssci.2007.05.012]
47. Pousette, A.; Larsson, S.; Torner, M. Safety climate cross-validation, strength and prediction of safety behaviour. Saf. Sci.; 2008; 46, pp. 398-404. [DOI: https://dx.doi.org/10.1016/j.ssci.2007.06.016]
48. Gao, Y.F.; Gonzalez, V.A.; Yiu, T.W. Exploring the Relationship between Construction Workers’ Personality Traits and Safety Behavior. J. Constr. Eng. Manag.; 2020; 146, 13. [DOI: https://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0001763]
49. Hofmann, D.A.; Morgeson, F.P. Safety-related behavior as a social exchange: The role of perceived organizational support and leader-member exchange. J. Appl. Psychol.; 1999; 84, pp. 286-296. [DOI: https://dx.doi.org/10.1037/0021-9010.84.2.286]
50. Meng, X.C.; Zhai, H.Y.; Chan, A.H.S. Development of Scales to Measure and Analyse the Relationship of Safety Consciousness and Safety Citizenship Behaviour of Construction Workers: An Empirical Study in China. Int. J. Environ. Res. Public Health; 2019; 16, 1411. [DOI: https://dx.doi.org/10.3390/ijerph16081411] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31010171]
51. Zhang, J.J.; Zhai, H.Y.; Meng, X.C.; Wang, W.X.; Zhou, L. Influence of Social Safety Capital on Safety Citizenship Behavior: The Mediation of Autonomous Safety Motivation. Int. J. Environ. Res. Public Health; 2020; 17, 866. [DOI: https://dx.doi.org/10.3390/ijerph17030866]
52. Fang, D.P.; Jiang, Z.M.; Zhang, M.Z.; Wang, H. An experimental method to study the effect of fatigue on construction workers’ safety performance. Saf. Sci.; 2015; 73, pp. 80-91. [DOI: https://dx.doi.org/10.1016/j.ssci.2014.11.019]
53. Kao, K.Y.; Spitzmueller, C.; Cigularov, K.; Wu, H. Linking Insomnia to Workplace Injuries: A Moderated Mediation Model of Supervisor Safety Priority and Safety Behavior. J. Occup. Health Psychol.; 2016; 21, pp. 91-104. [DOI: https://dx.doi.org/10.1037/a0039144]
54. Leung, M.Y.; Chan, I.Y.S.; Yu, J.Y. Preventing construction worker injury incidents through the management of personal stress and organizational stressors. Accid. Anal. Prev.; 2012; 48, pp. 156-166. [DOI: https://dx.doi.org/10.1016/j.aap.2011.03.017] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/22664679]
55. Zhang, M.; Fang, D. A cognitive analysis of why Chinese scaffolders do not use safety harnesses in construction. Constr. Manag. Econ.; 2013; 31, pp. 207-222. [DOI: https://dx.doi.org/10.1080/01446193.2013.764000]
56. Shin, D.P.; Gwak, H.S.; Lee, D.E. Modeling the predictors of safety behavior in construction workers. Int. J. Occup. Saf. Ergon.; 2015; 21, pp. 298-311. [DOI: https://dx.doi.org/10.1080/10803548.2015.1085164]
57. Li, S.Q.; Wu, X.Y.; Wang, X.Z.; Hu, S.H. Relationship between Social Capital, Safety Competency, and Safety Behaviors of Construction Workers. J. Constr. Eng. Manag.; 2020; 146, 10. [DOI: https://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0001838]
58. Goh, Y.M.; Sa’adon, N.F.B. Cognitive Factors Influencing Safety Behavior at Height: A Multimethod Exploratory Study. J. Constr. Eng. Manag.; 2015; 141, 8. [DOI: https://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0000972]
59. Kapp, E.A. The influence of supervisor leadership practices and perceived group safety climate on employee safety performance. Saf. Sci.; 2012; 50, pp. 1119-1124. [DOI: https://dx.doi.org/10.1016/j.ssci.2011.11.011]
60. He, C.Q.; Jia, G.S.; McCabe, B.; Sun, J.D. Relationship between leader-member exchange and construction worker safety behavior: The mediating role of communication competence. Int. J. Occup. Saf. Ergon.; 2021; 27, pp. 371-383. [DOI: https://dx.doi.org/10.1080/10803548.2019.1579484]
61. Kaufman, B.R.; Cigularov, K.P.; Chen, P.; Hoffmeister, K.; Gibbons, A.M.; Johnson, S.K. Interactive effects of leader justice and support for safety on safety performance. J. Organ. Eff. People Perform.; 2014; 1, pp. 296-315. [DOI: https://dx.doi.org/10.1108/joepp-05-2014-0023]
62. Choudhry, R.M.; Fang, D.; Ahmed, S.M. Safety management in construction: Best practices in Hong kong. J. Prof. Issues Eng. Educ. Pract.; 2008; 134, pp. 20-32. [DOI: https://dx.doi.org/10.1061/(ASCE)1052-3928(2008)134:1(20)]
63. Liang, H.K.; Lin, K.Y.; Zhang, S.J.; Su, Y.K. The Impact of Coworkers’ Safety Violations on an Individual Worker: A Social Contagion Effect within the Construction Crew. Int. J. Environ. Res. Public Health; 2018; 15, 773. [DOI: https://dx.doi.org/10.3390/ijerph15040773]
64. Sun, X.L.; Chong, H.Y.; Liao, P.C.; Fang, D.P.; Wang, Y.Q. A System Dynamics Model of Prevention through Design towards Eliminating Human Error. KSCE J. Civ. Eng.; 2019; 23, pp. 1923-1938. [DOI: https://dx.doi.org/10.1007/s12205-019-0845-0]
65. Leung, M.Y.; Liang, Q.; Olomolaiye, P. Impact of Job Stressors and Stress on the Safety Behavior and Accidents of Construction Workers. J. Manag. Eng.; 2016; 32, 10. [DOI: https://dx.doi.org/10.1061/(ASCE)ME.1943-5479.0000373]
66. Yang, J.; Ye, G.; Zhang, Z.; Liu, X.; Liu, Y. Linking construction noise to worker safety behavior: The role of negative emotion and regulatory focus. Saf. Sci.; 2023; 162, 106093. [DOI: https://dx.doi.org/10.1016/j.ssci.2023.106093]
67. Mneymneh, B.E.; Abbas, M.; Khoury, H. Vision-Based Framework for Intelligent Monitoring of Hardhat Wearing on Construction Sites. J. Comput. Civil. Eng.; 2019; 33, 20. [DOI: https://dx.doi.org/10.1061/(ASCE)CP.1943-5487.0000813]
68. Fang, W.L.; Zhong, B.T.; Zhao, N.; Love, P.E.; Luo, H.B.; Xue, J.Y.; Xu, S.J. A deep learning-based approach for mitigating falls from height with computer vision: Convolutional neural network. Adv. Eng. Inform.; 2019; 39, pp. 170-177. [DOI: https://dx.doi.org/10.1016/j.aei.2018.12.005]
69. Fang, W.L.; Ding, L.Y.; Zhong, B.T.; Love, P.E.D.; Luo, H.B. Automated detection of workers and heavy equipment on construction sites: A convolutional neural network approach. Adv. Eng. Inform.; 2018; 37, pp. 139-149. [DOI: https://dx.doi.org/10.1016/j.aei.2018.05.003]
70. Ding, L.Y.; Fang, W.L.; Luo, H.B.; Love, P.E.D.; Zhong, B.T.; Ouyang, X. A deep hybrid learning model to detect unsafe behavior: Integrating convolution neural networks and long short-term memory. Autom. Constr.; 2018; 86, pp. 118-124. [DOI: https://dx.doi.org/10.1016/j.autcon.2017.11.002]
71. Awolusi, I.; Marks, E.; Hallowell, M. Wearable technology for personalized construction safety monitoring and trending: Review of applicable devices. Autom. Constr.; 2018; 85, pp. 96-106. [DOI: https://dx.doi.org/10.1016/j.autcon.2017.10.010]
72. Ke, J.; Zhang, M.; Luo, X.; Chen, J. Monitoring distraction of construction workers caused by noise using a wearable Electroencephalography (EEG) device. Autom. Constr.; 2021; 125, 103598. [DOI: https://dx.doi.org/10.1016/j.autcon.2021.103598]
73. Chen, J.; Song, X.; Lin, Z. Revealing the “Invisible Gorilla” in construction: Estimating construction safety through mental workload assessment. Autom. Constr.; 2016; 63, pp. 173-183. [DOI: https://dx.doi.org/10.1016/j.autcon.2015.12.018]
74. Hasanzadeh, S.; Esmaeili, B.; Dodd, M.D. Measuring the Impacts of Safety Knowledge on Construction Workers’ Attentional Allocation and Hazard Detection Using Remote Eye-Tracking Technology. J. Manag. Eng.; 2017; 33, 04017024. [DOI: https://dx.doi.org/10.1061/(ASCE)ME.1943-5479.0000526]
75. Noghabaei, M.; Han, K.; Albert, A. Feasibility Study to Identify Brain Activity and Eye-Tracking Features for Assessing Hazard Recognition Using Consumer-Grade Wearables in an Immersive Virtual Environment. J. Constr. Eng. Manag.; 2021; 147, 04021104. [DOI: https://dx.doi.org/10.1061/(ASCE)CO.1943-7862.0002130]
76. Xu, Q.; Chong, H.-Y.; Liao, P.-c. Exploring eye-tracking searching strategies for construction hazard recognition in a laboratory scene. Saf. Sci.; 2019; 120, pp. 824-832. [DOI: https://dx.doi.org/10.1016/j.ssci.2019.08.012]
77. Ray, S.J.; Teizer, J. Real-time construction worker posture analysis for ergonomics training. Adv. Eng. Inform.; 2012; 26, pp. 439-455. [DOI: https://dx.doi.org/10.1016/j.aei.2012.02.011]
78. Zhang, M.; Cao, T.; Zhao, X. Applying sensor-based technology to improve construction safety management. Sensors; 2017; 17, 1841. [DOI: https://dx.doi.org/10.3390/s17081841] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28800061]
79. Hwang, S.; Seo, J.; Jebelli, H.; Lee, S. Feasibility analysis of heart rate monitoring of construction workers using a photoplethysmography (PPG) sensor embedded in a wristband-type activity tracker. Autom. Constr.; 2016; 71, pp. 372-381. [DOI: https://dx.doi.org/10.1016/j.autcon.2016.08.029]
80. Zhao, D.; Lucas, J. Virtual reality simulation for construction safety promotion. Int. J. Inj. Control Saf. Promot.; 2015; 22, pp. 57-67. [DOI: https://dx.doi.org/10.1080/17457300.2013.861853]
81. Le, Q.T.; Pedro, A.; Park, C.S. A social virtual reality based construction safety education system for experiential learning. J. Intell. Robot. Syst.; 2015; 79, pp. 487-506. [DOI: https://dx.doi.org/10.1007/s10846-014-0112-z]
82. Rokooei, S.; Shojaei, A.; Alvanchi, A.; Azad, R.; Didehvar, N. Virtual reality application for construction safety training. Saf. Sci.; 2023; 157, 105925. [DOI: https://dx.doi.org/10.1016/j.ssci.2022.105925]
83. Sacks, R.; Perlman, A.; Barak, R. Construction safety training using immersive virtual reality. Constr. Manag. Econ.; 2013; 31, pp. 1005-1017. [DOI: https://dx.doi.org/10.1080/01446193.2013.828844]
84. Joshi, S.; Hamilton, M.; Warren, R.; Faucett, D.; Tian, W.; Wang, Y.; Ma, J. Implementing Virtual Reality technology for safety training in the precast/prestressed concrete industry. Appl. Ergon.; 2021; 90, 103286. [DOI: https://dx.doi.org/10.1016/j.apergo.2020.103286] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33080561]
85. Kim, N.; Kim, J.; Ahn, C.R. Predicting workers’ inattentiveness to struck-by hazards by monitoring biosignals during a construction task: A virtual reality experiment. Adv. Eng. Inform.; 2021; 49, 101359. [DOI: https://dx.doi.org/10.1016/j.aei.2021.101359]
© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Details
; Wei, Yuhu 2
; Li, Hujun 3
; Huang, Jianling 4 ; Chen, Huihua 4 1 School of Civil Engineering, Central South University, Changsha 410083, China; Department of Architecture and Civil Engineering, City University of Hong Kong, Hong Kong 999077, China
2 College of Design and Engineering, National University of Singapore, Singapore 119077, Singapore
3 School of Civil Engineering, Central South University, Changsha 410083, China; School of Civil Engineering, Henan Polytechnic University, Jiaozuo 454099, China
4 School of Civil Engineering, Central South University, Changsha 410083, China