Elsevier

Safety Science

Volume 106, July 2018, Pages 66-78
Safety Science

Perceiving interactions and dynamics of safety leadership in construction projects

https://doi.org/10.1016/j.ssci.2018.03.007Get rights and content

Highlights

  • A SEM model consisting of 5 latent variables and 26 observed variables is established.

  • Common method bias, multiple group invariances, and goodness-of-fit tests are conducted.

  • It perceives causal relationships among stakeholders from survey data.

  • It identifies leading roles on managing construction safety performance.

  • It reveals the transfer of safety leadership as construction process advances.

Abstract

This paper develops a Structural Equation Model (SEM) based approach to identify roles of leadership in managing construction safety performance in a changing project environment. A model consisting of 5 latent variables and 26 observed variables is established to reveal relationships between different stakeholders’ leaderships and construction safety performance according to a survey of 464 valid respondents. The built model is verified through several tests, including common method bias test, hypothesis test, multiple group invariance test, and goodness-of-fit tests. Impacts of path coefficient, stakeholder participation, and construction process dynamics on construction safety performance are investigated and analyzed in detail. Results indicate that: (i) At the pre-construction phase, the owner party has a strongest total effect on managing construction safety performance; (ii) At the construction phase, the contractor party plays a dominant role in managing construction safety performance associated with a highest value in both total and direct effects; and (iii) The role of the indirect effect that participating stakeholders play on construction safety performance becomes weakened as the project gradually steps from pre-construction into construction phase. The novelty of the developed approach lies in its capabilities in (a) perceiving causal relationships among stakeholders from survey data, (b) identifying leading roles on managing construction safety performance, and (c) revealing the transfer of duty and responsibility within different stakeholders as the construction advances. The research findings can assist to allocate the duty on construction safety management among different stakeholders and further facilitate the enhancement of safety performance in the construction industry.

Introduction

Construction is one of the most dangerous industries worldwide (Waehrer et al., 2007, Sacks et al., 2009), leading to a common interest in improving construction safety performance due to humanitarian reasons and rising costs of worker compensation and Occupational Safety and Health Administration (OSHA) fines (Jaselskis et al., 1996). Statistics from USA, UK, and Hong Kong reveal no significant reduction in the number of fatalities in the construction industry (Zhang and Fang, 2013). In the Chinese construction market, the number of construction accidents and the fatalities remain stubbornly high (Wu et al., 2015b, Zhang et al., 2017). Numerous researchers suggested the continuous unsafe conditions were mainly because of a misalignment of management commitment and stakeholders' actions, that is, lack of safety leadership (Tam et al., 2004, Martin and Lewis, 2013). Effective leadership plays an important role in ensuring the success of temporary construction onsite organizations that are facing a high degree of uncertainty (Tyssen et al., 2014). Strong safety leadership should be the key to enhancing the construction safety performance, particularly for the countries where construction industries are facing significant safety challenges (Roundtable, 2012, Wu et al., 2015c).

Leadership is defined as “a process of social influence in which one person can enlist the aid and support of others in the accomplishment of a common task” (Chemers, 1997). Safety leadership has emerged within the Occupational Health and Safety (OHS) literature as a key construct in construction safety management. Wong et al. (2017) concluded that workplace supervisors had a substantial influence on the safety performance of their employees. Much of the research has considered safety leadership as an antecedent to lagging indicators. Traditional safety control in the construction industry is typically based on lagging indicators (e.g. Total Recordable Incident Rate (TRIR) and Experience Modification Rate (EMR) on worker’s compensation) (Hinze et al., 2013, Nassar and Abourizk, 2014). However, these indicators have often been criticized with little predictive value to system improvement (Carder and Ragan, 2003, Cooper and Phillips, 2004). Grabowski et al. (2007) noted that a growing number of safety professionals questioned the value of lagging indicators and argued that lagging indicators did not provide enough information or insight to effectively avoid future accidents. Mengolini and Debarberis (2008) stated that an unbalanced focus on lagging after-the-fact based measures may convey an unintended message that safety prevention was less important. Choudhry (2014) indicated traditional indicators focused on the afterward analysis but paid less attention to internal factors. In general, the criticism of incident-based indicators coincided with the growing attention on exploring leading factors and cause-effect relationships between internal factors and construction safety performance. Sinelnikov et al. (2015) demonstrated that safety leadership potentially enables leading indicators of OHS, and indicated that it would be easier to take preventive action through leading indicators when leading indicators received more attention in management coordination and collaboration.

Many industrial participants/stakeholders are involved in safety issues in construction, and effective collaboration of those stakeholders can contribute to ensuring construction safety performance. The organizational commitment to safety within each stakeholder has a significant impact on cultivating a positive OHS culture (Ng and Tang, 2001). Chan et al. (2004) indicated that the most influential factor driving safety performance in the construction industry was the organizational safety policy and behaviors. Erickson (2000) stated that improvements in organizational structure, safety responsibility and accountability, communication, employee involvement, and employee responses and behavior could help improve construction safety performance. Wong et al. (1999) noted that safety system (safety committees and departments), written safety policies, measurable safety targets, and communication of safety policies to various concerned parties should be essential to construction safety. According to Tam and Fung IV (1998), safety awards or incentive schemes, safety training schemes, safety committees and level of subcontracting should be recommended for consideration of the improvement of construction safety performance. Overall, many factors that exert significant impacts on construction safety performance have been identified by different scholars. However, causal links between those internal factors and external unwanted outcomes have not been properly addressed (Chinda and Mohamed, 2007). The relationship between those safety-related internal and external factors within different stakeholders in construction projects has not been fully validated. The main reason is that conclusions rely largely on cross-sectional studies, where empirical links or correlations, rather than causal relations, are obtained between variables. Longitudinal studies of safety leadership, however, are few, where causalities should be validated with time-series data and panel data (Wooldridge, 2010).

A long debate on the allocation of duty and responsibility for construction safety management always exists among main stakeholders. Gambatese (2000) indicated that owners held a proactive role in safety and can significantly influence the safety experience on a construction project. Huang and Hinze (2006) noted that the owner’s involvement was especially important when the construction firm was not fully committed to safety. Wu et al. (2015a) concluded that the owner’s greatest leverage was the leadership to influence safety perception, motivation, and behavior of other stakeholders. Hadidi and Khater (2015) emphasized the significance in selecting contractors at the bidding stage to assure loss prevention and better safety orientation during project implementation. Kapp et al. (2003) recognized that the competitive bidding process forced contractors to cut safety and quality budget items in order to win the job, which would exert great negative impacts on construction safety management. Conchie et al. (2013), on the other hand, supported the importance of supervisors’ safety leadership in promoting construction safety. Behm (2005) highlighted the importance of the role of the designer in construction safety and suggested linking construction fatalities to the design for construction safety concept, after a review of 224 fatality investigation reports. In general, a common agreement on which party should play a leading role of leadership in construction safety has not yet been reached. This research intends to propose a Structural Equation Model (SEM) based approach to address these deficiencies, aiming to reveal causal relationships and interactions among different industry stakeholders as to perceive their safety leadership and improve construction safety performance.

In recent years, two statistical approaches, including SEM and exploratory factor analysis (EFA), have been proposed to empirically reveal and test inter-relationships in the hypothetical model (Zhang et al., 2016). EFA is capable of identifying the structure among the questionnaire items, where no prior knowledge of factors or patterns of measured variables exists (Fabrigar and Wegener, 2011). Compared with EFA, SEM is not only able to perceive complex structural relationships among variables, but also estimate all coefficients in the model simultaneously (Xiong et al., 2015). Actually, SEM is a statistical methodology that takes a confirmatory approach to the analysis of a structural theory. It represents causal processes that generate observations on multiple variables, which can handle complex relationships among variables, where some variables can be hypothetical or unobserved (latent variables). SEM is able to estimate all coefficients in the model simultaneously, and thus, one is able to assess the significance and strength of a particular relationship in the context of the complete model. In addition, the hypothesized model can be tested statistically in a simultaneous analysis of the entire system of variables to determine the extent to which it is consistent with the data (Dion, 2008, Martínez et al., 2010). Several research studies have reported the successful use of SEM. For instance, Martínez-Córcoles et al. (2011) used SEM to analyze the interaction between leadership and employees’ perceived safety behaviors and indicated that the leader behavior generated a higher safety climate among members. Hsu et al. (2012) conducted a study on the application of SEM in systematic safety performance model fitness verifications and revealed that the safety performance can be composed of four different dimensions. Al-Refaie (2013) examined effects of organizational safety management and workgroup level factors on safety self-efficacy, safety awareness, and safety behavior in Jordanian companies using SEM. Wu et al. (2015c) proposed a systematic SEM-based approach to the evaluation of prospective safety performance on construction sites, with causal relationships and interactions between enablers and goals taken into account. Generally, SEM has proven to be an effective tool in testing and investigating inter-relationships among hypothesized models. However, this kind of the inter-relationships is static at a specific time point, leading to an unknown situation when certain parameters or constraints are changed over time. So far, no SEM models have ever investigated structural dynamics during the model evolution over time.

The main objective is to answer the following questions: (i) How to perceive interactions among multiple stakeholders during the construction safety management practices? and (ii) Who plays a leading role in safety leadership in a changing project environment? In this research, an SEM-based causal model for revealing relationships between behaviors of different stakeholders and construction safety performance is hypothesized and tested using data gathered via a questionnaire survey. Impacts of path coefficients on construction safety performance are explored to gain insight on inter-dependence for stakeholder participation in safety leadership. The evolutionary dynamics of safety leadership as construction process progresses are investigated by comparing the fitted SEM at different construction phases (e.g., pre-construction and construction phases). This research attempts to provide a dynamic perspective to perceive the updating interactions among different stakeholders as the construction process is moving forward. The novelty of this research lies in (a) the state of the knowledge by proposing a systematic methodology that can deeply mine the survey data to model and identify the dynamic behaviors of multiple stakeholders as the construction process advances; and (b) the state of the practice by gain a better understanding of identifying leading roles and factors for managing construction safety in a changing project environment.

The remainder of the paper is structured as follows. In Section 2, related studies on safety leadership are reviewed. In Section 3, a methodology consisting of participants, hypothesis, framework, and measures is developed. In Section 4, data regarding construction safety performance were gathered via a questionnaire survey and verified via a bias test. In Section 5, the causal construction safety model is built and verified by means of hypothesis and the goodness-of-fit tests. In Section 6, the verified model is used to investigate impacts of path coefficient, stakeholder participation, and construction process dynamics on construction safety. In Section 7, more discussions on advantages and capacities of the developed approach are presented. In Section 8, conclusions and future works are drawn up.

Section snippets

Literature review on safety leadership

Safety leadership is the process of interaction between leaders and followers in order to achieve organizational safety goals. Several empirical studies have demonstrated the importance of leadership to safety (Barling et al., 2002, Griffin and Hu, 2013). For instance, Fang et al. (2015) indicated that the root cause of accident occurrence lied in the upper management within organizations. Huang and Hinze, 2006, Toole, 2002 analyzed the roles that project managers from all parties of

Participants

In recent years, the ramifications of construction accidents are growing with a trend towards more large-scale and more complex construction projects (Lee et al., 2011), especially in developing countries, like China. In 2012, occupational accidents in Chinese construction industry led to 2437 deaths, which exceeded any other industry including coal-mining and became the most dangerous industry in China (Safety, 2013). Particularly, the rapid development of huge and complex infrastructure

Data collection

In this research, almost 2000 questionnaires are distributed to workers in the construction industry by a variety of survey tools, including emails, mails, websites, and hand-out types. All the surveyed respondents were front-line workers, including project managers, on-site engineers, designers, supervisors, foremen, and others in construction projects, where the involved projects were mainly located in the south, central, and northeast China. As shown in Table 2, the first 24 items reveal the

SEM development

SEM comprises two types of models: namely measurement model and structural model. In regard to the required sample size for the construction of an SEM, some of the principles should be followed up with: (i) Thompson (2000) indicated that the sample size should be at least 10 to 15 times as much as that of the observed variables; (ii) Byrne (2001) noted that the maximum likelihood (that was the default method used in AMOS) required the samples to be more than 400 as to better estimate the

Analysis of results

In the above sections, the SEM model for construction safety performance is built and verified. It is thus rational and reliable to apply the established model to many kinds of analysis. In this research, in order to explore the cause-effect interaction within different stakeholders, the impacts of path coefficient, stakeholder participation and construction process on construction safety performance are mainly investigated.

Discussions

In this research, a total of 464 valid responses at pre-construction phase and 560 valid responses at construction phase are collected and further used for the model establishment, model verification, and model analysis. A holistic SEM-based approach to exploring causal-effect relationships on construction safety performance from a stakeholder’s perspective is proposed. The advantages and capacities of the developed approach are discussed as follows:

  • (1)

    Perceiving causal relationships. Causality is

Conclusions and future works

This paper has presented an SEM-based approach to exploring interactions among different stakeholders and perceiving leading roles in managing construction safety. A causal model consisting of 26 variables has been established by using the survey data, in order to reveal relationships between behaviors of different stakeholders and construction safety performance. The built model has been verified through the hypothesis test and the goodness-of-fit test. The impacts of path coefficient,

Acknowledgments

National Key Research Projects of China (Grant No. 2016YFC0800208) and the National Natural Science Foundation of China (Grant Nos. 51778262, 51708241, and 71571078) are acknowledged for their financial support of this research.

References (109)

  • Y. Feng et al.

    Exploring the interactive effects of safety investments, safety culture and project hazard on safety performance: An empirical analysis

    Int. J. Project Manage.

    (2014)
  • B. Fernandez-Muniz et al.

    Relation between occupational safety management and firm performance

    Saf. Sci.

    (2009)
  • M.A. Griffin et al.

    How leaders differentially motivate safety compliance and safety participation: the role of monitoring, inspiring, and learning

    Saf. Sci.

    (2013)
  • L.A. Hadidi et al.

    Loss prevention in turnaround maintenance projects by selecting contractors based on safety criteria using the analytic hierarchy process (ahp)

    J. Loss Prev. Process Ind.

    (2015)
  • J. Hinze et al.

    Leading indicators of construction safety performance

    Saf. Sci.

    (2013)
  • I.Y. Hsu et al.

    Analysis of business safety performance by structural equation models

    Safety Sci.

    (2012)
  • A.E. Keegan et al.

    Transformational leadership in a project-based environment: a comparative study of the leadership styles of project managers and line managers

    Int. J. Project Manage.

    (2004)
  • T.H. Li et al.

    Evaluating stakeholder satisfaction during public participation in major infrastructure and construction projects: a fuzzy approach

    Autom. Constr.

    (2013)
  • C.-S. Lu et al.

    Safety leadership and safety behavior in container terminal operations

    Saf. Sci.

    (2010)
  • M. Martínez-Córcoles et al.

    Leadership and employees’ perceived safety behaviours in a nuclear power plant: a structural equation model

    Saf. Sci.

    (2011)
  • M. Martínez-Córcoles et al.

    Empowering team leadership and safety performance in nuclear power plants: a multilevel approach

    Saf. Sci.

    (2013)
  • M. Martinez-Córcoles et al.

    Linking empowering leadership to safety participation in nuclear power plants: a structural equation model

    J. Saf. Res.

    (2012)
  • A. Mengolini et al.

    Effectiveness evaluation methodology for safety processes to enhance organisational culture in hazardous installations

    J. Hazard. Mater.

    (2008)
  • F. Nasirzadeh et al.

    Dynamic modeling of the quantitative risk allocation in construction projects

    Int. J. Project Manage.

    (2014)
  • S. Olander et al.

    Evaluation of stakeholder influence in the implementation of construction projects

    Int. J. Project Manage.

    (2005)
  • H.-C. Seo et al.

    Analyzing safety behaviors of temporary construction workers using structural equation modeling

    Saf. Sci.

    (2015)
  • S. Sinelnikov et al.

    Using leading indicators to measure occupational health and safety performance

    Safety Sci.

    (2015)
  • C. Smallman et al.

    British directors perspectives on the impact of health and safety on corporate performance

    Saf. Sci.

    (2001)
  • C. Tam et al.

    Identifying elements of poor construction safety management in China

    Saf. Sci.

    (2004)
  • A.K. Tyssen et al.

    The challenge of transactional and transformational leadership in projects

    Int. J. Project Manage.

    (2014)
  • G.M. Waehrer et al.

    Costs of occupational injuries in construction in the united states

    Accid. Anal. Prev.

    (2007)
  • C. Wu et al.

    Roles of owners' leadership in construction safety: The case of high-speed railway construction projects in China

    Int. J. Project Manage.

    (2015)
  • C. Wu et al.

    How safety leadership works among owners, contractors and subcontractors in construction projects

    Int. J. Project Manage.

    (2016)
  • Afp, 2008. 21 dead in china subway accident: State media. November 15, 2008....
  • R.P. Bagozzi et al.

    On the evaluation of structural equation models

    J. Acad. Market. Sci.

    (1988)
  • J. Barling et al.

    Development and test of a model linking safety-specific transformational leadership and occupational safety

    J. Appl. Psychol.

    (2002)
  • B.M. Byrne

    Structural Equation Modeling with Amos: Basic Concepts, Applications and Programming

    (2001)
  • B.M. Byrne

    Testing for multigroup invariance using amos graphics: a road less traveled

    Struct. Equ. Model.

    (2004)
  • A.P. Chan et al.

    Factors affecting the success of a construction project

    J. Constr. Eng. Manage.

    (2004)
  • A.T. Chan et al.

    Impact of perceived leadership styles on work outcomes: Case of building professionals

    J. Constr. Eng. Manage.

    (2005)
  • M. Chemers

    An Integrative Theory of Leadership, Lawrence Elrbaum Associates

    (1997)
  • Y.Q. Chen et al.

    Risk perception and propensity in bid/no-bid decision-making of construction projects

    Eng. Constr. Archit. Manage.

    (2015)
  • Chinda, T., Mohamed, S.a.M., 2007. Causal relationships between enablers of construction safety culture. In:...
  • Daily, P., 2015. A collapse accident on construciton site in jiangxi....
  • A. Diamantopoulos et al.

    Formative versus reflective indicators in organizational measure development: a comparison and empirical illustration

    Br. J. Manage.

    (2006)
  • P.A. Dion

    Interpreting structural equation modeling results: a reply to martin and cullen

    J. Bus. Ethics

    (2008)
  • J.A. Erickson

    Corporate culture: the key to safety performance

    Occupational Hazards

    (2000)
  • Fabrigar, L.R., Wegener, D.T., 2011. Exploratory factor analysis Oxford University...
  • D. Fang et al.

    Impact of the supervisor on worker safety behavior in construction projects

    J. Manage. Eng.

    (2015)
  • J.M. Finney

    Indirect effects in path analysis

    Sociol. Method. Res.

    (1972)
  • Cited by (43)

    • Trustworthy AI and robotics: Implications for the AEC industry

      2022, Automation in Construction
      Citation Excerpt :

      Similarly, Liu et al. used a combination of SEM and EFA to identify the critical risk factors and advise safety control in a complex metro construction project [94]. Zhang et al. created a SEM-based strategy to determine dynamic safety leadership roles during different phases of a construction project, which enabled stakeholders to better assign responsibility for safety management [95]. There is also a different strategy that can be used for reasoning.

    • Organizational factors and specific risks on construction sites

      2022, Journal of Safety Research
      Citation Excerpt :

      The contribution of F3, embracing internal organization structure and job planning, is very important because it reaches significant effects on 8 of 10 risk indicators (except on RV2 and RV4). These results are consistent with the relationships analyzed by other studies that related aspects of the organizational structure and internal planning of the work with impacts on the performance of the H&S (Abas et al., 2020; Bavafa et al., 2018; Dillon et al., 2017; Forteza et al., 2017; Grill & Nielsen, 2019; Gunduz & Laitinen, 2017; Zhang et al., 2018). The highest relationship between F3 and risk indicators corresponds to fall of height (RV5, r2 = 0.74), followed by Process (RV7, r2 = 0.53).

    • Investigating the impact of emerging technologies on construction safety performance

      2024, Engineering, Construction and Architectural Management
    View all citing articles on Scopus
    View full text