Safety Leadership and Safety Performance: An integrative Model for Ghanaian Construction Industry ()
1. Introduction
Occupational health and safety (H&S) continue to be a significant global issue in the construction industry. Despite advancements in legislation, technology, and training, the construction sector continues to have a disproportionately high rate of accidents, injuries, and fatalities (Fang et al., 2020). Structural characteristics include disjointed supply chains, transient workgroups, extensive subcontracting, and variable site conditions that create inherently high-risk scenarios (Wu et al., 2015; Xia et al., 2020). These dangers have considerable human and economic repercussions, such as reduced production, legal disputes, reputational damage, and fatalities (Debela et al., 2022; Rezagholi, 2023; Roche et al., 2024).
Leadership is widely recognised as a key determinant of safety outcomes. Effective leaders not only ensure compliance with regulations but also model safe behaviours, encourage engagement, and cultivate a positive safety culture (Lingard et al., 2019). Such leadership fosters accountability, trust, and proactive safety behaviour; whereas weak leadership is associated with disengagement, low compliance, and elevated accident rates (Mullen et al., 2011; Wu et al., 2022). However, the specific processes via which leadership affects measurable safety outcomes—particularly through mediators such as LMX, the safety climate, and organisational culture—remain insufficiently investigated across many settings (e.g., Shen et al., 2017; He et al. 2021; Sun et al., 2022).
Ghanaian construction industry significantly contributes to national development; however, it is characterised by hazardous practices, inadequate regulatory enforcement, and insufficient institutional oversight (Kheni & Afatsawu 2022). Cultural and socio-institutional factors—such as hierarchical authority, significant power distance, and collectivist norms—may further hinder the effectiveness of safety leadership, as workers often demonstrate superficial compliance and are hesitant to report hazards (Danso et al., 2023, Kheni et al., 2007, Sherratt & Aboagye Nimo, 2022).
Recent research on safety leadership in Ghana has mostly concentrated on non-construction sectors (e.g., Addo & Dartey-Baah, 2020; Dartey-Baah et al., 2021; Ansong et al., 2025; Adotey et al., 2025), thereby creating a gap in understanding leadership-mediated safety performance in the construction industry. Although substantial evidence highlights safety shortcomings in SMEs, empirical knowledge of safety performance on large, high-risk building sites remains limited. This research explores leadership trajectories across different contractor tiers in Ghana.
Global studies on leadership and construction safety often draw generalised conclusions that presuppose stable regulations and uniform enforcement—characteristics that are not applicable in Ghana. Empirical research indicates that safety results in Ghana are influenced by inadequate institutional monitoring and elevated power-distance cultural norms that affect leadership dynamics (Kheni & Afatsawu, 2022; Danso et al., 2023). These circumstances promote surface compliance rather than deeply internalised safety practices (Sherratt & Aboagye Nimo, 2022), thereby constraining the efficacy of conventional safety interventions—such as compliance-based inspections, rule-driven safety training and top-down enforcement in resource-limited and disjointed construction settings (Fang et al., 2020; Wu et al., 2015; Xia et al., 2020).
This study develops and validates an integrative, Ghanaian-centric model that integrates leadership behaviour (LB), leader-member exchange (LMX), safety climate (SC), safety culture (SCult), and worker safety behaviour (WSB) to address these shortcomings. The model directly connects these dimensions to leadership’s influence on safety performance (HSP), therefore identifying practical strategies for enhancing compliance, participation, and overall safety in high-risk construction settings. Against this background, the study was conducted based on the following objectives:
1) To examine the influence of LB and LMX on employee HSP.
2) To evaluate the impact of LB on SCult. and SC, including intermediary impacts on HSP.
3) To validate an integrative model linking LB, LMX, SC, SCult., and WSB to construction safety performance in Ghana.
This study explicitly contextualises LB and its mediators within Ghana’s socio-cultural and institutional framework, generating novel insights into the mechanisms that translate leadership actions into measurable safety outcomes; thus, it offers both theoretical contributions and practical guidance for improving construction safety performance.
2. Literature Review
2.1. Theoretical Framework
This research synthesises leadership, relational, motivational, and cultural theories to elucidate the influence of leadership on safety performance in Ghana’s construction industry. Transformational and transactional leadership theories serve as the fundamental processes within the framework, illustrating how leaders affect employee safety behaviour via role modelling, communication, feedback, and enforcement. Transformational leadership elucidates discretionary, proactive safety behaviours that go beyond basic standards, whereas transactional leadership pertains to compliance-driven acts facilitated by incentives and remedial measures (Bass, 1985; Clarke, 2013; Kelloway et al., 2012). Collectively, these ideas establish leadership behaviour (LB) as a crucial precursor to the model.
Ethical leadership emphasises moral and relational dimensions. By prioritising justice, accountability, and employee wellness, ethical leadership cultivates trust, enhances leader-member exchange (LMX), and creates a safe climate, therefore encapsulating the relational aspect through which leadership influences behaviour (Brown & Treviño, 2006). Social Exchange Theory (SET) clarify the connection between leadership and Leader-Member Exchange (LMX) in relation to voluntary employee safety participation, demonstrating how supportive leadership fosters reciprocal safety engagement instead of enforced compliance.
Motivational theories, such as expectancy theory, planned behaviour theory, and social cognitive theory, illustrate individual-level processes. They elucidate how leadership and relationship dynamics influence employees’ motivation, self-efficacy, outcome expectations, and perceived control over safety behaviours (Vroom, 1964; Ajzen, 1991; Bandura, 1986). These theories validate the inclusion of workers’ safety behaviour (WSB) in the model as an outcome indicative of both compliance and proactive participation.
Cultural and identity-based theories contextualise behaviour within Ghana’s sociocultural framework. Social Identity Theory expounds how leadership and safety climate foster a collective safety identity, thereby reinforcing norms and communal commitment, while Human Capital Theory emphasises the significance of skill development in facilitating safe behaviour (Tajfel, 1972; Becker, 1964; Fang et al., 2022). These ideas support safety culture (SCult) as a higher-order concept that encompasses shared norms and competencies.
This integration connects leadership behaviour (LB) to relational quality (LMX), perceptual context (safety climate), cultural norms (SCult), and individual safety behaviour (WSB), offering a multi-level elucidation of safety performance. This framework is particularly relevant in Ghana’s hierarchical and resource-constrained construction industry, where insufficient enforcement and superficial compliance necessitate relational trust, collective identity, and intrinsic motivation to attain effective safety outcomes (Danso et al., 2023; Kheni et al., 2007; Sherratt & Aboagye Nimo, 2022).
2.2. Empirical Review
Leadership is recognised as a critical determinant of safety performance in construction, impacting outcomes through communication, trust, engagement, and compliance (Wu et al., 2018; Lingard et al., 2019). Transactional leadership enhances compliance via regulations, oversight, and corrective measures, but transformational leadership—particularly when bolstered by a high-quality LMX—fosters profound behavioural change that transcends mere conformity to rules (Shen et al., 2017).
Global studies indicate contextual variations in the operation of these processes. In mainland China, strong leader-member exchange (LMX) enhances safety behaviour by positively affecting both the safety climate and employees’ psychological capital (He et al., 2021; Zhang et al., 2018). Similarly, evidence from the China-Hong Kong region, which functions under a unique administrative and occupational safety regulatory framework distinct from mainland China, demonstrates that supervisors’ safety-related behaviours substantially impact workers’ safety performance and their perceptions of the safety climate (Zhang et al., 2017). Research from Iran underscores management commitment as a cornerstone of organisational safety culture, highlighting the need for ongoing leadership engagement (Tehrani et al., 2019; Ardeshir & Mohajeri, 2018). In South Africa, transformational leadership also enhances safety culture and compliance (Skeepers & Mbohwa, 2015). Evidence from Ghana indicates that safety climate modulates the impact of leadership on employee behaviour (Novieto, 2023).
Research from other high-risk industries corroborates these tendencies. Transformational leadership regularly improves engagement and safety performance, whereas passive supervision detracts from results (Mullen et al., 2011; Prinsloo & Hofmeyr, 2022). Relational and cultural factors, including organisational norms and psychological resources, also promote compliance and performance (Gracia et al., 2020; Supardi et al., 2021).
LMX is pivotal in enhancing staff engagement, influencing safety climate, and bolstering psychological capital (Bentoy et al., 2022; He et al., 2021; Sun et al., 2022). The climate and culture enhance performance via supervisory signals, peer impact, and collective norms (Gao et al., 2016; Xia et al., 2020; Fang et al., 2022; Adzivor et al., 2024).
Safety performance is eventually affected by the interplay of leadership, relationship quality, safety perceptions, safety values, and workers’ actions within the construction system, as shown in Table 1.
Table 1. Key constructs, definitions, levels, references and expected influences.
Construct/ Variable |
Definition/Description |
Level of Analysis |
Key References |
Expected Influence on Safety Performance |
Leadership Safety Behaviours |
Enhance safety through communication, exemplification, oversight, mentoring, and assistance. |
Organisational/ Supervisor |
Fang et al. (2020), Lingard et al. (2019), Wu et al. (2022). |
Direct beneficial impact; indirect influence via LMX, Climate, and culture |
LMX |
The quality of leader-worker relationships, founded on trust, respect, and mutual support, facilitates the expression of safety concerns. |
Dyadic (Supervisor-Employee) |
Graen & Uhl-Bien (1995), He et al. (2021), Bentoy et al. (2022). |
Exerts a more pronounced influence on participation than on compliance, and it modulates the dynamics of leadership. |
Safety Climate |
The collective views of employees reflect the management’s emphasis and dedication to safety. |
Group/ Organisational |
Zohar (2002), Gao et al. (2016), Xia et al. (2020) |
Direct beneficial impact by facilitating leadership and enhancing LMX. |
Safety Culture |
Fundamental ideals and standards directing shared duty and accountability for safety. |
Organisational |
Cooper (2000), Fang et al. (2022), Adzivor et al. (2024). |
The role of the distal mediator is to support sustained performance throughout time. |
Workers’ Safety Behaviours |
Occupational Safety Practices require strict adherence to regulations and active participation in voluntary initiatives such as reporting, peer assistance, and proactive measures. |
Individual |
Neal & Griffin (2006), Sun et al. (2022), Wu et al. (2018) |
A direct predictor serves as a conduit for leadership, climate, and culture. |
Safety Performance |
Include accident reduction, regulatory compliance, proactive involvement, and site enhancements. |
Organisational/Site |
Ta et al. (2022), Khan et al. (2018), Novieto (2023). |
All other constructs influence the dependent variable. |
Ghana garners less academic attention, but worldwide data underscores the high-risk characteristics of its construction industry. This study emphasises the need for context-specific research that incorporates leadership dynamics, LMX, SC, SCult., and employee behaviour. It contends that leadership and organisational culture influence essential safety conditions—such as equipment availability and regulatory compliance—that promote behavioural transformation. The study offers practical insights by investigating the relational and contextual factors that affect safety outcomes; therefore, it addresses a notable gap in the current literature.
2.3. Hypotheses Development
Leadership influences construction safety through relational, cultural, and behavioural channels. Leaders who articulate standards, exemplify safe behaviours, and exhibit dedication cultivate trust, improve team collaboration, and integrate safety as a collective value, thereby reinforcing the underpinnings of safety performance (Fang et al., 2022; Lingard et al., 2019). Consequently:
H1: Safety leadership positively influences LMX.
H2: Safety leadership positively influences safety climate.
H5: Safety leadership positively influences overall safety performance.
H6: Safety leadership positively influences safety culture.
H7: Safety leadership positively influences workers’ safety compliance behaviour.
H8: Safety leadership positively influences workers’ safety participation behaviour.
LMX assesses the quality of leader-subordinate relationships via trust and reciprocal support. Superior exchanges promote transparency and voluntary participation (He et al., 2021; Sun et al., 2022), especially within hierarchical settings such as Ghana’s construction sector.
Safety climate represents workers’ impressions of management’s safety goals, while safety culture signifies persistent norms that govern behaviour (Zohar, 2002; Gao et al., 2016; Adzivor et al., 2024). A favourable climate and a developed culture enhance compliance, promote participation, and maintain performance.
H4: Safety climate positively influences overall safety performance.
H10: Safety climate positively influences workers’ safety participation behaviour.
H12: Safety climate positively influences workers’ safety compliance behaviour.
H13: Safety climate positively influences workers’ safety performance.
H14: Safety culture positively influences overall safety performance.
Employee compliance and engagement serve as behavioural conduits via which leadership, LMX, the organisational environment, and culture influence performance (Neal & Griffin, 2006; Wu et al., 2018). Compliance guarantees obedience to procedures, while participation signifies proactive enhancement of safety.
Source: Authors’ work.
Figure 1. Conceptual framework for the study.
Figure 1 illustrates the proposed integrative safety leadership model, showing leadership as the central driver influencing safety performance directly and indirectly via LMX: safety climate, culture, participation, and compliance, summarising the hypothesised pathways for subsequent SEM and qualitative analyses.
3. Methodology
This research was performed in six regions of Ghana—Greater Accra, Ashanti, Central, Western, Northern, and Upper East—to include varied construction circumstances, worker profiles, and safety protocols. Greater Accra (91.7% urbanised, 154,357 workers) and Ashanti (106,902) are significant centres, whilst Central, Western, Northern, and Upper East (59,613; 29,738; 20,444; 11,578) exhibit moderate to smaller rural labour forces with inadequate safety infrastructure (GSS, 2021). The inclusion of Accra and Kumasi, crucial construction hubs, enhanced representativeness (Simpson & Sam, 2020).
A mixed-methods approach based on pragmatism, supplemented by critical realism (CR), was used. Pragmatism facilitated actionable insights while recognising diverse realities (Creswell, 2008; Greene, 2007, p. 20, as cited in Leavy, 2017: p. 168), whereas critical realism offered an ontological perspective to investigate how outcomes arise from institutional, organisational, and behavioural mechanisms, employing the empirical-actual-real framework and context-mechanism-outcome heuristic (Park & Peter, 2022; Tennant et al., 2020).
A convergent parallel design of equal status simultaneously gathered quantitative surveys and qualitative interviews (see Table S2 and Appendix 7-9 for the details of the quantitative and qualitative data items), analysed them separately, and merged them to improve triangulation, complementarity, and analytical thoroughness (Creswell & Plano Clark, 2018; Greene et al., 1989). Integration occurred at both methodological and interpretative levels (Schoonenboom & Johnson, 2017), guaranteeing stringent cross-validation.
The analytical methodology integrates induction, deduction, and abduction. Induction facilitated the organic emergence of themes from qualitative data (Strauss & Corbin, 1998), deduction evaluated hypotheses using SEM based on theoretical frameworks (McAbee et al., 2017), and abduction explicates unforeseen outcomes by connecting empirical evidence to conceptual rationale (Thomas, 2022; Timmermans & Tavory, 2012). This hybrid logic improved theory development and encapsulated intricate behavioural processes.
A cross-sectional approach offered an overview of safety perceptions and behaviours across 79 sites, appropriate for exploratory research when experimental manipulation is unfeasible (Allen, 2017; Thompson, 2022). The target population consisted of professionals and labourers engaged in large (D1K1/A1B1—the highest contractor rating for buildings and roads) and medium-sized (D2K2/A2B2—the second-tier classification) construction projects across six regions of Ghana, as per the official contractor ranking system that categorises firms based on financial and technical capabilities (Ayarkwa et al., 2012). The participants were engineers, architects, project managers, safety officials, supervisors, consultants, and artisans. Non-probability sampling methods—purposive, expert, maximum variation, convenience, and snowball techniques—were used owing to obsolete records, inaccessible locations, and the need to engage key informants (Emmel, 2013; Etikan & Bala, 2017; Parker et al., 2019). Sites were chosen based on operational activity, project magnitude, and observable safety protocols, with participant’s participation requiring a minimum of six months of site experience.
The data collection utilised structured, validated questionnaire items from Clarke and Ward (2006), Wu et al. (2018), Zhang et al. (2020), Sun et al. (2022), Ni et al. (2020), Wang et al. (2021), Zahoor et al. (2017), Christian et al. (2009), and Yu et al. (2021) to evaluate LB, LMX, WSC, WSP, SC, SCult., and HSP (see Table 2 for the items). All questions were evaluated using 7-point Likert scales (Yaddanapudi & Yaddanapudi, 2019), with a range from 1 = strongly disagree to 7 = strongly agree. The biodata and personal records section initiated the questionnaire, documenting respondents’ gender, age range, educational attainment, employment position, business classification and type, firm/site location, and years of experience with both the present employer and the broader construction sector. The semi-structured interview guide (Appendix 8 and Appendix 9), formulated in accordance with Kallio et al. (2016) for both management and employees, included sections addressing the interviewee’s biographical details, leadership behaviours, their impact on employees, satisfaction levels, principal challenges, and suggestions. Both instruments included preambles stating the study’s purpose, focus, expected outcomes, respondent’s role and ethical assurances.
Reliability and validity were established via pretesting, pilot testing, Cronbach’s alpha, composite reliability, and AVE ≥ 0.5 (Jackson, 2003; Pallant, 2011; Ghazali, 2016), while qualitative credibility was enhanced through triangulation, member checking, systematic coding, and peer validation (Sandelowski, 2008).
Data were obtained from six regions; however, no interviews were performed in the Northern Region. During the survey data collection, the Northern Region had just one active construction site that satisfied the study’s inclusion requirements, resulting in a restricted number of questionnaire respondents. Therefore, data from surveys conducted in the Northern and Upper East areas were amalgamated for analysis to guarantee analytical rigour. Quantitative data (n = 534, mostly in-person surveys) were processed in SPSS version 26, with confirmatory factor analysis (CFA) and structural equation modelling (SEM) executed in Python 3.13.5, using Factor Analyser and Semopy with maximum likelihood estimation. Model fit was assessed using χ2/df, CFI, TLI, RMSEA, and SRMR (Kline, 2023; Meshcheryakov et al., 2021). Qualitative data (n = 36 interviews) were examined with thematic analysis and grounded theory (Braun & Clarke, 2006; Birks & Mills, 2015; Ju & Rowlinson, 2014), corroborated by triangulation, member validation, and peer review for credibility (Elo et al., 2014; Creswell, 2015). Findings were synthesised by triangulation (Table 3) to ascertain convergence, complementarity, and divergence (Creswell, 2015). The department review committee verbally authorised the instruments.
This methodology assessed the interactions among LB, LMX, SC, SCult., and WSB in shaping safety outcomes, yielding empirically validated, context-specific insights that enhance construction safety research in underdeveloped nations, particularly within the unique context of Ghana. Ethical standards were upheld in accordance with the Nuremberg Code, ensuring informed permission and secrecy (Shuster, 1997; Leavy, 2017; Pillay et al., 2023). Although constrained by a cross-sectional, non-probability design, the mixed-methods approach reduced bias and yielded a thorough, mechanism-sensitive comprehension of safety performance in Ghanaian construction projects.
4. Findings
4.1. Quantitative
This study evaluated a holistic safety leadership framework for Ghana’s construction industry by integrating structural equation modelling (SEM) with grounded theory (GT). SEM constituted a significant portion of the variation in safety performance (R2 = 0.628), (Figure 2), while GT illuminated the contextual, social, and environmental mechanisms behind these effects (Table 3, Figure 3). The results together demonstrate how leadership, relationship dynamics, attitudes, and culture influence safety outcomes in Ghana’s hierarchical and resource-limited construction environment.
4.1.1. Instrumentation and Demographics
Out of 650 questionnaires sent, 540 were returned, resulting in a response rate of 83.1%. After addressing the 44 item-level missing data points through modal imputation (Wright et al., 2015). 24 were retained whereas 20 were discarded. In all, 534 valid responses were analysed, including 38 online submissions through google form (Bhalerao, 2015). Thirty-six interviews were analysed using grounded theory, since one interview was removed due to incompleteness, resulting in comprehensive contextual insights. This combination ensured robust empirical and qualitative triangulation.
The predominant demographic of survey participants was male (89.7%), with ages ranging from 26 to 33 years (38.2%), and possessing either a bachelor’s degree (35.8%) or a senior high school/WASSCE/Diploma (31.1%). Roles included managers (35.8%), engineers (21.7%), and artisans (29.0%), mostly within large firms (82.6%) located in Greater Accra (67.2%). The predominant proportion of responders (78.1%) possessed fewer than five years of tenure. Moreover, local construction companies employed 50.2% of the respondents, multinational construction companies employed 42.5%, and foreign construction companies employed 7.3% (Table S1).
4.1.2. Evaluation and Assessment of Structural Model and Hypotheses
1) Validation of the measurement model
The measuring model exhibited robust reliability and validity. Internal consistency was validated using Cronbach’s α, composite reliability, and omega coefficients, ranging from 0.830 to 0.947. Convergent validity was shown by factor loadings ranging from 0.555 to 0.872 and AVE values between 0.470 and 0.664. Discriminant validity was confirmed using HTMT ratios under 0.90 and adherence to the Fornell-Larcker criterion (Fornell & Larcker, 1981). KMO values (0.762 - 0.938) and a significant Bartlett’s test (p < 0.001) confirmed sample adequacy and model appropriateness. The findings affirm the alignment of leadership, relational, perceptual, and behavioural components pertinent to safety performance in Ghana’s construction industry (see Tables S2-S6 for the details).
2) Validation of the structural model
The structural model demonstrated satisfactory fit indices: Comparative Fit Index (CFI) = 0.917, Tucker–Lewis Index (TLI) = 0.901, Root Mean Square Error of Approximation (RMSEA) = 0.079, standardised Root Mean Square Residual (SRMR) = 0.071, and chi-squared (χ2) = 98.320. This signifies that the suggested structural linkages sufficiently reflect the observed facts (Hu & Bentler, 1999). Leadership had the most significant impact on LMX (β = 0.543, p < 0.001), underscoring its crucial role in influencing relationship dynamics within construction teams. SCult. emerged as the foremost predictor of HSP (β = 0.412, p < 0.01), highlighting the essential influence of shared norms and values in promoting safe work practices. The SC, WSC, and WSP were highly interconnected, indicating that leadership influences safety performance both directly and indirectly via many channels.
3) Explained variance
The structural model accounted for significant variation in essential constructs: LMX (R2 = 0.295), safety climate (R2 = 0.487), safety culture (R2 = 0.472), participation (R2 = 0.498), and compliance (R2 = 0.531). It represented 62.8% of the variance in overall safety performance (R2 = 0.628), demonstrating substantial predictive capability. The findings underscore LB and SCult. as essential factors for improving compliance and proactive engagement and, therefore, augmenting safety outcomes on construction sites in Ghana.
4) Hypotheses testing
All fifteen proposed hypotheses (Table 2) were validated, indicating leadership’s significant influence on relational, perceptual, and behavioural processes. Leadership had the greatest significant impact on LMX (β = 0.543), while SCult. emerged as the primary predictor of HSP (β = 0.412). The relationship between SC, WSC, and WSP showed both direct and indirect effects of leadership. The link from LB to LMX and SC to WSP was very strong. Leadership, culture, climate and employee behaviour accounted for 62.8% of the variation in safety performance, highlighting the significance of leadership and cultural interventions in Ghanaian construction sites.
Table 2. Hypotheses testing result.
Hypothesis |
Path |
β |
SE |
p-value |
Result |
H1 |
LB → LMX |
0.543 |
0.078 |
<0.001 |
Strongly Supported |
H2 |
LB → SC |
0.392 |
0.091 |
<0.001 |
Strongly Supported |
H3 |
LMX → SC |
0.241 |
0.094 |
0.010 |
Supported |
H4 |
SC → HSP |
0.312 |
0.086 |
<0.001 |
Strongly Supported |
H5 |
LB → HSP |
0.218 |
0.095 |
0.021 |
Supported |
H6 |
LB → SCult. |
0.287 |
0.088 |
0.001 |
Supported |
H7 |
LB → WSC |
0.245 |
0.092 |
0.008 |
Supported |
H8 |
LB → WSP |
0.198 |
0.096 |
0.039 |
Supported |
H9 |
LMX → WSP |
0.276 |
0.089 |
0.002 |
Supported |
H10 |
SC → WSP |
0.334 |
0.082 |
<0.001 |
Strongly Supported |
H11 |
WSP → HSP |
0.245 |
0.093 |
0.008 |
Supported |
H12 |
SC → WSC |
0.298 |
0.088 |
0.001 |
Supported |
H13 |
WSC → HSP |
0.287 |
0.089 |
0.001 |
Supported |
H14 |
SCult. → HSP |
0.412 |
0.084 |
<0.001 |
Strongly Supported |
H15 |
WSP → WSC |
0.267 |
0.090 |
0.003 |
Supported |
4.1.3. Structural Equation Model Interpretation
The SEM (Figure 2) validates LB as the primary upstream determinant of safety performance in Ghanaian construction. LB is the most significant predictor of LMX (β = 0.543), highlighting relational qualities as the principal means by which leaders influence employees’ attitudes and behaviours. It also improves SC (β = 0.392) and SCult. (β = 0.287), as well as WSP (β = 0.198) and WSC (β = 0.245), indicating that leadership influence functions via both relational and operational channels. The SC serves as a fundamental mediator, substantially forecasting WSP (β = 0.334), WSC (β = 0.298), and HSP (β = 0.312). The SCult. has the most significant direct impact on performance (β = 0.412), suggesting that ingrained standards surpass transient views. Behavioural pathways are significant: WSP (β = 0.245) and WSC (β = 0.287) both directly influence performance, with WSP also enhancing WSC (β = 0.267). The model explains 62.8% of the variation in safety performance, indicating an interconnected system in which leadership influences results via relationship exchanges, safety attitudes, cultural integration, and behavioural processes. Enhancing safety in the industry requires coordinated actions across many paths rather than standalone initiatives.
![]()
Source: Authors’ work.
Figure 2. Structural equation model.
4.2. Qualitative Findings
4.2.1. Demographic and Occupational Characteristics of Participants
A total of 37 interviews were conducted; however, one (SO18) was terminated after 14 minutes, and its corresponding data was eliminated. The sample was mostly male (94.44%), illustrating the gender disparity in the construction sector, and primarily included those aged 30 - 39 (47.22%), followed by those aged 20 - 29 (25%) and 40 - 49 (22.22%), suggesting that construction sites are mainly manned by young to mid-career professionals. Educational achievement was elevated: over half (52.78%) of participants had bachelor’s degrees, 13.89% held master’s degrees, 11.11% had diplomas, and 8.33% obtained other tertiary credentials, with a minimal number owning NVTI certificates. Significantly, almost all individuals in safety-related positions (28 participants) had professional certificates such as either NEBOSH, OSHA, ISO, EOSH, IOSH, or HSE, highlighting considerable proficiency in occupational safety.
Employment roles were mostly held by safety management jobs, with safety officers (76.92%) and safety managers (15.39%) being most common, in addition to coordinators, assistants, artisans (8.33%), engineers (5.56%), and project managers or supervisors (13.89%). The construction experience varied from one to 30 years, with an average of 7.8 years. The majority of participants (77.78%) had one to ten years of experience, while 5.56% had between 11 and 20 years, creating a balance between novice and experienced experts. Interview lengths varied from 25 to 35 minutes, indicating the level of participation. Telephone interviews constituted 55.56% while the remaining took place in person at various locations or site offices. Interviews were sometimes arranged during evening hours to suit participants.
Participants primarily included local companies (over 50%), while multinationals represented about 25%, along with a smaller percentage from foreign companies, facilitating a comparative analysis of safety performance across Ghana’s contractor levels. The majority were linked to prime contractors (75%), while others were subcontractors (19.44%) and project management consultants. The projects exemplified the diversity of Ghana’s construction sector, encompassing residential (apartments, houses), commercial (offices, multipurpose facilities), institutional (hospitals, lecture halls, stadiums, sports complexes), industrial (factories, data centres), and infrastructure developments (harbours). These projects were in different phases: foundation, superstructure, finishing, and completion. Geographically, 78% were situated in Greater Accra, with supplementary projects in Ashanti (11.11%), Central (5.56%), Northern (2.78%), and Western (2.78%), confirming Greater Accra’s status as the preeminent centre of construction activity.
4.2.2. Safety Leadership’s Influence on Safety Performance
The qualitative data indicates leadership is a critical element in enhancing safety performance at construction sites in Ghana. Leaders establish safety standards, enforce regulations, educate and inspire employees, balance production requirements, and promote open communication and trust via their behaviours and strategies. These measures foster a robust safety culture that promotes adherence to safe procedures and decreases accidents. Participants, for example, made the following remarks and are categorised as follows:
Leadership and Establishing Safety Protocols: Leaders serve as essential exemplars by continually exhibiting safe behaviours. One participant said, “Leaders don personal protective equipment... Our leaders inspire and motivate us to perform” (A1), while another highlighted, “They exemplify best practices, assist in inspections and training... act as role models and mentor employees” (SM3). This observable dedication motivates employees to emulate, establishing a culture of safety adherence.
Enforcement and Accountability: Leaders uphold safety standards via vigilant monitoring and rectification of hazardous activity. For example, a worker said, “Our safety officer prohibits unsafe actions...” and issued a nonconformance to a supervisor” (A2). Another emphasised point is, “Reporting nonconformance on the WhatsApp platform... provides corrective measures and deadlines” (SM2), illustrating how accountability systems bolster safety regulations.
Continuous Safety Education: Continuous safety education is essential for augmenting workers’ knowledge and awareness. Participants acknowledged the leadership’s responsibility in delivering training and coaching: “They participate in safety meetings, deliver safety presentations, educate, motivate, and assess tasks” (SO12), and “Supply PPEs, perform regular inspections, and engage in safety coaching” (OD). These exercises enable employees to identify and alleviate threats.
Motivation and Positive Reinforcement: Leaders cultivate motivation via encouragement and rewards, thereby enhancing positive safety behaviour: “Most senior managers reward employees for exemplary safety behaviour” (SM4), and “I consistently remind employees about safety, encourage and educate them, and occasionally impose penalties” (SS). Positive reinforcement fosters dedication to safe habits.
Balancing Safety and Production Demands: Certain participants underscored the difficulties in reconciling safety with productivity pressures: “Leadership prioritises production over safety... They perceive insistence on safety as a hindrance to progress” (HSSC), whereas others accentuated the importance of effective leadership by stating, “Integrating transformational and transactional leadership styles is essential for ensuring safety at our site” (OD). This equilibrium is essential to avoid jeopardising safety for production objectives.
Communication, Collaboration, and Trust Development: Leaders promote transparent communication and collaboration, thereby augmenting shared safety accountability: “Integrating sanctions, incentives, dialogue, respect, and training... This method has improved safety adherence and performance” (SM2), and “Synchronise activities, mobilise resources, engage in open communication and collaboration” (SO4). Such inclusive methodologies enhance the safety culture on-site.
4.3. Triangulation of Results
Triangulation (Table 3) substantiates that leadership is the primary determinant of safety performance; it influences relational (LMX), perceptual (climate and culture), and behavioural (compliance and participation) processes. Quantitative findings indicate that all leadership channels are relevant; however, qualitative insights demonstrate how leaders exemplify safety, enforce regulations, communicate transparently, mentor employees, and offer ongoing training. SCult. and SC serve as essential mediators, strengthening collective norms and fostering situations that encourage WSP and WSC. WSC and WSP subsequently transform these organisational circumstances into quantifiable safety results. The comprehensive data indicates that leadership stimulates interrelated relational, cultural, and behavioural pathways that jointly improve safety performance in Ghana’s construction industry.
Table 3. Triangulation of quantitative hypotheses and qualitative insights.
Quantitative Relationship (Hypothesis) |
Statistical Outcome |
Supporting Qualitative Themes/Evidence |
Integrated Interpretation |
H5: LB → HSP (β = 0.218) |
Confirmed |
The positive behaviours of leaders (role modelling, enforcement, encouragement) significantly impact employee safety behaviour and performance. |
Leadership is the fundamental catalyst for safety performance via demonstrable commitment and the implementation of safety standards. |
H6: LB → SCult. (β = 0.287) |
Confirmed |
Leaders create a strong safety culture by honest communication, teamwork, consequences, incentives, discussion, and respect. |
Leadership cultivates a culture of collective accountability and compliance with safety protocols. |
H14: SCult. →HSP (β = 0.412) |
Strongly
Confirmed |
A culture of safety fosters adherence to regulations and decreases accidents. Employees assimilate safety rules via the reinforcement of leadership. |
An effective safety culture acts as a conduit connecting leadership to safety results. |
H2: LB→SC (β = 0.392) |
Strongly
Confirmed |
Leaders serve as exemplars by using PPE, enforcing regulations, doing inspections, supplying PPE, and mentoring employees. |
Leadership distinctly influences the workplace atmosphere (safety climate) by exemplifying and upholding safety norms. |
H7: LB → WSC (β = 0.245) |
Affirmed |
Theme of Enforcement and Accountability: Leaders oversee, restrict hazardous behaviours, and generate nonconformance reports. |
Leadership mandates adherence to regulations and responsibility, hence promoting safer behaviours on-site. |
H12: SC → WSC (β = 0.298) |
Affirmed |
A favourable environment fostered by consistent leadership behaviour promotes adherence. |
Employees replicate the attitudes of their leaders, thereby converting the organisational into compliance with regulations. |
H13: WSC → HSP (β = 0.287) |
Affirmed |
Sanctions, incentives, and training improve compliance with safety protocols and results. |
Compliance functions as a behavioural mechanism that converts leadership and organisational climate into performance. |
H1: LB → LMX (β = 0.543) |
Strongly
Confirmed |
Relational leadership involves the equitable treatment of employees, fostering care, providing motivation, and cultivating trust. |
Superior leader-member connections improve communication and foster mutual respect for safety. |
H3: LMX → SC (β = 0.241) |
Confirmed |
The amicability, communication, and mentorship of leaders foster trust and collective safety standards. |
LMX functions as a social mechanism that connects leadership style to a favourable safety climate. |
H8: LB → WSP (β = 0.198) |
Affirmed |
Leaders promote and incentivise engagement by acknowledgement, education, and transparent communication. |
Motivational leadership fosters proactive employee engagement in safety initiatives. |
H10: SC → WSP (β = 0.334) |
Strongly
Confirmed |
A supportive environment fosters employee participation in idea generation and safety measures. |
An affirmative environment promotes psychological safety and proactive participation. |
Continuous Training and Education (Cross-disciplinary) |
— |
Leaders provide training sessions, toolbox discussions, and safety mentorship. |
Ongoing education maintains awareness and enhances climate, compliance, and performance. |
Production Pressure (Contextual Nuance) |
— |
Leaders sometimes prioritise output above safety. |
Emphasises the need for a harmonious blend of transformational and transactional leadership to ensure safety integrity. |
4.4. Mixed Methods Model
Source: Authors’ work.
Figure 3. Integrated conceptual model for the influence of health and safety leadership practices on health and safety performance.
The mixed-methods model (Figure 3) indicates that LB is the principal determinant of HSP in Ghanaian construction industry and functions via several mediating mechanisms. LB significantly influences LMX (β = 0.543), corroborating the qualitative findings that employees perform optimally when supervisors exhibit transparency, maintain constant communication, and acknowledge contributions. LB also influences SC (β = 0.392), indicating employees’ trust in leadership, assessment of remedial actions, and acknowledgement of safe procedures. SC impacts WSP (β = 0.334) and WSC (β = 0.298), demonstrating that perceived leadership commitment fosters safe behaviour. Aside from SC, SCult—the collective standards of zero harm and mutual accountability—exerts the most significant direct influence on HSP (β = 0.412). WSP and WSC convert these views and norms into action, substantially influencing performance (β = 0.245 and 0.287). The quantitative SEM exhibits significant explanatory power (R² = 0.628), whereas qualitative insights augment the robustness and application of these projections by contextualising leadership and safety dynamics throughout routine organisational practice. Dashed routes underscore contextual issues highlighted qualitatively—temporary contracts, informal employment, understaffing, hierarchical frameworks, and production pressures—that may undermine LB, restrict SC and SCult, and obstruct WSP or WSC. The model indicates that leadership influences safety via relational, perceptual, cultural, and behavioural processes, all of which are influenced by Ghana’s distinct sociocultural and institutional setting (Figure 3).
5. Discussion
The findings indicate that LB is the principal factor influencing HSP in Ghanaian construction projects and functions via relational, perceptual, cultural, and behavioural processes. Quantitatively, LB had the most significant impact on LMX, β = 0.543, suggesting that relationship quality is the primary mechanism by which leaders influence employees’ safety perceptions and behaviours. Qualitative data supports this, with findings demonstrating that leaders actively coach staff, exemplify safe behaviours, enforce standards, and cultivate trust, thereby inspiring compliance with safety regulations. These results correspond with findings from Mainland China region, China-Hong Kong region, Iran, South Africa, and other high-risk sectors, indicating that superior LMX improves compliance, participation, and safety outcomes (He et al., 2021; Zhang et al., 2017; Skeepers & Mbohwa 2015; Mullen et al., 2011).
SC and SCult. have emerged as crucial mediating categories. SC significantly impacted WSP (β = 0.334) and WSC (β = 0.298), but SCult. had the most substantial direct influence on health and safety performance (HSP, β = 0.412). Qualitative data indicate that leaders strengthen norms by continuous training, feedback, coaching, and acknowledgement, cultivating collective attitudes of safety and a healthy corporate culture. The results are consistent with global studies showing that climate affects short-term perceptions and actions, while culture promotes long-term adherence through established norms and shared responsibility (Zohar, 2002; Fang et al., 2022; Adzivor et al., 2024).
Behavioural mechanisms—WSP and WSC—convert perceptions and cultural norms into action. WSP enhances WSC (β = 0.267), suggesting that active participation strengthens compliance to rules. Qualitative results substantiate that positive reinforcement, coaching, supervisory oversight, and mentoring enhance both participation and compliance, corroborating findings from transformational and transactional leadership studies (Shen et al., 2017; Lingard et al., 2019).
The study furthermore revealed other secondary results that explain the context-specific dynamics of safety leadership in Ghana. Initially, participants saw the difficulty of reconciling safety with production pressures, since operational objectives may contradict compliance with safety regulations (“Leadership prioritises production over safety… insistence on safety as an impediment”—HSSC). This study indicates a moderating effect, in which production goals may diminish leadership’s impact on SC, SCult., and WSB, aligning with global findings from rapid construction settings (Wu et al., 2015; Xia et al., 2020).
Secondly, motivation and positive reinforcement surfaced as a pivotal process. Leaders promote participation and compliance by acknowledging outstanding safety behaviour, offering rewards, and occasionally enforcing penalties (“Most senior managers reward employees for exemplary safety behaviour”—SM4), in alignment with transformational leadership theories that highlight discretionary safety engagement (Lingard et al., 2019; Shen et al., 2017).
Third, communication, cooperation, and cultivating trust were emphasised as essential relationship practices. Leaders foster transparent communication, collaboration, and the strategic implementation of rewards and penalties, thereby augmenting trust and collective responsibility (SM2, SO4), which aligns with international research that demonstrates the importance of engagement and relational quality in influencing safety outcomes (Mullen et al., 2011; Bentoy et al., 2022).
Fourth, continuous safety education was considered essential for improving workers’ knowledge, attitudes, and actions. Leaders engage in safety meetings, provide information, offer coaching, and guarantee the accessibility of personal protective equipment (“They participate in safety meetings, deliver safety presentations, educate, motivate, and assess tasks”—SO12). These approaches correspond with global data that demonstrates the importance of continuous education in converting leadership into quantifiable safety results (Fang et al., 2022; Gao et al., 2016).
The study identifies structural and institutional limitations that hinder leadership effectiveness, such as short-term contracts, informal labour arrangements, understaffing, hierarchical norms, and inadequate enforcement. These challenges affect leadership’s impact on safety performance, aligning with literature on high-risk, resource-constrained environments where structural support and institutional enforcement are crucial (Wu et al., 2015; Xia et al., 2020; Kheni & Afatsawu, 2022).
The SEM indicated that leadership behaviour significantly influences safety climate, culture, and worker behaviour, collectively accounting for 62.8% of the variance in safety performance. The qualitative data clarifies this relationship by detailing how relational quality, climate, culture, and worker behaviours impact safety outcomes in Ghana’s construction sector. Overall, these findings underscore the crucial role of leadership in improving construction safety.
Research Contributions
This study enhances construction safety research by developing and verifying a context-specific safety leadership model for Ghana, which includes LB, LMX, SC, SCult., and WSB. This study situates these constructs within Ghana’s hierarchical and socio-institutional framework, thereby enhancing global leadership-safety literature and addressing deficiencies in comprehending how leadership correlates with quantifiable safety outcomes in construction settings of developing countries (Danso et al., 2023; Sherratt & Aboagye Nimo, 2022). The results indicate that relational, perceptual, cultural, and behavioural mechanisms collectively mediate the impact of leadership on safety performance, offering empirical validation for SET, transformational and transactional leadership, and motivational frameworks (Bass 1985; Kelloway et al., 2012; Blau, 1964; Ajzen, 1991). The amalgamation of SEM data with qualitative insights provides a sophisticated viewpoint seldom documented in previous studies (e.g., Addo & Dartey-Baah, 2020; Dartey-Baah et al., 2021).
The research identifies critical leverage points—LB, LMX, SC, SCult, and WSB—those organisations may affect via mentorship, coaching, training, recognition, and structured feedback to enhance compliance, participation, and overall safety performance. The results underscore the need for harmonising safety and production goals, guaranteeing that operational efficiency does not undermine compliance. Construction managers and safety officers in Ghana must therefore emphasise leadership development, foster a strong safety culture, promote peer engagement, provide regular feedback, and ensure ongoing monitoring to translate organisational policies into enhanced site safety outcomes.
Insights on policy and institutions are also apparent. Structural and regulatory limitations—temporary contracts, informal employment, hierarchical norms, and inadequate enforcement—diminish leadership efficacy, indicating areas that need improvement. Policymakers and regulators may improve safety by intensifying enforcement, formalising employment contracts, and enhancing the capabilities of supervisors and safety authorities. Incorporating leadership-focused safety protocols into legislation may enhance compliance and encourage proactive worker involvement, hence cultivating safer construction workplaces (Kheni & Afatsawu, 2022; Novieto, 2023).
This research presents evidence from a low-resource, hierarchical, and collectivist construction environment, with findings that align broadly with previous studies conducted in high-risk contexts such as Mainland China region, China Hong Kong region, Iran, and South Africa (He et al., 2021; Zhang et al., 2017; Skeepers & Mbohwa, 2015; Mullen et al., 2011). The findings clarify relational, perceptual, cultural, and behavioural processes, providing insights for other developing countries with analogous socio-cultural and institutional contexts and a framework for cross-contextual benchmarking.
By combining SEM, which explains 62.8% of the variance in safety performance, with the qualitative data that offers contextual insight into the mechanisms through which leadership influences safety performance, the study demonstrates the methodological value of the mixed-methods approach. This integration strengthens knowledge of the statistical relationships and offers a strong foundation for future research and practical safety solutions.
6. Conclusions and Recommendations
6.1. Conclusions
This study illustrates that leadership behaviour is the primary determinant of safety performance in Ghanaian construction sector, functioning via relational, perceptual, cultural, and behavioural channels. High-quality LMX, a favourable SC, an entrenched SCult., and proactive WSP and WSC account for 62.8% of the variation in HSP. Qualitative findings demonstrate how mentoring, role modelling, coaching, and open communication convert leadership into quantifiable enhancements in safety behaviour. The findings underscore the difficulties in reconciling safety with production requirements, the essential importance of motivation and positive reinforcement, and the moderating influence of structural and institutional limitations. The study offers a sophisticated, contextually aware comprehension of leadership-influenced safety performance in a construction environment of a developing nation by amalgamating quantitative data with practical contextual insights.
Limitation
Notwithstanding its merits, the research had drawbacks. The cross-sectional methodology limits causal conclusions, whereas longitudinal research might more effectively explain the progression of leadership and safety policies over time. The emphasis on large- and medium-scale projects in Ghana may restrict their applicability to small-scale projects or other industries. Social desirability bias may influence self-reported survey and interview data; however, triangulation with qualitative evidence alleviates this issue. Employing purposive and snowball sampling may result in selection bias, limiting the generalisability of the results to a wider population of construction workers and managers in Ghana. Ultimately, while structural and institutional restrictions were qualitatively analysed, their quantitative moderating effects were not investigated, suggesting a potential avenue for future study.
6.2. Recommendations
Based on the study’s empirical results and theoretical integration, the following suggestions are made to improve safety performance in Ghana’s construction sector by focusing on leadership behaviours, institutional capabilities, and future research requirements:
Construction organisations should prioritise visible, active, and supportive leadership that incorporates both transformational and transactional methodologies for practice.
Leaders must actively teach and train staff, reward safety behaviours via recognition and feedback, and cultivate trust and open communication to improve compliance and participation.
Policy and regulatory organisations, together with industry groups, should enhance enforcement mechanisms, formalise labour agreements, and provide capacity-building programmes for supervisors and safety officials. Policies that institutionalise leadership-driven safety interventions may guarantee that structural constraints—such as short-term contracts, understaffing, and hierarchical norms do not compromise safety performance.
Future study should use longitudinal designs to examine causation and assess the moderating impact of contextual variables on leadership and safety outcomes. Broadening the study across various industries, project sizes, and cultural settings would improve generalisability and offer comparative perspectives on international construction safety practices.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Data Availability
The data that support the findings for this study are available from the corresponding author upon reasonable request.
Declaration on Generative AI and AI-Enhanced Technology in the Writing Process
We used ChatGPT-5 and QuillBot to improve the wording and clarity in the preparation of this paper. Upon using these tools or service, we meticulously evaluated and revised the material as necessary and assume complete responsibility for the publication’s content.
Appendices
Appendix 1
Table S1. Demographics and occupational characteristics of respondents.
Category |
Sub-category |
Frequency |
Percent |
Gender |
Male |
479 |
89.7 |
Female |
55 |
10.3 |
Total |
534 |
100.0 |
Age Bracket |
<18 years |
1 |
0.2 |
18 - 25 years |
141 |
26.4 |
26 - 33 years |
204 |
38.2 |
34 - 41 years |
140 |
26.2 |
42 - 49 years |
39 |
7.3 |
50 - 57 years |
8 |
1.5 |
>57 years |
1 |
0.2 |
Total |
534 |
100.0 |
Qualification |
No Education |
1 |
0.2 |
Primary |
2 |
0.4 |
Junior High/MLSC |
16 |
3.0 |
Inter/Senior High/WASSCE |
166 |
31.1 |
Advance/CTC |
47 |
8.8 |
NEBOSH |
3 |
0.6 |
HND |
67 |
12.5 |
Bachelor |
191 |
35.8 |
Masters |
41 |
7.7 |
Total |
534 |
100.0 |
Work Schedule |
Managers |
38 |
7.1 |
Engineers |
116 |
21.7 |
Supervisors |
191 |
35.8 |
Artisans |
155 |
29.0 |
Unskilled Personnel |
34 |
6.4 |
Total |
534 |
100.0 |
Company Classification |
D1K1/A1B1 |
441 |
82.6 |
D2K2/A2B2 |
75 |
14.0 |
D3K3/A3B3 |
15 |
2.8 |
D4K4/A4B4 |
2 |
0.4 |
Undisclosed |
1 |
0.2 |
Total |
534 |
100.0 |
Project Site Location |
Greater Accra |
359 |
67.2 |
Ashanti |
105 |
19.7 |
Western |
38 |
7.1 |
Northern |
24 |
4.5 |
Central |
8 |
1.5 |
Total |
534 |
100.0 |
Company Category |
Local |
268 |
50.2 |
Foreign |
39 |
7.3 |
Multinational |
227 |
42.5 |
Total |
534 |
100.0 |
Years with Current Company |
<1 year |
137 |
25.7 |
1 - 5 years |
280 |
52.4 |
6 - 10 years |
79 |
14.8 |
11 - 15 years |
25 |
4.7 |
16 - 20 years |
6 |
1.1 |
21 - 25 years |
5 |
0.9 |
>30 years |
2 |
0.4 |
Total |
534 |
100.0 |
Years in the Construction Industry |
<1 year |
64 |
12.0 |
1 - 5 years |
207 |
38.8 |
6 - 10 years |
125 |
23.4 |
11 - 15 years |
72 |
13.5 |
16 - 20 years |
23 |
4.3 |
21 - 25 years |
23 |
4.3 |
26 - 30 years |
15 |
2.8 |
>30 years |
5 |
0.9 |
Total |
534 |
100.0 |
Appendix 2
Table S2. Factor loadings.
Construct |
Item and Code |
Factor Loading |
Leadership Behaviours (6 items) |
Integration of Safety and Production Goals (LSBLL1) |
−0.767 |
Involvement in Safety Activities (LSBLL2) |
−0.838 |
Positive Safety Climate (LSBLL3) |
−0.784 |
Discussion of Values with Workers (LSBLL4) |
−0.734 |
Hierarchical Reward and Punishment System (LSBLL5) |
−0.737 |
Task Understanding and Motivation Enhancement LSBLL6 |
−0.751 |
Leader-Member Exchange (11 items) |
Workers respect for knowledge and competence (LMX1) |
0.703 |
Worker’s perception of friendship (LMX2) |
0.634 |
Worker’s willingness to apply extra efforts for safety (LMX3) |
0.730 |
Workers proactive engagement in safety effort (LMX4) |
0.712 |
Worker’s willingness to invest in safety (LMX5) |
0.719 |
Workers admiration for professional skills in safety (LMX6) |
0.749 |
Worker’s awareness of my satisfaction of safety (LMX7) |
0.772 |
Workers understanding of job problems and needs (LMX8) |
0.726 |
Recognition of workers potential (LMX9) |
0.696 |
Worker’s perception of supportive authority LMX10 |
0.691 |
Workers’ confidence in leadership (LMX11) |
0.629 |
Worker Safety Compliance (8 items) |
Adherence to Safety Construction Rules (WSBSC1) |
0.817 |
Timely Reporting of Risks and Incidents (WSBSC2) |
0.734 |
Participation in Safety Meetings and Training (WSBSC3) |
0.779 |
Compliance with Safety Procedures: (WSBSC4) |
0.872 |
Use of Safety Equipment (WSBSC5) |
0.850 |
Application of Risk Reduction Techniques (WSBSC6) |
0.801 |
Implementation of Management Regulations (WSBSC7) |
0.842 |
Avoidance of Over-Familiarity Issues (WSBSC8) |
0.815 |
Worker Safety Participation (10 items) |
Encouragement of Peer Involvement (WSBSP1) |
0.809 |
Prevention of Safety Violations (WSBSP2) |
0.824 |
Reporting Safety Matters to Management (WSBSP3) |
0.765 |
Collaboration on Safety Improvements (WSBSP4) |
0.769 |
Reporting Driving Hazards (WSBSP5) |
0.773 |
Monitoring Colleague Compliance (WSBSP6) |
0.758 |
Alerting Co-workers to Safety Concerns (WSBSP7) |
0.802 |
Promotion of Safety Programmes (WSBSP8) |
0.863 |
Assistance in Hazardous Conditions (WSBSP9) |
0.784 |
Participation in Safety Planning and Evaluation (WSBSP10) |
0.732 |
Safety Climate (9 items) |
Perception of Company Commitment to Health and Safety (SC1) |
0.786 |
Adequacy of Health and Safety Training (SC2) |
0.766 |
Consistent Use of PPEs (SC3) |
0.733 |
Team Commitment to Health and Safety (SC4) |
0.809 |
Quality of Communication About Health and Safety (SC5) |
0.834 |
Availability of Resources for Health and Safety (SC6) |
0.821 |
Reasonableness of Time Pressures (SC7) |
0.753 |
Peer Reactions to Safety Violations (SC8) |
0.754 |
Prohibition of Working with Defective Equipment (SC9) |
0.637 |
Safety Culture (11 items) |
Effectiveness of Training and Supervision for Safe Work Delivery (SCult 1) |
0.738 |
Adherence to Safe Work Procedures (SCult 2) |
0.774 |
Involvement of Workers in Safety Procedures (SCult 3) |
0.793 |
Regular Reviews and Updates of Safety Procedures (SCult 4) |
0.845 |
Effective Consultation on Safety Issues (SCult 5) |
0.766 |
|
Safety Reporting Procedures for Incidents (SCult 6) |
0.820 |
Follow-Up on Safety Incident Reports (SCult 7) |
0.814 |
Review of Safety Training After Incidents (SCult 8) |
0.822 |
Management Commitment to Safety (SCult 9) |
0.787 |
Management’s Priority for Safety as Role Models (SCult 10) |
0.760 |
Clear Procedures for Reporting Injuries (SCult 11) |
0.743 |
Safety Performance (6 items) |
Fewer Safety Incidents (SP1) |
−0.713 |
Reduction in Injuries (SP2) |
−0.776 |
Decrease in Equipment Failure (SP3) |
−0.555 |
Low Economic Losses (SP4) |
−0.746 |
Improved Compliance Behaviour (SP5) |
−0.708 |
Improved Participative Behaviour (SP6) |
−0.584 |
Appendix 3
Table S3. Complete psychometric properties summary.
Construct |
No. Items |
α |
ω |
KMO |
AVE |
CR |
Variance |
Bartlett’s χ2 |
p-value |
1st Eigenvalue |
Leadership Behaviours |
6 |
0.894 |
0.592 |
0.866 |
0.592 |
0.897 |
0.659 |
1816.48 |
<0.001 |
3.956 |
LMX |
11 |
0.915 |
0.499 |
0.937 |
0.499 |
0.916 |
0.544 |
2973.81 |
<0.001 |
5.989 |
Worker Safety Compliance |
8 |
0.940 |
0.664 |
0.930 |
0.664 |
0.940 |
0.705 |
3378.12 |
<0.001 |
5.642 |
Worker Safety Participation |
10 |
0.941 |
0.622 |
0.938 |
0.622 |
0.943 |
0.660 |
4050.20 |
<0.001 |
6.597 |
Safety Climate |
9 |
0.926 |
0.590 |
0.934 |
0.590 |
0.928 |
0.635 |
3137.59 |
<0.001 |
5.711 |
Safety Culture |
11 |
0.947 |
0.621 |
0.937 |
0.621 |
0.947 |
0.655 |
4589.94 |
<0.001 |
7.208 |
Safety Performance |
6 |
0.830 |
0.470 |
0.762 |
0.470 |
0.840 |
0.554 |
1406.17 |
<0.001 |
3.327 |
Notes:
α = Cronbach’s Alpha; ω = McDonald’s Omega; KMO = Kaiser-Meyer-Olkin
AVE = Average Variance Extracted; CR = Composite Reliability
All Bartlett’s tests significant (p < 0.001), confirming factorability
Appendix 4
Table S4. Factor loading performance analysis.
Construct |
No. Items |
Loading Range |
Mean Loading |
SD |
Lowest Loading |
Highest Loading |
Leadership Behaviours |
6 |
−0.838 to −0.734 |
−0.769 |
0.035 |
−0.838 |
−0.734 |
LMX |
11 |
0.629 to 0.772 |
0.706 |
0.041 |
0.629 |
0.772 |
Worker Safety Compliance |
8 |
0.734 to 0.872 |
0.814 |
0.041 |
0.734 |
0.872 |
Worker Safety Participation |
10 |
0.732 to 0.863 |
0.788 |
0.036 |
0.732 |
0.863 |
Safety Climate |
9 |
0.637 to 0.834 |
0.766 |
0.056 |
0.637 |
0.834 |
Safety Culture |
11 |
0.738 to 0.845 |
0.787 |
0.033 |
0.738 |
0.845 |
Safety Performance |
6 |
−0.776 to −0.555 |
−0.680 |
0.082 |
−0.776 |
−0.555 |
Interpretation: All factor loadings exceed 0.60 threshold (in absolute value), demonstrating strong indicator reliability.
Appendix 5
Table S5. Methodological quality and statistical significance.
Construct |
Sample Size |
Bartlett’s χ2 |
df |
p-value |
First Eigenvalue |
Kaiser Criterion |
Eigenvalue Ratio |
Leadership Behaviours |
534 |
1816.48 |
15 |
<0.001 |
3.956 |
Met (4.0×) |
6.4:1 |
LMX |
534 |
2973.81 |
55 |
<0.001 |
5.989 |
Met (6.0×) |
7.4:1 |
Worker Safety Compliance |
534 |
3378.12 |
28 |
<0.001 |
5.642 |
Met (5.6×) |
10.1:1 |
Worker Safety Participation |
534 |
4050.20 |
45 |
<0.001 |
6.597 |
Met (6.6×) |
9.7:1 |
Safety Climate |
534 |
3137.59 |
36 |
<0.001 |
5.711 |
Met (5.7×) |
7.7:1 |
Safety Culture |
534 |
4589.94 |
55 |
<0.001 |
7.208 |
Met (7.2×) |
10.7:1 |
Safety Performance |
534 |
1406.17 |
15 |
<0.001 |
3.327 |
Met (3.3×) |
3.6:1 |
Interpretation: All factor loadings exceed 0.60 threshold (in absolute value), demonstrating strong indicator reliability.
Appendix 6
Table S6. Merged discriminant validity matrix (HTMT and Fornell-Larcker combined).
Construct |
LB |
LMX |
WSC |
WSP |
SC |
SCult. |
SP |
Leadership Behaviours (LB) |
(0.769) |
0.708 (0.639) |
0.706 (0.646) |
0.696 (0.637) |
0.766 (0.696) |
0.714 (0.656) |
0.339 (0.262) |
Leader-Member Exchange (LMX) |
|
(0.707) |
0.728 (0.674) |
0.805 (0.748) |
0.795 (0.729) |
0.741 (0.688) |
0.394 (0.316) |
Worker Safety Compliance (WSC) |
|
|
(0.815) |
0.825 (0.774) |
0.810 (0.752) |
0.762 (0.717) |
0.363 (0.290) |
Worker Safety Participation (WSP) |
|
|
|
(0.789) |
0.817 (0.761) |
0.775 (0.733) |
0.341 (0.273) |
Safety Climate |
|
|
|
|
(0.809) |
0.864 (0.809) |
0.392 (0.314) |
Safety Culture |
|
|
|
|
|
(0.788) |
0.466 (0.384) |
Safety Performance |
|
|
|
|
|
|
(0.685) |
Threshold: HTMT < 0.85 (conservative) or < 0.90 (liberal). Assessment: Most values below 0.85, with Climate-Culture (0.864) slightly above, suggesting generally adequate discriminant validity; Fornell-Larcker Criterion: The square root of AVE (diagonal) should exceed inter-construct correlations in parentheses. Assessment: Some correlations approach diagonal values (e.g., Climate-Culture: 0.809 vs √AVE: 0.768), indicating potential overlap but acceptable discriminant validity overall.
Appendix 7: Preamble for Data Collection
Akenten Appiah-Menka University of Skill Training and Entrepreneurial Development
Faculty of Technical Education
Department of Construction Technology and Management
Dear Respondent,
An integrative model for health and safety performance improvement of construction project sites in Ghana is the theoretical basis for our research at the aforementioned University. This study primarily not only seek to document but also to educate key figures and other stakeholders and in the building sector how Safety Leadership practices affect safety behaviours at construction sites and the need for improvement. The study subsequently intends to develop an integrative predictive model that takes into account Safety leadership behaviour, employee behaviour, and the extent to which such behaviours enable the proper Safety climate and culture which ultimately impact on the improvement of Safety performance at construction project sites and in the construction industry as a whole.
You are therefore cordially asked to answer the questions honestly and completely as a stakeholder in the construction industry in order to get a reliable result that will aid in improving the performance of construction health and safety. Please understand that this is primarily for academic purposes, and that any information you provide will be kept totally private and anonymous. However, if at any time during the study you change your mind, you are free to do so.
We appreciate your assistance in advance.
Appendix 8: Interview Guide for Construction Leadership
Section A: Bio Data and Personal Records of Interviewee
1) Please can you briefly tell me about yourself in terms of your
a) Age bracket ………………………………………………………………………………
b) Highest academic qualification…………………………………………………………
c) Years of experience in the construction industry………………………………………
d) Current company and your job schedule………………………………………………
Section B: Leadership Behaviour and Its Influence on Workers
2a) Please, what are some of the leadership behaviours at your sites in terms of:
1) Managing health and safety delivery?....................................................................................
2) Influencing health and safety of workers with their leadership style?............……………
2b) Please, how do your safety leaders’ behaviours influence your workers’ safety behaviours in terms of safety compliance and safety participation at your sites?.....................................
2c) Please, how do the leadership behaviours at your sites influence the health and safety performance at your site?...............................................................................................................
Section C: Satisfaction
3a) Please, how satisfied are you with regards to the safety behaviour and the safety performance at your sites?............................................................................................................
Section D: Challenges
4) Please, what are some of the challenges you face in health and safety delivery at your site?....................................................................................................................................................
Section E: Recommendations
5) Please, kindly state your recommendations if any in an attempt to improve the health and safety performance at your site……………………………………………………………
Appendix 9: Interview Guide for Construction Workers
Section A: Bio Data and Personal Records of Interviewee
1) Please can you briefly tell me about yourself in terms of your
a) Age bracket …………………………………………………………………………………
b) Highest academic qualification……………………………………………………………
c) Years of experience in the construction industry…………………………………………
d) Current company and your job schedule…………………………………………………
Section B: Safety Leadership Behaviour and Their Influence
2a) Please, how do your safety leadership (managers, supervisors and gang leaders) behave towards health and safety management and performance at your sites?.............……………
2b) How do their behaviours impact your own health and safety behaviour and that of your coworkers?...................................................……………………………………………………
Section C: Satisfaction
3) Please, how satisfied are you with regards to the safety behaviour and the safety performance at your sites?.......................................................................................................................
Section D: Challenges
4) Please what challenges do you face at your site as regards health and safety delivery?
………………………………………………………………...……………………………
Section E: Recommendations
Please indicate your recommendations as to how best health and safety performance can be holistically improved at your sites?............…………………………………………………