The relationships between multidimensional construction project risk exposure and overall risk-control strategy effectiveness in Oman: a quantitative PLS-SEM study

1 Introduction

Construction projects are exposed to uncertainty through long implementation periods, tightly interdependent activities, fragmented contractual arrangements, and changing external conditions. These characteristics allow disruption in one part of the project system to spread into cost, time, quality, safety, compliance, and organizational-continuity consequences. The present study organizes these uncertainties under the overarching independent-variable domain of Construction Project Risk Exposure. This domain is multidimensional rather than unitary: it comprises financial, environmental, operational, logistical, regulatory and legal, health and safety, and market risks. Treating exposure as a multidimensional domain is important because the seven dimensions arise from different mechanisms and may be associated with management outcomes in different directions and with different magnitudes (Ale et al., 2022; Demirkesen and Tezel, 2022).

The Omani construction sector provides a particularly relevant setting for examining these relationships. National infrastructure and industrial-development programs require contractors to deliver complex projects while managing administrative procedures, labor-market constraints, supply-chain exposure, environmental obligations, and changing commercial conditions. These demands are amplified in the Duqm Special Economic Zone, where distance from established supply centers, dependence on transport and port systems, evolving institutional arrangements, and the concentration of major projects increase the need for coordinated and context-sensitive risk control. Evidence from Oman indicates that successful construction delivery depends on aligning organizational responses with local market, regulatory, and operational conditions rather than transferring generic practices without adaptation (Powmya et al., 2023).

International standards and established risk-management approaches provide general principles for identifying, assessing, and controlling project risks, but they do not indicate which dimensions of exposure have the largest conditional associations with strategy effectiveness in the specific context of Duqm. Average exposure ratings and documented control procedures are also insufficient for determining whether variation in individual risk dimensions corresponds to variation in the effectiveness of the overall control system. A quantitative assessment is therefore necessary to distinguish the perceived level of each risk category from its association with management effectiveness. Accordingly, the dependent variable in this study - Overall Risk-Control Strategy Effectiveness - captures respondents’ assessment of the integrated effectiveness of the control system across seven corresponding strategy domains. In the statistical output, this endogenous construct is labelled Overall Strategy Effectiveness (OSE).

Accordingly, this paper examines a conceptually integrated but analytically differentiated quantitative model. Construction Project Risk Exposure represents the overarching independent-variable domain, while its seven dimensions are entered as separate exogenous constructs in the PLS-SEM model to estimate their unique statistical relationships with OSE. OSE is treated as the single endogenous construct and is represented by financial, environmental, operational, logistical, regulatory, health and safety, and market strategy-effectiveness indicators. The paper has three objectives: first, to assess the perceived level of each risk-exposure dimension; second, to evaluate the perceived effectiveness of the corresponding risk-control strategy domains; and third, to determine the direction, magnitude, and statistical significance of the relationship between each risk-exposure dimension and OSE. The cross-sectional design supports association-based interpretation rather than causal inference.

2 Literature review

2.1 Theoretical foundation: integrated and context-sensitive risk management

The study is grounded in an integrated and context-sensitive view of construction risk management. A construction project is a system of interdependent technical, commercial, institutional, environmental, and human activities; consequently, exposure in one domain can alter conditions in several others. Complexity-oriented research therefore emphasizes interactions, feedback, and cascading effects rather than isolated risk lists (Caldarelli et al., 2021). Strategic risk perspectives further argue that exposure assessment must be linked to governance, resource allocation, response design, monitoring, and organizational learning if risk management is to contribute to project objectives (Hubbard, 2020; Kaplan and Mikes, 2020).

This position is particularly appropriate for Duqm because project organizations must simultaneously manage site-level uncertainty and external regulatory, market, environmental, and logistical pressures. The conceptual model therefore distinguishes two levels. At the higher conceptual level, Construction Project Risk Exposure provides the overarching independent-variable domain. At the analytical level, its seven dimensions are retained as distinct constructs to reveal whether some forms of exposure are more strongly associated with OSE than others. This balance between integration and differentiation supports a stable risk taxonomy while preserving the flexibility required to respond to local regulations, supply conditions, workforce realities, and project-specific priorities (Alblooshi et al., 2025; Salim et al., 2026).

2.2 Study variables and construct specification

The study contains one overarching independent-variable domain and one dependent variable. Construction Project Risk Exposure comprises seven dimensions: Financial Risks (FIN), Environmental Risks (ENV), Operational Risks (OPR), Logistical Risks (LOG), Regulatory and Legal Risks (REG), Health and Safety Risks (HSS), and Market Risks (MAR). Conceptually, these dimensions represent complementary forms of exposure within the broader construction-project risk environment. Analytically, they are entered as seven separate exogenous constructs rather than combined into a single higher-order path. The dependent variable is Overall Risk-Control Strategy Effectiveness, reported as Overall Strategy Effectiveness (OSE) in SmartPLS. OSE was specified reflectively as a global perception of integrated risk-control capability expected to manifest across the seven strategy domains. This specification captures shared variance in respondents’ evaluations and does not imply that the domains are operationally interchangeable. Alternative composite or formative specifications should be examined in future validation studies.

2.3 Financial risk and strategy effectiveness

Financial risk is conceptualized as exposure to inadequate or unstable funding, delayed receipts, cash-flow disruption, cost escalation, exchange-rate movements, and other conditions that threaten the liquidity and resource base required for project execution (Shibani et al., 2024). In Duqm, this exposure is especially consequential because long implementation horizons, imported inputs, geographically extended supply arrangements, and payment dependencies can transmit a financial disturbance into procurement, subcontractor continuity, scheduling, and quality. The strategic risk-management perspective adopted in the present study holds that risk information becomes useful when it is translated into decisions, resource allocation, controls, monitoring, and continuing review (Hubbard, 2020). Within the present model, Financial Risks (FIN) are therefore positioned as an exogenous dimension of Construction Project Risk Exposure, whereas Overall Risk-Control Strategy Effectiveness (OSE) is the single endogenous outcome representing the integrated effectiveness of the seven control domains. Evidence cited in the present study indicates that formal risk-management practices are associated with construction-project performance (Alsaadi and Norhayatizakuan, 2021), while focused financial controls are needed to manage payment, liquidity, and cost uncertainty (Shibani et al., 2024). Research on mega-infrastructure projects further suggests that risk-reduction strategies contribute to project success when they are aligned with material exposures (Malek and Bhatt, 2024). Accordingly, variation in financial exposure is expected to be systematically related to the effectiveness of the overall risk-control system.

H1 Financial risks have a significant relationship with overall risk-control strategy effectiveness in Omani construction projects.

2.4 Environmental risk and strategy effectiveness

Environmental risk refers to exposure created by climatic and site conditions, natural events, sustainability requirements, environmental approvals, and compliance obligations that can alter project execution (El-Sayegh et al., 2021). This dimension is important in Duqm because remote-site conditions, infrastructure development, environmental sensitivity, and coordination with public authorities can affect access, sequencing, material performance, worker protection, and regulatory continuity. From the strategic risk-management perspective used in the present study, environmental information should not remain a stand-alone compliance record; it should inform contingency capacity, cross-functional coordination, resource decisions, monitoring, and adaptive control (Hubbard, 2020). In the study model, Environmental Risks (ENV) form one exogenous dimension of Construction Project Risk Exposure and are linked directly to OSE, the single endogenous construct representing the effectiveness of the integrated response system. The present study draws on evidence that sustainable construction introduces distinct certification, material, energy-performance, and regulatory exposures that require systematic identification and control (El-Sayegh et al., 2021). Related evidence from Gulf construction settings shows that risk-management capability shapes how critical risks are translated into project outcomes (Alblooshi et al., 2025), while risk-reduction strategies are most useful when adapted to the project and risk category concerned (Malek and Bhatt, 2024). Because rising environmental exposure may activate stronger controls or exceed existing control capacity, a significant relationship with OSE is expected without prespecifying its direction.

H2 Environmental risks have a significant relationship with overall risk-control strategy effectiveness in Omani construction projects.

2.5 Operational risk and strategy effectiveness

Operational risk is conceptualized as uncertainty in the organizational and process capabilities through which project plans are converted into coordinated site execution, including planning quality, knowledge sharing, workforce capability, subcontractor coordination, design change, and equipment reliability (Khoa and Chinda, 2023). In Duqm construction projects, such exposure matters because dispersed work packages, specialist-labor requirements, time pressure, and interdependent contractors can turn a localized process failure into delay, rework, cost escalation, or quality loss. The strategic risk-management perspective informing the present study requires organizations to connect observed exposure with practical decisions, accountable routines, resource allocation, and iterative learning rather than relying on formal registers alone (Hubbard, 2020). Accordingly, Operational Risks (OPR) are specified in the paper as an exogenous dimension of Construction Project Risk Exposure that is modelled as directly associated with OSE, the integrated endogenous assessment of risk-control strategy effectiveness. Previous work cited in the present study indicates that construction competitiveness depends partly on the implementation of knowledge-management processes that support the capture and use of operational knowledge (Khoa and Chinda, 2023). Evidence from Oman also emphasizes contractor strategies that translate planning and governance requirements into project delivery practice (Powmya et al., 2023), and broader mega-infrastructure research links well-targeted risk-reduction strategies with project success (Malek and Bhatt, 2024). These arguments support the expectation that differences in operational exposure will be associated with differences in the effectiveness of the overall control system.

H3 Operational risks have a significant relationship with overall risk-control strategy effectiveness in Omani construction projects.

2.6 Logistical risk and strategy effectiveness

Logistical risk refers to exposure arising from uncertain or disrupted movement and availability of materials, equipment, and information across the project supply chain, including transport constraints, customs and port delays, supplier dependence, and coordination failures (He and Han, 2022). This construct is particularly important in Duqm because distance from established supply centers and reliance on transport, port, and import systems can leave limited recovery options when a critical input is delayed. Under the present study’s strategic risk-management perspective, logistics exposure should guide dependency mapping, supplier and inventory decisions, schedule buffers, coordination responsibilities, and ongoing monitoring (Hubbard, 2020). In the conceptual model, Logistical Risks (LOG) are positioned as an exogenous dimension of Construction Project Risk Exposure with a direct path to OSE, the single endogenous measure of the integrated control system’s effectiveness. Prior research used in the present study identifies logistics as a central source of uncertainty in overseas engineering, procurement, and construction projects and emphasizes the need to evaluate its key drivers systematically (He and Han, 2022). Omani evidence further indicates that contractor delivery strategies must be adapted to local project conditions (Powmya et al., 2023), while mega-infrastructure research suggests that aligned risk-reduction practices enhance the prospect of project success (Malek and Bhatt, 2024). Because logistical disruptions can propagate rapidly into cost, time, and operational consequences, greater logistical exposure is expected to be significantly related to overall risk-control strategy effectiveness.

H4 Logistical risks have a significant relationship with overall risk-control strategy effectiveness in Omani construction projects.

2.7 Regulatory and legal risk and strategy effectiveness

Regulatory and legal risk is conceptualized as exposure arising from permits and approvals, changing requirements, contractual interpretation, enforcement uncertainty, claims, documentation duties, and coordination with multiple authorities (Powmya et al., 2023). This dimension is central to Duqm projects because mobilization, design, procurement, labor deployment, environmental compliance, and handover can all depend on timely and consistent institutional decisions. The strategic risk-management perspective adopted in the present study implies that regulatory knowledge must be converted into governance arrangements, schedule checkpoints, contractual safeguards, assigned accountability, and continuous review (Hubbard, 2020). Within the paper’s model, Regulatory and Legal Risks (REG) constitute an exogenous dimension of Construction Project Risk Exposure and are modelled as directly associated with OSE, the single endogenous construct capturing the effectiveness of the integrated risk-control system. The present study cites Omani evidence showing that successful contractor strategies depend on adapting project delivery to local administrative and institutional conditions (Powmya et al., 2023). It also relies on evidence that formal risk-management practices are associated with construction-project performance (Alsaadi and Norhayatizakuan, 2021), while research on major infrastructure projects indicates that targeted risk-reduction strategies are associated with stronger project outcomes (Malek and Bhatt, 2024). These findings suggest that the capacity of the integrated control system cannot be separated from the regulatory and contractual exposure confronting the project; therefore, a significant relationship between REG and OSE is expected.

H5 Regulatory and legal risks have a significant relationship with overall risk-control strategy effectiveness in Omani construction projects.

2.8 Health and safety risk and strategy effectiveness

Health and safety risk refers to exposure to accidents, unsafe conditions, inadequate protective practices, weak emergency preparedness, and failures to capture and disseminate information from incidents and near misses (Al Shaaili et al., 2023). In Duqm construction projects, this exposure is important because hazardous activities, multiple contractors, demanding site conditions, and communication across work teams can affect both worker welfare and continuity of project delivery. Consistent with the strategic risk-management perspective used in the present study, safety information should support accountable decisions, preventive resource allocation, monitoring, learning, and adaptation rather than functioning only as procedural documentation (Hubbard, 2020). In the study model, Health and Safety Risks (HSS) are specified as an exogenous dimension of Construction Project Risk Exposure with a direct path to OSE, the single endogenous assessment of overall risk-control strategy effectiveness. Omani construction evidence emphasizes that the analysis and dissemination of near-miss information are important for organizational learning and preventive action (Al Shaaili et al., 2023). The present study also draws on evidence that structured risk-management practices contribute to construction-project performance (Alsaadi and Norhayatizakuan, 2021), and that project outcomes benefit when risk-reduction measures are aligned with salient exposures (Malek and Bhatt, 2024). Accordingly, differences in health and safety exposure are expected to be significantly related to the effectiveness of the integrated control system, even where baseline safety procedures are comparatively institutionalized.

H6 Health and safety risks have a significant relationship with overall risk-control strategy effectiveness in Omani construction projects.

2.9 Market risk and strategy effectiveness

Market risk refers to exposure created by changes in demand, material and labor prices, competition, financing conditions, availability of specialized resources, and wider economic volatility that can alter project assumptions and commercial viability (Al-Mhdawi, 2022). This construct is important in Duqm because large projects depend on imported inputs, specialist capacity, contractor margins, and investment conditions that may change during lengthy delivery periods. The strategic risk-management perspective applied in the present study treats such uncertainty as a decision problem requiring updated forecasts, flexible allocation of resources, scenario analysis, monitoring, and adaptive contracting (Hubbard, 2020). Within the paper’s conceptual model, Market Risks (MAR) are positioned as an exogenous dimension of Construction Project Risk Exposure and are linked directly to OSE, the single endogenous construct representing the effectiveness of the integrated control system. The present study uses evidence from construction under extreme conditions to show that economic and market uncertainty requires flexible and context-sensitive risk responses (Al-Mhdawi, 2022). Gulf construction research further indicates that risk-management capability can condition the relationship between critical risk factors and project success (Alblooshi et al., 2025), while evidence from mega-infrastructure projects supports aligning risk-reduction strategies with changing project exposures (Malek and Bhatt, 2024). Because market change can affect financing, procurement, resources, and strategic choices simultaneously, MAR is expected to have a significant relationship with overall risk-control strategy effectiveness.

H7 Market risks have a significant relationship with overall risk-control strategy effectiveness in Omani construction projects.

2.10 Conceptual framework

Figure 1 presents the study’s conceptual framework. Construction Project Risk Exposure is the overarching independent-variable domain, and its seven dimensions are modelled as distinct exogenous constructs associated with Overall Risk-Control Strategy Effectiveness (OSE). The paths follow the hypothesis order: financial risks (H1), environmental risks (H2), operational risks (H3), logistical risks (H4), regulatory and legal risks (H5), health and safety risks (H6), and market risks (H7). The arrows represent seven direct statistical relationships with the single dependent construct; they do not establish causal effects.

3 Methodology

3.1 Research design

The present paper reports exclusively the quantitative phase of the broader doctoral study. A cross-sectional quantitative design was adopted to examine relationships between the seven dimensions of Construction Project Risk Exposure and Overall Risk-Control Strategy Effectiveness in Omani construction projects. The questionnaire was administered once to construction professionals working in the Duqm Special Economic Zone. SPSS was used for data screening, demographic analysis, descriptive statistics, sampling-adequacy tests, and correlation analysis, while SmartPLS 4 was used to assess the measurement model and estimate the dimension-specific structural relationships with OSE. Because the design is cross-sectional and observational, all structural paths are interpreted as conditional associations rather than causal effects.

3.2 Population, sample, and data collection

The empirical setting was the Duqm Special Economic Zone. Participants were construction professionals working in engineering, project management, site supervision, planning, and related technical or supervisory roles within Galfar Engineering and Contracting and Sarooj Construction Company. Convenience sampling was adopted because a complete randomized list of eligible professionals was unavailable and project staffing was dynamic. A total of 250 questionnaires were distributed. Two hundred and twenty were returned, 10 incomplete responses were removed, and 210 usable cases were retained for analysis, producing an 84.0% usable response rate. The questionnaire introduction explained the academic purpose of the study, assured confidentiality and anonymity, and participation was voluntary.

3.3 Measurement and instrument development

The structured questionnaire contained three parts: respondent characteristics; construction-project risk exposure in Duqm; and the perceived effectiveness of strategies used to control those risks. Item development followed a structured process. An initial item pool was generated from the construction-project-risk and risk-management literature and was then reviewed by three academic experts specializing in construction risk management and project management, who evaluated each item for clarity, relevance, and correspondence to its intended construct; items were retained, reworded, or removed on the basis of this review. The revised instrument was subsequently administered to a pilot sample of 30 construction professionals, drawn separately from the main survey sample, to assess item clarity and preliminary reliability before full-scale distribution. All items were measured on a five-point Likert scale. Construction Project Risk Exposure was operationalized through seven first-order constructs: FIN, ENV, OPR, LOG, REG, HSS, and MAR. OSE was represented by FIN_S, ENV_S, OPR_S, LOG_S, REG_S, HSS_S, and MAR_S. The pilot stage (n = 30) reported Cronbach’s alpha values above 0.80 for all constructs, supporting preliminary internal consistency ahead of full-scale distribution; this pilot reliability evidence is reported alongside, and does not substitute for, the expert content review described above. To reduce common-source concerns procedurally, the questionnaire separated exposure and strategy sections, used standardized response formats, and assured respondent anonymity. OSE was retained as a reflective global-perception construct for this analysis, while alternative composite specifications are reserved for future validation.

3.4 Data analysis procedures

Preliminary screening was performed in SPSS. Missing and incomplete cases were examined, outliers were reviewed, and distributional characteristics were assessed using skewness, kurtosis, Kolmogorov-Smirnov, and Shapiro-Wilk tests. Sampling adequacy was evaluated through the Kaiser-Meyer-Olkin statistic and Bartlett’s test of sphericity. Descriptive statistics summarized the level of each risk-exposure dimension and the perceived effectiveness of the corresponding strategy domain, while Pearson correlations examined bivariate associations among the constructs.

The measurement and structural models were analysed using SmartPLS 4. PLS-SEM was selected because the study was prediction-oriented, involved multiple latent constructs, and did not require strict multivariate normality. The measurement model was evaluated through outer loadings, Cronbach’s alpha, composite reliability, average variance extracted, HTMT, and the Fornell-Larcker criterion; cross-loadings were treated as supplementary diagnostics rather than decisive evidence. The structural model was assessed using R², adjusted R², effect sizes (f²), path coefficients, t-statistics, two-tailed p-values, and model-fit indices. Statistical significance was evaluated at the 5% level using |t| > 1.96, and the p-values reported in this article were aligned with the reported t-statistics. The hypothesis order was FIN (H1), ENV (H2), OPR (H3), LOG(H4), REG (H5), HSS (H6), and MAR (H7) in relation to OSE (Hair et al., 2019; Kassem, 2022; Legate et al., 2023; Rasoolimanesh, 2022).

3.5 Quantitative interpretation of the empirical model

The quantitative findings were interpreted by integrating descriptive results, measurement-model evidence, and structural-model estimates. Construct-level means were used to describe perceived exposure and strategy effectiveness, while path coefficients and f² values were used to distinguish coefficient magnitude from unique incremental contribution. The analysis did not rank practical importance from path coefficients alone. No interview findings, thematic-analysis results, regulatory evidence, or qualitative framework-development procedures are reported in this paper.

4 Results

4.1 Response profile and data screening

The final dataset contained 210 complete responses. Men represented 64.8% of the sample and women 35.2%. The largest age group was 31–40 years (48.6%), and 75.2% of respondents were Omani. More than half held a bachelor’s degree (52.4%), while 30.0% held a master’s degree or above. Engineers formed the largest professional group (40.5%), followed by project managers (26.2%) and supervisors (21.4%). Most respondents had substantial professional exposure: 34.3% reported 6–10 years of experience and 22.9% reported 11–15 years. In addition, 90.0% had studied risk-management or project-management courses, and 91.4% assessed their risk-management knowledge as fair or advanced. Table 1 reports the complete profile used in Chapter 4. Detailed results are presented in Tables 2–6.

TABLE 1. Demographic profile of respondents (N = 210).

The screening results supported continued analysis. The KMO, statistic was 0.722 and Bartlett’s test was significant (χ² = 4580.175, df = 91, p < 0.001), indicating adequate factorability. Skewness ranged from −1.986 to 0.688 and kurtosis from −0.621 to 5.284, remaining within the criteria used in the study. Kolmogorov-Smirnov and Shapiro-Wilk tests were significant for all constructs, but PLS-SEM, was retained because it does not require strict multivariate normality and is suitable for prediction-oriented analysis.

TABLE 2. Descriptive statistics for perceived risk exposure and strategy effectiveness.

TABLE 3. Final construct reliability and convergent validity.

TABLE 4. Heterotrait-monotrait ratio of correlations (HTMT).

TABLE 5. Structural relationships and hypothesis decisions.

The global model-fit results require caution. SRMR, was 0.174 for both the saturated and estimated models, NFI, was 0.453, d_ULS, was 8.323, d_G was 10.476, and chi-square was 5628.888. These values indicate weak overall fit. The structural coefficients should therefore be interpreted together with measurement quality, R², f², the substantial correlations among exogenous constructs, and the context-specific nature of the sample. Because inner VIF, statistics and the complete bootstrapping configuration were not preserved in the available quantitative output, the paper does not claim relative dominance among the risk dimensions or universal model adequacy.

TABLE 6. Explanatory power and model-fit indices.

4.2 Descriptive statistics and correlations

Within the Construction Project Risk Exposure domain, regulatory and legal risk recorded the highest construct-level mean (M = 3.471, SD = 0.717), followed by environmental risk (M = 3.356, SD = 0.735) and health and safety risk (M = 3.119, SD = 0.651). Financial risk recorded the lowest mean (M = 2.614, SD = 0.615). For the seven strategy-effectiveness indicators of OSE, environmental strategies received the highest rating (M = 3.757, SD = 0.693), followed by logistical strategies (M = 3.519, SD = 0.753), while financial strategies received the lowest rating (M = 3.312, SD = 0.449). These rankings are based on construct means and are not directly comparable with item-level Relative Importance Index rankings reported elsewhere in the broader study.

The correlation matrix indicated that the seven exposure dimensions and the seven strategy-effectiveness indicators were interconnected but not interchangeable. The largest reported associations among risk dimensions were market with financial risk (r = 0.683, p < 0.001), logistical with health and safety risk (r = 0.659, p < 0.001), and regulatory with financial risk (r = 0.659, p < 0.001). Environmental risk was not significantly associated with operational strategies (r = 0.006, p = 0.927) or environmental strategies (r = −0.011, p = 0.871). The correlation pattern supports multidimensional analysis but also indicates substantial shared variance that requires cautious interpretation of structural coefficients.

4.3 Measurement model

The initial measurement model met construct-level reliability and convergent-validity criteria, but FIN3 loaded at 0.429 and was removed. In the refined model, Cronbach’s alpha ranged from 0.759 to 0.941, composite reliability from 0.832 to 0.953, and AVE from 0.721 to 0.894. Most retained indicators exceeded the preferred loading threshold. REG2 (0.666), ENV3 (0.711), and FIN_S (0.613) were comparatively lower, but their associated constructs retained acceptable internal consistency and AVE. OSE was represented by the seven domain-specific strategy indicators as a global perceived-effectiveness construct.

Discriminant validity was assessed primarily through HTMT and the Fornell-Larcker criterion. Every HTMT value was below 0.85; the largest was 0.820 between operational and health and safety risks, followed by 0.777 between regulatory and environmental risks and 0.776 between regulatory and health and safety risks. Cross-loadings were inspected as supplementary diagnostics and were not treated as conclusive evidence because a small number of indicators displayed substantial secondary loadings. The constructs were retained for the present exploratory analysis, but the affected indicators and the reflective OSE specification require independent replication.

Figure 2 shows that most retained indicators load strongly on their intended constructs and that construct-level reliability is acceptable. ENV3, REG2, and FIN_S are comparatively weaker and therefore warrant cautious interpretation. The model supports a multidimensional representation of risk exposure and a shared overall strategy-effectiveness construct, but it should not be read as proving that all strategy domains are interchangeable or that the measurement specification is definitive.

4.4 Structural model and hypothesis testing

The seven dimensions of Construction Project Risk Exposure jointly explained 43.0% of the variance in OSE (adjusted R² = 0.410), indicating moderate explanatory power. Financial risk was positively associated with OSE (H1: β = 0.285, p = 0.017), environmental risk was negatively associated with OSE (H2: β = −0.445, p = 0.030), operational risk was positively associated with OSE (H3: β = 0.502, p = 0.008), logistical risk was positively associated with OSE (H4: β = 0.590, p = 0.009), and regulatory and legal risk was positively associated with OSE (H5: β = 0.427, p = 0.016). Health and safety risk was not statistically significant (H6: β = 0.319, p = 0.408), while market risk was positively associated with OSE (H7: β = 0.778, p = 0.044). Environmental risk had the largest f² value (0.193), followed by regulatory risk (0.077) and logistical risk (0.021). Market and operational risks had large coefficients but negligible f² values (0.008 and 0.002), showing that coefficient magnitude should not be equated with unique practical contribution.

5 Discussion

5.1 Theoretical implications

The findings support a two-level specification of the study variables. Construction Project Risk Exposure is meaningful as an overarching conceptual domain because the seven dimensions describe a connected project-risk environment. At the same time, the dimension-specific paths show that the domain is not statistically homogeneous. Regulatory and legal risk recorded the highest construct-level mean, whereas market risk produced the largest positive coefficient. These quantities answer different questions: means summarize perceived exposure, coefficients describe conditional associations, and f² values estimate unique incremental contribution. They should not be used interchangeably.

Financial, operational, logistical, regulatory and legal, and market risks showed significant positive associations with OSE. However, the very small f² values for financial, operational, and market risks indicate that their unique incremental contributions were limited after the other dimensions were included. Environmental risk showed the largest f² value, while regulatory and logistical risks contributed smaller increments. This pattern is consistent with substantial shared variance across the construction-risk environment and supports an integrated but cautious interpretation rather than a ranking based solely on standardized coefficients (Kassem and Ali, 2022; Malek and Bhatt, 2024).

Environmental risk supported H2 but produced a negative coefficient, even though environmental strategies received the highest average effectiveness rating. The result means that, conditional on the other dimensions in this cross-sectional model, respondents reporting higher environmental exposure tended to report lower OSE. It does not establish that environmental exposure causes strategy effectiveness to decline. The comparatively larger f² value nevertheless identifies environmental exposure as an important area for further measurement refinement, stress testing, and longitudinal investigation.

Health and safety risk was the only dimension without a statistically significant direct association with OSE, and H6 was not supported. This result does not imply that health and safety is unimportant or that safety controls should be reduced. The nonsignificant path may reflect limited independent variance after correlated risk dimensions are included, contextual institutionalization of safety procedures, or mechanisms not represented in the present model, such as leadership, safety climate, reporting, and resilience (Zhang and Zhang, 2023).

The measurement results support construct-level reliability and convergent validity, while HTMT values provide the principal evidence of discriminant validity. The secondary cross-loading patterns and the reflective specification of OSE prevent stronger claims that the measurement model is definitive. The moderate R² demonstrates useful explanatory capacity, but the weak SRMR and NFI, unavailable inner VIF output, and context-specific sample require cautious interpretation. The contribution is therefore an exploratory, association-based account that requires independent replication and alternative measurement testing.

5.2 Managerial implications

The quantitative findings provide an empirical basis for prioritizing risk-control review in Omani construction projects, but management attention should not be determined from path coefficients alone. Construct means, f² values, operational consequences, and current strategy ratings should be considered together. Environmental exposure deserves particular review because it combined a significant negative association with the largest f² value. Regulatory and logistical risks also showed non-negligible incremental contributions. Market and operational risks remain managerially relevant, but their large coefficients were accompanied by negligible f² values and should not be described as the dominant unique drivers of OSE.

Financial strategies received the lowest effectiveness rating and should be strengthened through cash-flow forecasting, payment-risk monitoring, contingency allocation, cost-escalation provisions, and contractor-client coordination. Environmental strategies received the highest rating, yet the negative conditional association between environmental exposure and OSE supports stress testing under severe or changing conditions. Health and safety controls should remain fully resourced despite the nonsignificant path because their ethical and operational importance is not determined by one structural coefficient. Overall, resource allocation should combine statistical evidence with project-specific consequences and professional judgement. The quantitative evidence can inform future refinement of context-sensitive risk-management practices. The seven-dimensional classification provides a structured basis for organizing risks, while construct means, path coefficients, and f² values offer different forms of decision evidence. However, the design, sequencing, and validation of any formal framework extend beyond this cross-sectional quantitative analysis and require independent implementation-based evaluation.

5.3 Limitations and future research directions

The findings should be interpreted in light of several limitations. Convenience sampling from two companies (Galfar Engineering and Contracting and Sarooj Construction Company) restricts generalizability to the broader Omani construction sector, and the cross-sectional design does not establish temporal order or causality. No a priori statistical-power analysis was conducted; consequently, the adequacy of the achieved sample of 210 for detecting the smaller incremental effects reported here cannot be independently verified and should be borne in mind when interpreting the structural results. Because all 210 respondents were drawn from only two firms, observations may not be fully independent: firm-specific risk procedures, organizational climate, or supervisory practices could induce within-firm dependence that is not represented in the reported standard errors, constituting a further threat to inference beyond sampling bias alone. Predictor and outcome measures were self-reported by the same respondents; procedural separation and anonymity were used, but no ex post common-method test was available at the time of submission. Five of the seven exposure constructs are measured with only two retained indicators each, which limits empirical coverage of these broad risk domains and increases sensitivity to individual item wording. The removal of FIN3 (loading = 0.429) was a post hoc, data-driven decision rather than one anticipated in the original measurement design; while consistent with standard PLS-SEM refinement practice, it was not accompanied by independent content-validity evidence for the revised two-item FIN construct. The global-fit indices indicate more than a minor qualification: SRMR = 0.174 and NFI = 0.453 reflect substantial misfit between the specified model and the covariance structure of the data, and the structural results are best read as associations generated by an imperfectly fitting model rather than as evidence of a well-specified structural system. Several indicators displayed substantial secondary cross-loadings, and the reflective OSE specification - notwithstanding the conceptual position that the seven strategy domains are not operationally interchangeable - remains open to alternative composite or formative modelling in future work. Inner VIF statistics and the complete bootstrapping configuration, including confidence intervals for the structural paths, were not preserved in the available quantitative output; their absence means that path significance and the stability of coefficients for correlated predictors (e.g., market and operational risk) cannot be independently verified here. PLS-SEM was justified partly on prediction-oriented grounds, but no out-of-sample predictive assessment was conducted; R² = 0.430 is an in-sample explanatory statistic and does not by itself validate this predictive rationale. Finally, environmental risk exposure was essentially uncorrelated with its corresponding environmental-strategy rating at the bivariate level (r = −0.011, p = 0.871) yet emerged as a significant negative predictor in the multivariate structural model; given the correlated exogenous constructs and the unavailable collinearity diagnostics, this pattern is consistent with either a suppression effect or coefficient instability, and the negative relationship implied by the path coefficient should not be interpreted at face value without further diagnostic work. The conclusions must therefore remain specific to the seven dimension-level associations reported here, and several of the constraints above are addressed as priorities for future research in Section 5.4.

Future research should use longitudinal and multi-source designs, combine perceptual measures with verified project indicators, and replicate the model across construction sectors and regions. Researchers should compare the present seven-predictor specification with hierarchical-component models for Construction Project Risk Exposure and alternative reflective, composite, or formative specifications for OSE. Full collinearity diagnostics, complete bootstrapping settings and confidence intervals, common-method assessments, and out-of-sample predictive tests should be reported. Potential mediators and moderators, including risk-management maturity, organizational learning, digital readiness, collaborative capacity, leadership, and safety climate, should also be investigated.

6 Conclusion

This paper examined the relationships between seven dimensions of Construction Project Risk Exposure and Overall Risk-Control Strategy Effectiveness using quantitative data from 210 construction professionals in Duqm. Regulatory and legal risk recorded the highest perceived exposure, environmental strategies received the highest effectiveness rating, and financial strategies received the lowest. The refined measurement model demonstrated acceptable internal consistency and convergent validity, while HTMT values supported construct distinctiveness. The seven exposure dimensions explained 43.0% of the variance in OSE. Financial, operational, logistical, regulatory and legal, and market risks showed significant positive associations; environmental risk showed a significant negative association; and health and safety risk was nonsignificant.

The results support an integrated yet differentiated view of construction-project risk exposure, but they do not establish causal effects or a universal model. Means, path coefficients, and f² values provide different information and should be interpreted together. In particular, large coefficients for market and operational risk were accompanied by negligible unique effect sizes, while environmental risk produced the largest f² value. Weak global fit, secondary cross-loading patterns, the cross-sectional design, and the context-specific sample require caution. The model should therefore be treated as an exploratory quantitative contribution that informs future risk-control research and validation rather than as a definitive causal framework.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Heng Li, Hong Kong Polytechnic University, Hong Kong, SAR China

Reviewed by: Milad Baghalzadeh Shishehgarkhaneh, Monash University, Australia

Irfan Prasetia, Universitas Lambung Mangkurat, Indonesia