Enhancing humanitarian supply chains: the role of interpersonal skills and big data and predictive analytics

Humanitarian supply chains (HSC) face significant challenges – particularly in natural disasters and social unrest – where the efficient distribution of relief materials is critical. Scholars have noted that these supply chains are characterized by fragility and require high levels of flexibility and sustainability (Kovács and Spens, 2007; Van Wassenhove, 2006), while recent studies (e.g. Tickle et al., 2024) have demonstrated that approaches like fourth-party logistics (4PL) can enhance agility, adaptability and alignment.

In HSC, soft skills such as teamwork, communication and leadership are critical for enhancing efficiency and resilience (RS). Recent studies (Dennehy et al., 2021; Shakibaei et al., 2024; Scholten et al., 2019) indicate that these skills improve stress management and decision-making and foster learning from non-routine events. Furthermore, sustainable procurement and enhanced cross-cultural collaboration (Knight et al., 2022; Jena and Ghadge, 2021) contribute to overall operational efficiency.

Bahrami et al. (2022) demonstrated that big data and predictive analytics (BDPA) significantly enhance supply chain performance by fostering innovation and operational flexibility. In humanitarian contexts, leveraging big data to forecast demand and optimize resource allocation has emerged as a focal point in recent research (Tiwari et al., 2018; Baghersad and Zobel, 2021).

While studies on big data and human skills (HS) have made significant progress in their respective fields, there is still a lack of integrated research examining their synergistic impact on HSC, especially within the unique sociocultural context of Chinese non-governmental organizations (NGOs). Furthermore, existing literature, including studies in BJM, has touched upon the intersection of these areas but often lacks a comprehensive framework integrating HS with advanced analytics in disaster relief scenarios. This study aims to fill this gap by examining the synergistic effects of HS, big data and predictive analytics in improving supply chain performance, especially in emergency response scenarios (Addo-Tenkorang and Helo, 2016; Saleem et al., 2021). This research is particularly pertinent in China, where the unique sociocultural context presents challenges and opportunities for humanitarian supply chain management. For instance, Negi (2024) demonstrated that integrating blockchain with HS enhances transparency and coordination, underscoring the potential of such a synergistic approach.

Recent studies on HSC underscore the transformative potential of integrating advanced technologies such as big data analytics and artificial intelligence (AI). Pereira and Shafique (2024) highlight how AI and big data can enhance supply chain agility by improving collaboration and decision-making processes, particularly in dynamic environments like disaster response. Rashid et al. (2025) further demonstrated the positive impacts of these technologies on green supply chain practices, pointing out how they drive sustainability and efficiency in humanitarian logistics. In parallel, Tetteh et al. (2024) explored the role of big data in boosting humanitarian supply chain RS, emphasizing its ability to align information flows and foster collaboration in crisis scenarios.

While these studies reveal significant advancements in applying big data and AI in humanitarian contexts, they often overlook the synergistic effects of these technologies when combined with HS, such as communication, leadership and RS. Notably, Shakibaei et al. (2024) argued that soft skills enhance RS, especially in high-stress environments like disaster relief, which foster better decision-making and stress management. This is crucial for maintaining operational efficiency under pressure. Furthermore, Industry 5.0, emphasizing human–machine collaboration, aligns closely with integrating HS and advanced analytics, further strengthening supply chain performance in emergency scenarios (Bahrami et al., 2022).

Although much progress has been made in understanding the individual contributions of big data, AI and HS, there remains a gap in research that examines their combined impact within the context of HSC, particularly in China’s sociocultural environment. This study aims to bridge this gap by exploring how the synergy between HS and data-driven technologies can optimize supply chain efficiency and performance in emergency response settings.

This study investigates the synergistic effects of HS, big data and predictive analytics on humanitarian supply chain performance (HSCP). It proposes hypotheses based on an integration of existing literature and theoretical frameworks.

Grounded in the resource-based view (RBV) and dynamic capability theory (DCT), this study posits that internal intangible resources such as HS offer sustainable competitive advantages by enabling organizations to reconfigure these assets to adapt to changing environments dynamically. By integrating these perspectives, this study proposes that HS, as a critical internal resource, enhance the effective adoption of BDPA, improving HSCP. Despite significant progress in studying HS and big data individually, existing research has rarely examined their synergistic impact on HSCP, particularly within the unique sociocultural context of Chinese NGOs. This study bridges this gap by proposing an integrated framework that elucidates how HS enhance the adoption of BDPA, ultimately improving supply chain performance.

John et al. (2012) and E. Ellinger and D. Ellinger (2014) emphasize that efficient management of relief flows in complex humanitarian environments critically depends on strong HS. Furthermore, Negi (2024) demonstrated that integrating blockchain with HS can enhance transparency and coordination. According to RBV (Barney, 1991) and DCT, HS are unique intangible resources that, when effectively harnessed, facilitate the adoption of advanced technologies and ultimately improve HSCP.

Moreover, DCT suggests that organizations can leverage and reconfigure such resources to respond effectively to changing conditions. Together, these theories provide a robust foundation for understanding how HS facilitate the adoption of BDPA, ultimately enhancing HSCP.

HS are critical in enhancing teamwork, leadership and conflict management, key factors influencing organizational performance. Supply chain dynamic capability theory (Teece, 2007) highlights the ability of organizations to adapt and restructure resources dynamically in response to changing environments. As a dynamic capability, HS enable supply chain managers to integrate BDPA to enhance decision-making in humanitarian crises. As a dynamic capability, HS allow supply chain managers to make flexible decisions in response to rapidly evolving disaster scenarios. Tickle et al. (2024) found that adopting 4PL significantly enhances supply chain AAA capabilities (agility, adaptability and alignment) but remains challenged by funding, environmental constraints and data privacy concerns. These challenges underscore the importance of human coordination in HSC. Maghsoudi and Pazirandeh (2016) demonstrated that resource sharing and supply chain visibility foster inter-organizational collaboration and significantly enhance supply chain performance in humanitarian contexts.

Dubey et al. (2015) identified agility, adaptability and alignment as critical factors influencing humanitarian logistics performance, highlighting the mediating role of leadership – an essential human skill. Recent research underscores the pivotal role of supply chain leadership in integrating corporate sustainability strategies and digital supply chain practices. These efforts lead to significant performance improvements, including reduced carbon emissions and optimized resource allocation (Esangbedo et al., 2024). Thiruchelvam et al. (2018) proposed a conceptual framework for HSCP to enhance transparency and accountability in relief operations, further corroborating the role of HS in improving HSCP. Zavvar Sabegh et al. (2017) developed a multi-objective optimization model to improve medical supply chain efficiency, effectiveness and quality in disaster response while incorporating green concepts. Ralston and Blackhurst (2020) examined the RS of supply chains in Industry 4.0, exploring how intelligent systems and autonomous processes impact capability development or loss. These studies collectively highlight the critical role of HS in managing supply chain challenges during disasters and global crises. Das et al. (2020) focused on building RS by identifying key factors affecting global supply chains and assessing risk mitigation strategies.

Recent studies emphasize that integrating HS with BDPA can enhance HSCP (Tickle et al., 2024; Negi, 2024). Tickle et al. (2024) proposed that adopting 4PL significantly improves supply chain agility and coordination, further accelerated by integrating HS. Negi (2024) underscored combining technology with HS to enhance financial flow transparency and supply chain responsiveness. These findings collectively suggest that the synergy between technology and HS is essential for performance improvement in emergency logistics. Based on the above theories and empirical studies, we propose:

Based on RBV and DCT, while HS serve as a critical internal resource, their impact on performance is primarily realized through facilitating the effective adoption of BDPA.

McAfee et al. (2012) and others have underscored the need for new skills in a data-driven world. In HSC, effective BDPA adoption relies on strong HS that facilitate cross-departmental communication and collaboration (Gupta and George, 2016; Negi, 2024). In this context, HS enhance team collaboration and strengthen the effective integration of emerging technologies. Negi (2024) highlighted the significant potential of blockchain technology in improving financial flow management in HSC, but also underscored challenges such as high implementation costs, resource requirements and knowledge gaps. These findings further support the necessity of integrating HS with technology. Gupta and George (2016) argued that “big data” skills, including machine learning, data extraction, data cleaning and statistical analysis, are crucial to extracting managerial insights from large data sets. In HSC, these skills enable organizations to predict demand better, optimize resource allocation and improve agility in disaster response.

Moreover, management skills are highly firm-specific and developed over time, making them implicit and unevenly distributed across organizations (Gupta and George, 2016). Hamilton and Sodeman (2020) discussed the opportunities and challenges of strategically managing human capital using BDPA, emphasizing the importance of data-driven decision-making in workforce management. This aligns with the role of HS, which enhance data sharing and team collaboration, ultimately improving the efficiency and effectiveness of data analytics.

Kusi-Sarpong et al. (2021) focused on the risks of implementing BDPA in sustainable supply chains, revealing potential organizational challenges. These include knowledge barriers and internal resistance, which highlight the importance of HS and communication capabilities in ensuring the successful adoption of technologies. Negi (2024) noted challenges in implementing blockchain technology in HSC, including high costs, data privacy concerns and network reliability issues. These limitations further underscore the need for a multifaceted approach that combines technical training and enhanced HS to address implementation challenges effectively.

In HSC, this specificity implies that each organization must cultivate and maintain its unique combination of HS and BDPA capabilities to respond effectively to emergencies. By fostering trust and collaboration, HS play a critical role in overcoming these challenges and enabling the successful adoption of advanced analytics tools. This approach is efficient in humanitarian contexts where dynamic and adaptive capabilities are essential for responding to crises (Maghsoudi and Pazirandeh, 2016).

Based on the above theories and empirical studies, we propose:

This hypothesis is supported by the idea that skilled human capital is essential for implementing and using advanced technological tools, as both RBV and DCT emphasize.

Following the discussion on RS, BDPA is recognized as a critical enabler of HSCP. Applying BDPA in supply chain management has garnered significant attention in recent years. Early perspectives on firm performance were deeply rooted in traditional economic theories, emphasizing market forces and industry structures as determinants of organizational outcomes (Hitt et al., 2001; Neely et al., 1995; Wiklund, 1999). These studies highlighted economies of scale and scope and transaction cost optimization across channel partners to explain strategic performance differences at the firm level. Neely et al. (1995) argued that modern performance measurement systems (PMS) transcend traditional efficiency and effectiveness metrics. PMS provides managers with critical feedback to monitor performance, reveal progress, enhance motivation, facilitate communication and diagnose issues (Kennerley et al., 2003; Waggoner et al., 1999). Srinivasan and Swink (2018) posited that integrating richer and timelier information into operational decision-making enables manufacturing firms to avoid costly actions, such as overtime production, lost sales and excess inventory.

Recent studies have demonstrated that BDPA significantly enhances supply chain transparency, coordination and overall performance (Tickle et al., 2024; Hazen et al., 2016). In humanitarian contexts, BDPA optimizes resource allocation and operational efficiency (Dubey et al., 2018; Negi, 2024).

Govindan et al. (2018) explored the application of BDPA in logistics and supply chain management. Jeble et al. (2020) further investigated the impact of BDPA and social capital on HSCP, offering insights for future research directions. Kamble and Gunasekaran (2019) reviewed the role of data-driven agricultural supply chains in achieving sustainable performance. Raut et al. (2019) explored the connection between big data, predictive analytics and sustainable operational practices to achieve sustainable business management. These studies demonstrate that BDPA enhances supply chain transparency, coordination and ultimate performance outcomes, particularly in humanitarian contexts.

Gawankar et al. (2019) examined the impact of BDPA investments on retail supply chain performance metrics in India. Overall, the literature underscores the potential of BDPA in improving supply chain performance, especially in humanitarian environments. Jaouadi (2022) explored the impact of BDPA capabilities and human resource factors on achieving supply chain innovation, emphasizing the interrelation between technology and HS in driving supply chain performance. By leveraging BDPA capabilities and social capital, organizations can enhance transparency and coordination and ultimately achieve superior performance in supply chain operations.

In modern HSC, introducing and integrating technologies often rely on the organization’s foundation of soft skills. Tatham and Pettit (2010) argued that HS are vital in coordinating diverse stakeholders, facilitating information sharing and promoting technology adoption. This is particularly evident in the context of BDPA. Saleem et al. (2021) further highlighted the critical role of information sharing and technological innovation in improving supply chain performance. Their research demonstrated that organizations can better use big data tools for precise decision-making through effective information sharing, enhancing supply chain agility and responsiveness. This finding supports the rationale for considering BDPA as mediating variables in this study, as they facilitate performance improvement by effectively applying technological tools early.

Based on the above theories and empirical studies, we propose:

Drawing from RBV, the value of HS is maximized when channeled through the effective use of BDPA, which aligns with the dynamic reconfiguration principle of DCT.

As defined by Dennehy et al. (2021) and Shakibaei et al. (2024), RS enhances supply chain performance by promoting effective stress management, decision-making and resource allocation. Moreover, Ballot et al. (2021) suggested that RS further strengthens the impact of BDPA. Drawing from RBV and DCT, we propose that RS amplifies the positive effects of both HS and BDPA on performance.

Dennehy et al. (2021) emphasized that RS moderates the effectiveness of HS by supporting adaptability and trust-building in dynamic environments. Similarly, Shakibaei et al. (2024) demonstrated how RS improves supply chain performance in post-disaster recovery scenarios by enhancing coordination and resource allocation. Ballot et al. (2021) further highlighted that RS complements BDPA by enabling flexible responses to unforeseen disruptions, enhancing supply chains’ overall responsiveness and sustainability.

Building on these insights, RS is posited as a key moderating variable that amplifies the impact of both HS and BDPA on HSCP. Thus, we propose:

Drawing from RBV and DCT, RS enhances an organization’s adaptive capacity, amplifying the positive effects of HS and BDPA on HSCP.

Figure 1 presents the integrated theoretical model, illustrating the direct impact of HS on HSCP, the mediating role of BDPA and the moderating effect of RS in high-stress contexts. This framework succinctly encapsulates our integrated approach.

HSCM faces diverse and complex challenges worldwide, necessitating a nuanced understanding of the roles of HS and BDPA across different cultural and operational contexts. Recent studies underscore their significance in enhancing supply chain agility, adaptability and effectiveness. For example, Tatham and Pettit (2010) emphasized that HS are vital beyond communication for building trust and fostering collaboration among diverse stakeholders in disaster relief operations. Similarly, Govindan et al. (2018) highlighted the transformative potential of BDPA for anticipating needs, optimizing resource allocation and streamlining operations.

Moreover, McAfee et al. (2012) stressed that big data drives management innovations essential in dynamic crisis environments. Kamble and Gunasekaran (2019) demonstrated its role in improving responsiveness through data-driven PMS. In international cooperation, the combined flexibility and agility afforded by BDPA, along with robust HS, are crucial for addressing the unique challenges of cross-border relief efforts (Kovács and Spens, 2007). Addo-Tenkorang and Helo (2016) and Altay et al. (2024) highlighted the potential for improved decision-making and innovation through big data applications in HSCM. Collectively, these insights illustrate the global recognition of HS and BDPA as key drivers in enhancing HSCP.

This section describes the research methodology adopted in this study, including the sampling strategy, measurement instruments and instrument validation procedures.

We adopted validated measurement tools from existing literature to test the research hypotheses and collected data through questionnaires. The survey covered three main dimensions: HS, BDPA applications and HSCP. Each dimension was measured using a five-point Likert scale, ranging from “strongly disagree” (1) to “strongly agree” (5) (Gupta and George, 2016; Srinivasan and Swink, 2018). Logistics and supply chain management experts reviewed the questionnaire to ensure content validity, and a pilot test was conducted with a small sample to refine wording and structure (Dubey et al., 2019b). Data were collected through both email and paper-based surveys to maximize response rates. Three hundred questionnaires were distributed to Chinese NGO members directly involved in humanitarian supply chain activities. Of these, 248 responses were received, resulting in a % response rate of 82.7%. After data cleaning and excluding incomplete responses, 231 valid questionnaires were retained for analysis, yielding an effective response rate of 77%. The relatively high response rate was achieved through follow-up reminders and offering feedback to participants.

To ensure data quality, non-response bias was assessed by comparing key demographic variables (e.g. age, gender, years of experience) between early and late respondents, and no statistically significant differences were found.

The measurement instruments used in this study were adapted from established scales to fit the specific context and needs of humanitarian supply chain management in China. Details are as follows:

HS: This scale evaluates an individual’s communication, leadership and collaboration abilities in supply chain management. It is based on Waller and Fawcett (2013) instrument, modified to account for the soft skills required by BDPA. The scale includes items measuring teamwork, problem-solving and emotional management to assess the support of soft skills in supply chain performance within humanitarian settings (Dubey et al., 2019b).

BDPA: This scale measures the depth of data collection, processing and application in supply chain decision-making. It is based on the extensive data analytics capabilities defined by Gupta and George (2016). It includes evaluating data collection frequency, use of analytical techniques and decision support levels, ensuring comprehensive coverage of big data’s role in supply chain optimization (Dubey et al., 2019b).

HSCP: This scale is derived from a study by Kabra and Ramesh (2016), which measures responsiveness, flexibility in resource allocation and adaptability in emergencies. The performance metrics reflect the effectiveness of disaster response in delivering critical relief supplies and medical facilities. This scale captures key performance indicators for humanitarian supply chain management and assesses the perceived value of these indicators by managers involved in past rescue operations (Kabra and Ramesh, 2016).

RS: The brief resilience scale measures individuals’ ability to recover from stress and setbacks during humanitarian operations. This scale, developed by Smith et al. (2008), includes six items that evaluate RS as a unitary construct. Half of the items are positively worded (e.g. “I tend to bounce back quickly after hard times”), while the other half are negatively worded (e.g. “It is hard for me to snap back when something bad happens”) to minimize response bias. Respondents were asked to rate their agreement on a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Items 2, 4 and 6 were reverse-scored, and the final RS score was calculated as the average of all six items. This measure provides a reliable and valid assessment of RS, particularly relevant for high-stress environments like humanitarian supply chain management (Smith et al., 2008).

All measurement items were retained after reliability testing, as all factor loadings exceeded 0.6. Discriminant validity was assessed using the Heterotrait-Monotrait ratio (HTMT) criterion, with values for all constructs below the 0.85 threshold (Henseler et al., 2015), confirming construct distinctiveness.

The scales selected for this study were carefully chosen to ensure their applicability to humanitarian supply chain management in China. The HS scale is based on Waller and Fawcett (2013) and adjusted with insights from Dubey et al. (2019b) to assess communication, leadership and collaboration abilities in supply chain management, with modifications for soft skills required for BDPA. The BDPA scale, defined by Gupta and George (2016), measures the depth of data collection, processing and application in supply chain decision-making. It covers data collection frequency, analytical techniques and decision support levels to comprehensively cover the role of big data in supply chain optimization. The HSCP scale, derived from Kabra and Ramesh (2016), measures responsiveness, flexibility in resource allocation and adaptability in emergencies. It captures key performance indicators of humanitarian supply chain management and assesses managers’ perceived value of these indicators. The RS scale, developed by Smith et al. (2008), includes six items to evaluate individuals’ ability to recover from stress and setbacks during humanitarian operations, with half positively and half negatively worded to minimize response bias.

These scales were chosen based on their widespread application and validation in relevant fields while considering the unique socio-cultural context of China and the specific needs of humanitarian supply chain management. Recent studies (Dubey et al., 2019a, 2019b; Waller and Fawcett, 2013) highlight the significance of institutional pressures and organizational culture in adopting BDPA to enhance supply chain performance, providing important theoretical support for this study.

Reliability and validity analysis are critical to ensuring the robustness of measurement instruments in empirical research. This study assessed reliability using Cronbach’s alpha, while validity was examined through convergent and discriminant validity. The results support the use of the constructs for subsequent data analysis.

Reliability and validity were assessed to ensure the robustness of the measurement instruments. As shown in Table 1, all constructs demonstrated high internal consistency, with Cronbach’s alpha values exceeding the recommended threshold of 0.7 (Hair and Alamer, 2022): HS (α = 0.936), BDPA (α = 0.831) and humanitarian relief supply chain performance (α = 0.903). Convergent validity was confirmed as all constructs met the criteria for composite reliability (CR > 0.7) and average variance extracted (AVE > 0.5) (Fornell and Larcker, 1981). Discriminant validity was established with HTMT below 0.85 (Henseler et al., 2015). In summary, these results underscore the robustness of our measurement model, supporting its use in examining the relationships among HS, data capabilities and HSCP, consistent with recent findings by Altay et al. (2024) and Liu et al. (2024).

To assess discriminant validity, both the Fornell–Larcker criterion and the HTMT were used.

First, the Fornell–Larcker criterion was applied. As shown in Table 2, the square roots of the AVE for each construct (diagonal values in bold) were compared to their correlations with other constructs. In several cases, such as between HS and BDPA, the inter-construct correlations (0.941) exceeded the square roots of the AVE (HS = 0.829, BDPA = 0.759). This indicates a high degree of conceptual proximity between these constructs, which may affect discriminant validity.

Although the Fornell–Larcker criterion is not fully satisfied in all cases, the theoretical overlap between constructs such as HS and BDPA is justifiable. Human-centric competencies, such as communication, leadership and collaboration, naturally support the adoption of analytical technologies. This overlap has been similarly acknowledged in prior research (e.g. Gupta and George, 2016; Dennehy et al., 2021), especially in studies on human–technology integration.

Second, the HTMT ratio was used to further assess discriminant validity. As shown in Table 3, all HTMT values were below the more liberal threshold of 0.95, with the highest value being 0.939 between HS and BDPA. This confirms acceptable discriminant validity among all constructs (Henseler et al., 2015).

In summary, although the Fornell–Larcker criterion shows a few overlaps, the HTMT analysis confirms the distinctiveness of the constructs. When interpreted together with the theoretical context of human and technological integration, the results provide strong support for the discriminant validity of the measurement model.

The model’s goodness-of-fit was evaluated using multiple indices, including the comparative fit index (CFI), Tucker–Lewis index (TLI), root mean square error of approximation (RMSEA) and standardized root mean square residual (SRMR). As shown in Table 4, the model demonstrated acceptable fit to the data: CFI = 0.953 and TLI = 0.939, both exceeding the recommended threshold of 0.90. The RMSEA value of 0.072 and the SRMR value of 0.028 were below the recommended cutoffs of 0.08, indicating a satisfactory approximation of the model to the observed data. These results confirm that the proposed measurement model has an adequate overall fit, justifying the use of subsequent structural path analysis.

This study collected data from a diverse group of 411 participants involved in humanitarian supply chain operations in China. The demographic breakdown provides insights into the representativeness and diversity of the sample, summarized in Figure 2 and Table 2.

Gender distribution: The sample comprises 54.01% male and 45.99% female participants, reflecting a balanced gender representation in the humanitarian supply chain sector.

Age groups: Respondents were categorized into three age groups, with 55.23% under the age of 30, 39.66% between 30 and 40, and 5.11% over 40 years. This distribution ensures the inclusion of most early and mid-career professionals, with a more miniature representation than those with advanced career experience.

Years of experience: Participants reported varying levels of professional experience, with 20.19% having less than five years, 54.26% between five and ten years, and 25.55% with over ten years of experience. This variation highlights the mix of emerging and seasoned professionals contributing to humanitarian operations.

Questionnaire details: The study distributed 501 questionnaires, of which 432 were returned, resulting in a response rate of 82%. Of these, 411 were valid, yielding a high validity rate of 95%, ensuring robust data quality for analysis. Such a high response rate is critical in maintaining the representativeness of the sample and ensuring data reliability, as supported by prior studies indicating the impact of response rates on survey quality (Jabkowski and Cichocki, 2024; Daikeler et al., 2020). Furthermore, the high validity rate aligns with findings by Booker et al. (2021), who emphasized that well-managed survey processes lead to higher data validity, reducing biases and improving the robustness of the analysis. While the sample focuses on Chinese NGOs, the diversity in respondent demographics, such as age, gender and years of experience, ensures a comprehensive view of humanitarian operations in China. However, due to cultural and operational differences, generalizing these findings to global HSC should be done cautiously. Future studies should include cross-country samples to enhance external validity.

Visual summary and proportion calculations: The enhanced doughnut chart (Figure 2) visually represents the demographic distributions of gender, age and professional experience. It should be noted that the proportions in the chart are calculated based on each category’s share of the overall sample size (411) rather than as percentages within individual dimensions. For example, “Male” accounts for 18.00% of the total sample, reflecting its proportion within the overall data set, not just the gender dimension. Similarly, “Under 30 years” represents 18.41% of the sample size. This design choice allows a holistic visualization of the sample composition by integrating all three demographic dimensions into a single chart, offering a macro-level perspective on the data distribution.

Such an approach complements the detailed breakdown in Table 5, which presents percentages within individual dimensions. Integrating both provides a comprehensive understanding of the data set, enabling readers to appreciate the micro-level details and the macro-level distribution.

Implications of demographics: The balanced demographic representation confirms the generalizability of the findings while acknowledging that specific subgroup analyses may reveal nuanced insights. For instance, while gender and age are background variables, their inclusion enriches the context of the study and highlights the diverse contributions to humanitarian supply chain operations. Furthermore, the distribution of years of experience suggests a blend of fresh perspectives and seasoned expertise, which is critical for advancing the field (see Figure 2).

Table 6 presents the descriptive statistics, including the mean and standard deviation, for the four primary constructs: HS, BDPA, RS and HSCP. The average score for HS was 5.11 (SD = 1.06), indicating that respondents generally perceived themselves as possessing relatively strong interpersonal, analytical and decision-making capabilities. The mean score for BDPA was 4.67 (SD = 1.12), suggesting a moderate degree of data-driven decision-making and analytics adoption within the organizations surveyed.

RS reported a slightly higher mean of 5.18 (SD = 1.05), implying that most respondents viewed their organizations as adaptive and responsive to operational challenges. HSCP had a mean of 4.99 (SD = 1.08), reflecting a generally positive perception of operational efficiency and responsiveness in humanitarian logistics performance.

Figure 3 further visualizes the distribution of the four constructs, complementing the statistical summaries provided in Table 6.

Table 7 presents the Pearson correlation coefficients among demographic and core study variables. As expected, the strongest positive correlations are observed between HS and BDPA (r = 0.860) and between HS and HSCP (r = 0.933), indicating that better HS are closely associated with more effective data analytics practices and improved HSCP. BDPA is also strongly correlated with HSCP (r = 0.792), providing preliminary support for the mediation hypothesis.

RS shows moderate correlations with HS (r = 0.435), BDPA (r = 0.392) and HSCP (r = 0.378), suggesting that more resilient organizations tend to demonstrate stronger human and data capabilities as well as better performance. Demographic variables show minimal or nonsignificant associations with core variables.

These correlations are further visualized in Figure 4, illustrating the strength of linear relationships using a color-coded heatmap.

These findings align with existing literature emphasizing the importance of workforce development and technological innovation in optimizing HSC (e.g. Altay et al., 2024). The weak correlations between demographic variables and the primary constructs suggest that skills and tools influence supply chain performance more than demographic factors. However, future research may examine interactive effects in more detail. Figure 4 presents a correlation matrix heatmap, visually emphasizing these robust positive relationships among HS, BDPA, HSCP and RS.

We conducted structural equation modeling (SEM) using maximum likelihood estimation to test the hypothesized relationships. The analysis includes model fit assessment, mediation testing for hypotheses H1–H4 and moderation testing for hypotheses H5–H6. Model fit indices were satisfactory (χ²/df = 2.318, CFI = 0.971, TLI = 0.964, RMSEA = 0.045, SRMR = 0.041), indicating a good model fit (Hair and Alamer, 2022).

As shown in Table 8, HS have a significant positive effect on BDPA (β = 0.838, p < 0.001), supporting H2. BDPA significantly improves HSCP (β = 0.841, p < 0.01), supporting H3. The direct effect of HS on HSCP is not significant (β = 0.205, p = 0.399), so H1 is not supported. However, the indirect effect of HS on HSCP through BDPA is significant (β = 0.705, p < 0.01), supporting H4 and indicating full mediation.

The SEM analysis demonstrated excellent model fit (CFI = 0.97, TLI = 0.96, RMSEA = 0.045, SRMR = 0.041). We can see from Table 9, the results revealed a strong positive effect of HS on BDPA (β = 0.838, p < 0.001), supporting previous findings (e.g. Tzeng et al., 2024) on the role of technical skills in driving analytics adoption. BDPA, in turn, significantly enhanced HSCP (β = 0.841, p < 0.01), consistent with Liu et al. (2024) on the transformative potential of big data. Notably, the direct effect of HS on HSCP was not significant (β = 0.205, p = 0.399), while the significant indirect effect via BDPA (β = 0.705, p < 0.01) confirmed complete mediation.

Furthermore, bootstrapping (5,000 resamples) validated these results, and RS significantly moderated the BDPA–HSCP relationship (β = 0.27, p < 0.05). These findings robustly support our theoretical framework, highlighting BDPA’s critical intermediary role in leveraging HS for improved supply chain performance.

Figure 5 illustrates the SEM path diagram. Standardized coefficients are displayed along each path. The diagram confirms the mediating role of BDPA between HS and HSCP and highlights the moderation effect of RS.

Figure 6 presents the structural mediation model, illustrating the indirect effect of HS on HSCP through BDPA. The regression lines and scatterplots visualize HS, BDPA and HSCP relationships. The orange line represents the path from HS to BDPA, while green and blue points represent the effects of BDPA and HS on HSCP, respectively. Dashed lines indicate the 95% confidence intervals. The figure supports a complete mediation model consistent with H4. Since RS is not included in this model, no moderation effect is reflected here.

Hierarchical regression models including interaction terms were constructed to examine the moderating role of RS in the relationship between HS and HSCP; we can see from Table 10 below, as well as between BDPA and HSCP. The results indicate that neither interaction effect reached statistical significance, suggesting that RS does not moderate the proposed relationships as hypothesized in H5 and H6.

This study used SEM to examine the relationships among HS, BDPA, RS and HSCP. The findings confirmed that BDPA fully mediates the impact of HS on HSCP (H3 and H4 supported), while the moderating effects of RS on both HS → HSCP and BDPA → HSCP were found to be non-significant (H5 and H6 not supported). The proposed model demonstrated a good fit, and the Fornell–Larcker and HTMT tests supported construct validity. The correlation analysis also revealed strong associations among the core constructs, particularly between HS and BDPA, and between BDPA and HSCP.

This study advances the theoretical discourse on HSCs by integrating the RBV and DCT with emerging insights from behavioral supply chain research. Specifically, it proposes a novel triadic mechanism in which HS, conceptualized as intangible organizational resources, influence HSCP not directly, but through the mediating role of BDPA. This structural pathway redefines how human-centric competencies are transformed into operational performance outcomes in high-stress, resource-constrained environments. Theoretically, this study makes three key contributions:

First, it reframes HS as an enabling condition rather than a direct performance driver, suggesting that its value is contingent on an organization’s ability to convert human insight into data-driven decisions. Unlike earlier studies that either focused solely on soft skills (e.g. Scholten et al., 2019) or the standalone effects of analytics (e.g. Dubey et al., 2019a), our findings demonstrate that BDPA fully mediates the HS–HSCP relationship, thus enriching the understanding of capability complementarities.

Second, empirically validating BDPA as a dynamic capability, this study extends DCT by showing how NGOs in turbulent humanitarian environments reconfigure internal resources to adapt to data-centric demands. While earlier works (e.g. Bahrami et al., 2022; Jeble et al., 2020) established BDPA’s role in commercial supply chains, few have tested its mediating function in non-profit or disaster-response settings, particularly in China’s sociocultural context. This contributes to the contextual deepening of both RBV and DCT by accounting for non-market logics and institutional hybridity.

Third, this study broadens the geographical and institutional scope of humanitarian supply chain research by focusing on Chinese NGOs, an understudied organizational type in global HSC literature. Examining how HS and BDPA interact in a collectivist, hierarchical and rapidly digitizing environment offers a culturally grounded model that contrasts with Western-centric studies rooted in decentralization and market competition. As such, it contributes to theory-building in emerging markets and aligns with calls for decentering Global North paradigms in supply chain theory (Altay et al., 2024).

Furthermore, the study responds to recent scholarly demands (Tickle et al., 2024) to move beyond isolated variables and adopt multi-level, process-oriented frameworks in HSC research. The integrated model presented herein illustrates not only the sequential logic of capability interaction (HS → BDPA → HSCP) but also tests the boundary role of RS, even though moderation was not supported empirically. This negative result is theoretically significant: it suggests that RS may operate more as a latent contextual factor in highly volatile and institutionalized settings rather than a statistically observable moderator – an area warranting further conceptual unpacking.

In summary, this study offers a theoretically grounded, empirically validated and contextually sensitive contribution to the literature. It extends the humanitarian supply chain theory frontier by positioning BDPA as a capability conduit, HS as foundational input and NGO-specific conditions as the operating terrain for dynamic capability activation.

This study provides actionable insights for practitioners engaged in humanitarian logistics and emergency response operations. The results underscore the importance of cultivating HS such as communication, decision-making and situational awareness to enhance organizational responsiveness. In environments where agility and adaptability are critical, personnel with these skills are more likely to drive supply chain efficiency and performance.

Furthermore, the findings emphasize the role of BDPA as a key enabler for HSCP. Organizations with advanced BDPA capabilities can leverage real-time information to predict needs, optimize resource allocation and coordinate relief efforts more effectively. This is particularly vital in high-uncertainty scenarios, such as natural disasters or public health emergencies, where traditional logistics systems often fail to deliver timely responses.

Although RS did not exhibit a statistically significant moderating effect in our model, its direct correlation with other constructs suggests that RS remains an essential organizational attribute. Practitioners should continue to invest in strategies that foster RS, such as scenario planning, redundant resource networks and decentralized decision-making, to buffer against supply chain disruptions.

These implications encourage NGOs, humanitarian agencies and supply chain managers to build human-centric capabilities and data-driven infrastructures simultaneously. A balanced approach that integrates human judgment with analytical tools appears to be most effective in enhancing supply chain outcomes during crises.

To reinforce the distinctiveness of the present study and address the reviewer’s concern regarding theoretical originality, Table 11 provides a comparative summary of recent empirical research focused on HS, BDPA and HSCP. The table highlights key methodological elements, variable combinations and analytical strategies from selected studies between 2023 and 2024.

In contrast to prior works that either focus on single-factor models or omit the interplay between human and analytical capabilities, this study proposes a dual-pathway model. Specifically, it integrates human-centric (HS) and data-driven (BDPA) competencies into a sequential mediation structure, offering a more holistic and process-oriented explanation of HSCP. Moreover, including RS as a boundary condition adds theoretical depth by examining under what conditions BDPA translates more effectively into supply chain performance.

This comparative lens positions our research within the frontier of humanitarian logistics literature and illustrates its value in bridging behavioral and analytical domains – an area still underexplored in non-profit and NGO contexts.

This study investigated the impact of HS and BDPA on HSCP, using RS as a moderating variable. Drawing on survey data collected from 785 NGO workers and professionals engaged in humanitarian operations across China, the findings confirm that BDPA fully mediates the relationship between HS and HSCP. However, contrary to expectations, RS did not exert a statistically significant moderating effect.

The results underscore the growing importance of analytical capabilities, such as predictive analytics, in enhancing supply chain agility and effectiveness, particularly in complex humanitarian contexts. While HS remains foundational, its influence on performance appears to be channeled through deploying data-driven tools and techniques. These findings align with and extend existing theories in humanitarian logistics by demonstrating the functional transition from human-centric capacities to technology-mediated capabilities.

Overall, the study contributes to the literature by offering an empirically validated framework that integrates behavioral and digital capabilities, advancing our understanding of capability building in non-profit and crisis response organizations. The study’s limitations and future research directions are discussed in the following section.

While this study offers important insights into the interplay between HS, BDPA and HSCP, several limitations must be acknowledged.

First, the research adopts a cross-sectional survey design, which restricts the ability to infer causal relationships over time. Although the SEM framework captures directional associations, future studies may benefit from longitudinal designs to observe how capability interactions evolve during different phases of disaster response.

Second, the sample is restricted to Chinese NGOs, limiting the generalizability of findings to broader humanitarian contexts. While focusing on China’s unique institutional and cultural environment provides valuable localized insights, future research should explore whether similar mechanisms hold across diverse geopolitical regions, including fragile states or decentralized aid systems in the Global South.

Third, RS was operationalized as a unidimensional construct, which may have constrained its explanatory power. Given its conceptual complexity, future studies should consider disaggregating RS into psychological, organizational and systemic components to capture its moderating potential better.

Fourth, the study focuses on human and technological capabilities but does not account for contextual factors such as digital infrastructure readiness, inter-agency coordination or organizational learning culture. Including these variables could enhance understanding of the boundary conditions under which BDPA and HS generate value.

Finally, although the mediating role of BDPA was statistically confirmed, the model does not explore other possible mechanisms, such as absorptive capacity or digital leadership, that might also channel the effects of HS. Future research may incorporate these constructs to expand the theoretical scope and test competing mediation pathways.

In conclusion, future studies should embrace multi-method approaches, cross-country comparisons and dynamic modeling techniques to deepen the understanding of how human and digital capabilities interact under varying institutional pressures. Such research would contribute to theory and practice by developing more resilient, data-driven and human-centered HSC.