Economics and Management Empowered by Data Science: An Empirical Investigation of Data-Driven Decision-Making, Organizational Performance, and Sustainable Competitive Advantage

1. Introduction

The emergence of Industry 4.0 has fundamentally reshaped the global economic and managerial landscape by integrating cyber-physical systems, the Internet of Things (IoT), cloud computing, artificial intelligence (AI), robotics, blockchain, digital twins, and big data analytics into production and organizational processes. Unlike previous industrial revolutions that were driven by mechanization, electrification, and automation, Industry 4.0 is characterized by intelligent interconnected systems capable of autonomous communication, predictive learning, and real-time decision-making. These technological innovations have transformed organizations from data-generating entities into data-driven enterprises where information has become a strategic organizational asset. Consequently, economics and management have evolved beyond traditional theories of resource allocation and administrative control toward dynamic, evidence-based, and algorithm-assisted decision environments (Schwab, 2016; Kagermann et al., 2013; Lasi et al., 2014; Porter & Heppelmann, 2014; Lee et al., 2015; Xu et al., 2018; Frank et al., 2019; Vial, 2019).

Parallel to Industry 4.0 is the rapid acceleration of digital transformation, which has become a strategic imperative across governments, private enterprises, and non-profit organizations. Digital transformation extends beyond the mere adoption of information technologies; it encompasses the redesign of organizational structures, business models, customer engagement strategies, operational processes, and value creation mechanisms through digital technologies. Organizations increasingly leverage digital platforms, cloud infrastructures, mobile computing, social media analytics, and intelligent automation to improve operational efficiency, enhance customer experiences, and achieve sustainable competitive advantage. Studies consistently demonstrate that digital transformation positively influences organizational agility, innovation capability, knowledge management, and firm performance by enabling faster information processing and adaptive strategic responses (Bharadwaj et al., 2013; Hess et al., 2016; Warner & Wäger, 2019; Verhoef et al., 2021; Kane et al., 2015; Sebastian et al., 2017; Matt et al., 2015; Fitzgerald et al., 2014).

Central to digital transformation is the Big Data revolution, which has significantly altered the methods through which economic and managerial decisions are formulated. Contemporary organizations generate unprecedented volumes of structured and unstructured data from enterprise systems, social media platforms, digital transactions, sensors, wearable devices, satellite technologies, and Internet-enabled equipment. The exponential growth in data volume, velocity, variety, veracity, and value has created enormous opportunities for predictive analytics, operational optimization, customer intelligence, and strategic forecasting. Big data analytics enables organizations to uncover hidden patterns, forecast market trends, reduce operational uncertainties, and optimize resource allocation through advanced statistical and computational techniques. Consequently, organizations capable of transforming raw data into actionable knowledge consistently outperform competitors in innovation, operational excellence, and strategic adaptability (McAfee & Brynjolfsson, 2012; Chen et al., 2012; Davenport & Harris, 2007; George et al., 2014; Gandomi & Haider, 2015; Wamba et al., 2017; Akter et al., 2016; Dubey et al., 2019; Gupta & George, 2016).

The integration of artificial intelligence into economics and management has further accelerated the transition toward intelligent organizations. AI-driven economic systems employ machine learning, deep learning, natural language processing, reinforcement learning, and computer vision to automate complex analytical tasks, forecast economic variables, detect financial fraud, optimize pricing strategies, improve supply chain efficiency, and support evidence-based policymaking. In management, AI enhances strategic planning, human resource analytics, marketing personalization, financial risk assessment, and operational decision-making through real-time predictive insights. Rather than replacing managerial judgment, AI increasingly complements human expertise by reducing cognitive biases, improving forecasting accuracy, and facilitating faster responses to environmental uncertainty (Brynjolfsson & McAfee, 2014; Russell & Norvig, 2021; Agrawal et al., 2018; Jordan & Mitchell, 2015; Shrestha et al., 2019; Jarrahi, 2018; Davenport et al., 2020; Raisch & Krakowski, 2021).

The evolution of managerial decision-making reflects a gradual shift from intuition-based management to evidence-based and analytics-driven decision processes. Classical management theories emphasized managerial experience, hierarchical authority, and subjective judgment as primary determinants of organizational success. However, increasing environmental complexity, globalization, digital competition, and market volatility have rendered intuition alone insufficient for addressing contemporary organizational challenges. Modern organizations increasingly rely on descriptive, diagnostic, predictive, and prescriptive analytics to improve strategic planning, operational efficiency, and risk management. Data science has therefore emerged as a multidisciplinary capability integrating statistics, computer science, economics, operations research, and management science to support informed decision-making across organizational functions (Davenport, 2014; Provost & Fawcett, 2013; Power, 2008; Shmueli & Koppius, 2011; Davenport & Harris, 2017; Cao, 2017).

Despite substantial scholarly attention to digital transformation, big data analytics, and artificial intelligence, significant research gaps remain regarding the integrated influence of data science capabilities on economic decision-making and managerial performance. Existing studies predominantly examine isolated technological components, specific industries, or individual organizational outcomes, with relatively limited empirical evidence explaining the mechanisms through which data science simultaneously enhances managerial decision quality, organizational agility, innovation capability, and sustainable competitive advantage. Moreover, empirical findings remain fragmented across disciplinary boundaries, limiting the development of comprehensive theoretical frameworks that integrate economics, management, and data science. This fragmentation underscores the need for multidisciplinary empirical investigations capable of explaining how analytical capabilities function as strategic organizational resources in increasingly digital economies (Gupta & George, 2016; Wamba et al., 2020; Mikalef et al., 2020; Fosso Wamba et al., 2021).

Motivated by these theoretical and practical deficiencies, this study empirically examines how data science capabilities influence managerial decision-making, organizational agility, innovation capability, and organizational performance within contemporary organizations. Specifically, the study seeks to determine the direct and indirect relationships among these constructs using robust quantitative analytical techniques. The research aims to contribute to the growing literature on digital transformation by integrating perspectives from the Resource-Based View (RBV), Dynamic Capabilities Theory, and data-driven management to explain how organizations generate sustainable competitive advantage through analytics capabilities.

Accordingly, the specific objectives are to examine the influence of data science capability on managerial decision quality, investigate its effect on organizational agility and innovation capability, assess the mediating role of organizational agility in enhancing organizational performance, and develop an integrated empirical framework linking economics, management, and data science. By providing empirical evidence on these relationships, this study contributes to theory by extending existing knowledge on digital capabilities as strategic organizational resources, contributes methodologically through the application of advanced structural equation modelling, and offers practical guidance for managers and policymakers seeking to strengthen evidence-based decision-making, digital competitiveness, and sustainable economic development in the era of Industry 4.0 and artificial intelligence.

• Theoretical Framework

This study is anchored on four complementary theoretical perspectives: the Resource-Based View (RBV), Dynamic Capabilities Theory (DCT), Knowledge-Based View (KBV), and the Technology–Organization–Environment (TOE) Framework. Collectively, these theories provide a robust foundation for explaining how data science capabilities enhance managerial decision-making, organizational agility, innovation, and sustainable organizational performance in the digital economy.

The Resource-Based View (RBV), originally advanced by Barney (1991) and Wernerfelt (1984), posits that firms achieve sustainable competitive advantage by possessing valuable, rare, inimitable, and non-substitutable (VRIN) resources. In the context of Industry 4.0, data science capabilities—including big data infrastructure, artificial intelligence (AI), machine learning, cloud computing, predictive analytics, and skilled data professionals—constitute strategic intangible resources that are difficult for competitors to replicate. Organizations capable of acquiring, integrating, and exploiting these analytical resources can improve forecasting accuracy, optimize operational efficiency, reduce uncertainty, and enhance strategic decision-making. Numerous empirical studies have confirmed that analytics capabilities function as strategic organizational assets that significantly improve innovation, operational performance, and competitive advantage (Barney, 1991; Wernerfelt, 1984; Peteraf, 1993; Grant, 1996; Bharadwaj, 2000; Gupta & George, 2016; Akter et al., 2016; Wamba et al., 2017; Dubey et al., 2019; Mikalef et al., 2020).

While RBV explains the possession of valuable resources, Dynamic Capabilities Theory (DCT) extends this perspective by emphasizing an organization's ability to continuously renew, reconfigure, and transform these resources in response to rapidly changing business environments. Developed by Teece, Pisano, and Shuen (1997) and further refined by Teece (2007, 2018), DCT argues that sustainable performance depends not only on owning strategic resources but also on sensing opportunities, seizing emerging market prospects, and transforming organizational processes. Data science significantly strengthens these dynamic capabilities by enabling firms to process real-time information, predict market trends, automate decision processes, detect operational risks, and rapidly respond to environmental changes. Organizations that integrate AI-driven analytics into strategic management become more agile, adaptive, and resilient, particularly under conditions of technological disruption and economic uncertainty (Eisenhardt & Martin, 2000; Helfat et al., 2007; Teece, 2007, 2018; Warner & Wäger, 2019; Verhoef et al., 2021; Fosso Wamba et al., 2021).

The Knowledge-Based View (KBV) complements RBV by recognizing knowledge as the firm's most valuable strategic resource. Building on the works of Grant (1996), Kogut and Zander (1992), and Nonaka and Takeuchi (1995), KBV argues that organizational success depends on the acquisition, integration, creation, and application of knowledge. Data science serves as a knowledge-generation mechanism by transforming large volumes of structured and unstructured data into actionable insights that support managerial learning and evidence-based decision-making. Advanced analytics, machine learning algorithms, and business intelligence systems facilitate organizational knowledge creation, improve problem-solving, and stimulate continuous innovation. Consequently, organizations with stronger analytical knowledge capabilities are better positioned to exploit market opportunities, develop innovative products, and enhance organizational learning (Alavi & Leidner, 2001; Davenport & Prusak, 1998; Chen et al., 2012; George et al., 2014; Cao, 2017; Davenport & Harris, 2017).

The Technology–Organization–Environment (TOE) Framework, proposed by Tornatzky and Fleischer (1990), provides an organizational adoption perspective by explaining how technological innovation is influenced by technological readiness, organizational characteristics, and environmental pressures. Within this framework, the adoption of data science technologies is shaped by factors such as IT infrastructure, top management support, organizational culture, employee competencies, competitive intensity, regulatory environments, and digital ecosystems. The TOE framework has been widely applied to explain organizational adoption of big data analytics, cloud computing, artificial intelligence, and digital transformation initiatives. It suggests that firms possessing adequate technological resources, supportive leadership, and favorable institutional environments are more likely to achieve successful implementation of analytics capabilities and superior organizational outcomes (Oliveira & Martins, 2011; Gangwar et al., 2015; Gupta et al., 2018; Awa et al., 2017; Verhoef et al., 2021).

Integrating these four theoretical perspectives provides a comprehensive explanation of how data science empowers economics and management. RBV explains why analytics capabilities constitute strategic resources; Dynamic Capabilities Theory demonstrates how these resources are continuously reconfigured to sustain competitiveness; KBV highlights the transformation of data into organizational knowledge and innovation; while the TOE framework explains the contextual conditions that facilitate successful technology adoption. Together, these theories underpin the study's proposition that data science capability enhances managerial decision quality, organizational agility, innovation capability, and organizational performance, thereby generating sustainable competitive advantage in the digital economy. This integrated theoretical framework also addresses the interdisciplinary nature of modern economics and management by linking strategic resources, organizational learning, technological adoption, and dynamic adaptation into a unified empirical model.

Research Questions

• To what extent does Data Science Capability (DSC) influence Managerial Decision Quality (MDQ) in organizations?
• What is the effect of Data Science Capability (DSC) on Organizational Agility (OA)?
• How does Data Science Capability (DSC) influence Innovation Capability (IC)?
• To what extent does Organizational Agility (OA) influence Organizational Performance (OP)?
• What is the effect of Innovation Capability (IC) on Organizational Performance (OP)?
• Does Organizational Agility (OA) mediate the relationship between Data Science Capability (DSC) and Organizational Performance (OP)?

Hypotheses

• H1: Data Science Capability positively influences Managerial Decision Quality.
• H2: Data Science Capability positively affects Organizational Agility.
• H3: Data Science Capability positively influences Innovation Capability.
• H4: Organizational Agility positively influences Organizational Performance.
• H5: Innovation Capability positively affects Organizational Performance.
• H6: Organizational Agility mediates the relationship between Data Science Capability and Organizational Performance.

2. Literature Review

2.1. Data Science: Definition, Evolution, and Core Components

Data science has emerged as one of the most transformative interdisciplinary fields of the twenty-first century, fundamentally reshaping economics, management, engineering, healthcare, finance, public administration, and virtually every knowledge-intensive industry. Broadly, data science refers to the systematic process of collecting, managing, processing, analyzing, interpreting, and visualizing structured and unstructured data to generate actionable insights that support decision-making and innovation. It integrates concepts from statistics, computer science, mathematics, artificial intelligence (AI), machine learning, operations research, domain expertise, and information systems to transform raw data into valuable organizational knowledge (Dhar, 2013; Provost & Fawcett, 2013; Cao, 2017; Donoho, 2017; Cleveland, 2001; Saltz & Shamshurin, 2016). Unlike traditional statistical analysis, which primarily focuses on hypothesis testing using relatively small datasets, data science employs advanced computational algorithms, high-performance computing, and intelligent analytical models capable of processing massive volumes of structured, semi-structured, and unstructured data in real time (Chen et al., 2012; Kelleher & Tierney, 2018; Wickham & Grolemund, 2017).

The evolution of data science can be traced to the convergence of statistics, database management, artificial intelligence, and computational science during the late twentieth century. Early analytical practices relied heavily on descriptive statistics and relational databases to summarize historical information for managerial reporting. The rapid expansion of digital technologies, internet connectivity, mobile computing, and sensor networks generated unprecedented volumes of digital data, commonly referred to as the "Big Data Revolution." This phenomenon, characterized by the five dimensions of volume, velocity, variety, veracity, and value, necessitated the development of sophisticated analytical methods capable of extracting meaningful patterns from increasingly complex datasets (Laney, 2001; McAfee & Brynjolfsson, 2012; Gandomi & Haider, 2015; George et al., 2014; Mayer-Schönberger & Cukier, 2013). The advent of Industry 4.0 further accelerated the growth of data science through the integration of cyber-physical systems, the Internet of Things (IoT), cloud computing, blockchain technologies, digital twins, and intelligent automation, thereby enabling organizations to shift from reactive management to predictive and prescriptive decision-making (Schwab, 2016; Kagermann et al., 2013; Lasi et al., 2014; Porter & Heppelmann, 2014; Lee et al., 2015; Xu et al., 2018; Frank et al., 2019; Verhoef et al., 2021).

One of the foundational components of data science is machine learning, which refers to a class of computational techniques that enable computer systems to learn patterns from data and improve their performance without being explicitly programmed. Machine learning algorithms—including supervised, unsupervised, semi-supervised, and reinforcement learning models—are extensively employed for forecasting, customer segmentation, fraud detection, recommendation systems, predictive maintenance, financial risk assessment, and demand forecasting. The increasing availability of large datasets and advances in computational power have significantly enhanced the effectiveness of machine learning across economic and managerial applications, enabling organizations to automate complex analytical tasks and improve decision accuracy (Mitchell, 1997; Bishop, 2006; Murphy, 2012; Goodfellow et al., 2016; Hastie et al., 2009; Jordan & Mitchell, 2015; Alpaydin, 2020).

Closely related to machine learning is artificial intelligence (AI), which encompasses computational systems capable of performing tasks that traditionally require human intelligence, including reasoning, learning, perception, language understanding, and decision-making. AI integrates machine learning, natural language processing, expert systems, robotics, and computer vision to support intelligent organizational processes. In economics, AI is increasingly applied to macroeconomic forecasting, financial market analysis, credit scoring, algorithmic trading, and public policy evaluation. In management, AI facilitates strategic planning, human resource analytics, customer relationship management, marketing personalization, and supply chain optimization. Rather than replacing human managers, AI increasingly augments managerial capabilities by providing evidence-based recommendations and reducing cognitive biases in decision-making (Russell & Norvig, 2021; Agrawal et al., 2018; Brynjolfsson & McAfee, 2014; Jarrahi, 2018; Shrestha et al., 2019; Davenport et al., 2020; Raisch & Krakowski, 2021).

Another critical component is predictive analytics, which applies statistical modelling, machine learning, and forecasting techniques to estimate future outcomes based on historical and current data. Predictive analytics enables organizations to anticipate customer behaviour, forecast sales, optimize inventory, detect equipment failures, identify financial risks, and improve strategic planning under uncertainty. By transforming historical data into forward-looking insights, predictive analytics enhances organizational agility, resource allocation, and operational efficiency while supporting proactive rather than reactive management (Shmueli & Koppius, 2011; Siegel, 2016; Davenport & Harris, 2007; Provost & Fawcett, 2013; Power, 2008).

Data mining constitutes another essential pillar of data science by focusing on the discovery of hidden patterns, relationships, anomalies, and knowledge from large datasets using statistical, mathematical, and computational techniques. It employs clustering, classification, association rule mining, anomaly detection, and sequential pattern analysis to uncover valuable information that may not be immediately observable. Data mining has become indispensable across banking, healthcare, manufacturing, retail, telecommunications, and government sectors for fraud detection, customer profiling, quality control, market basket analysis, and policy evaluation. Its integration with machine learning and AI has significantly enhanced organizational intelligence and evidence-based management (Han et al., 2012; Witten et al., 2016; Fayyad et al., 1996; Berry & Linoff, 2011; Tan et al., 2019).

The rapid growth of data science has also been facilitated by cloud computing, which provides scalable computing resources, distributed storage, and on-demand analytical services over the internet. Cloud computing enables organizations to process massive datasets without substantial investments in physical infrastructure while supporting collaborative analytics, remote access, high-performance computing, and enterprise-wide digital transformation. Cloud-based platforms such as Software-as-a-Service (SaaS), Platform-as-a-Service (PaaS), and Infrastructure-as-a-Service (IaaS) have democratized access to advanced analytical capabilities, allowing organizations of all sizes to implement AI, machine learning, and big data solutions more efficiently (Mell & Grance, 2011; Armbrust et al., 2010; Marston et al., 2011; Hashem et al., 2015; Buyya et al., 2019).

Complementing these technologies is Business Intelligence (BI), which represents the organizational processes, architectures, and technologies used to collect, integrate, analyze, visualize, and communicate business information for strategic and operational decision-making. Business intelligence transforms raw organizational data into meaningful dashboards, reports, scorecards, and performance indicators that support executives in monitoring organizational performance and identifying strategic opportunities. Modern BI systems increasingly integrate with AI, machine learning, cloud computing, and predictive analytics to provide real-time insights and intelligent decision support. Empirical evidence consistently demonstrates that business intelligence significantly enhances organizational learning, innovation capability, operational efficiency, customer satisfaction, and competitive advantage across industries (Watson & Wixom, 2007; Negash, 2004; Wixom & Watson, 2010; Chen et al., 2012; Elbashir et al., 2011; Popovič et al., 2012). As organizations continue to navigate the complexities of Industry 4.0 and the digital economy, investments in integrated data science capabilities are increasingly recognized as essential drivers of productivity, resilience, and long-term organizational performance (Gupta & George, 2016; Akter et al., 2016; Wamba et al., 2017; Dubey et al., 2019; Mikalef et al., 2020; Fosso Wamba et al., 2021).

2.2. Economics and Data Science

The convergence of economics and data science has transformed the way economists, policymakers, financial institutions, and business organizations understand, predict, and respond to complex economic phenomena. Traditionally, economic analysis relied on theoretical models, econometric techniques, and relatively small datasets to explain market behavior and guide policy decisions. However, the exponential growth of digital technologies, high-frequency data, cloud computing, artificial intelligence (AI), and machine learning has significantly expanded the analytical capabilities available to economists. Data science enables the processing of massive volumes of structured and unstructured data, allowing economic relationships to be examined with greater precision, speed, and predictive accuracy than conventional econometric methods. Consequently, data-driven economics has emerged as a multidisciplinary field integrating economics, statistics, computer science, operations research, and artificial intelligence to improve forecasting, policy evaluation, financial analysis, and resource allocation (Varian, 2014; Athey, 2018; Mullainathan & Spiess, 2017; Einav & Levin, 2014; Brynjolfsson & McAfee, 2014; Agrawal et al., 2018; Provost & Fawcett, 2013).

One of the most significant applications of data science in economics is Gross Domestic Product (GDP) forecasting. GDP serves as a primary indicator of national economic performance, influencing monetary policy, fiscal planning, investment decisions, and international trade. Conventional GDP forecasting relies on macroeconomic indicators and econometric models such as autoregressive integrated moving average (ARIMA), vector autoregression (VAR), and dynamic stochastic general equilibrium (DSGE) models. While these models have proven useful, they often struggle to capture nonlinear relationships and rapidly changing economic conditions. Data science introduces machine learning algorithms—including random forests, support vector machines, artificial neural networks, gradient boosting, and deep learning—which significantly improve forecasting accuracy by integrating diverse data sources such as satellite imagery, electronic transactions, internet search activity, social media sentiment, and mobile phone data. These technologies enable policymakers and central banks to conduct near real-time economic monitoring and produce more reliable macroeconomic forecasts (Athey, 2018; Varian, 2014; Mullainathan & Spiess, 2017; Choi & Varian, 2012; McLaren & Shanbhogue, 2011; Chen et al., 2012; Gandomi & Haider, 2015).

Closely related is inflation modelling, where data science has substantially enhanced the ability of economists to forecast price movements and evaluate monetary policy effectiveness. Inflation is influenced by numerous interconnected factors including money supply, exchange rates, energy prices, consumer demand, wages, and global supply chains. Traditional econometric models often face limitations when addressing nonlinear interactions and structural economic changes. Machine learning algorithms overcome these constraints by identifying hidden patterns within large datasets and continuously updating predictions as new information becomes available. Central banks increasingly employ AI-driven forecasting models that integrate high-frequency financial data, online retail prices, consumer transactions, and web-based indicators to improve inflation prediction and policy formulation (Stock & Watson, 2007; Makridakis et al., 2018; Jordan & Mitchell, 2015; Athey, 2018; Gu et al., 2020; Brynjolfsson & McAfee, 2014).

Another important area is labour economics, where data science has revolutionized the analysis of employment patterns, wage determination, workforce productivity, skills development, and labour market dynamics. Governments and international organizations increasingly utilize administrative databases, digital employment platforms, mobile phone records, and online job advertisements to monitor labour demand and supply in real time. Machine learning techniques assist researchers in predicting unemployment trends, identifying skill shortages, evaluating education policies, and assessing the impact of technological change on employment. Human resource analytics also enables firms to optimize recruitment, employee retention, performance management, and workforce planning through predictive modelling and behavioral analytics. These developments provide more accurate and timely labour market intelligence than traditional survey-based approaches (Autor, 2015; Acemoglu & Restrepo, 2018; Einav & Levin, 2014; Brynjolfsson & McAfee, 2014; Bessen, 2019; Goldfarb & Tucker, 2019).

The integration of data science into financial markets has significantly transformed investment management, banking, insurance, and capital market operations. Financial institutions generate massive volumes of transactional data, market prices, trading activities, customer information, and macroeconomic indicators, making the sector particularly suitable for advanced analytics. Machine learning algorithms are extensively used in algorithmic trading, portfolio optimization, credit risk assessment, fraud detection, asset pricing, and financial forecasting. Artificial intelligence enables high-frequency trading systems to process millions of transactions within milliseconds while continuously adapting to changing market conditions. Likewise, predictive analytics assists banks in credit scoring, loan default prediction, anti-money laundering compliance, and cybersecurity. These innovations improve financial efficiency, market transparency, and risk-adjusted investment decisions (Fama, 1970; Lo, 2004; Gu et al., 2020; Hull, 2022; Jorion, 2007; Jordan & Mitchell, 2015; Agrawal et al., 2018).

Data science has also transformed the study of consumer behaviour by enabling organizations to understand customer preferences, purchasing patterns, and decision-making processes using large-scale behavioral datasets. Traditional consumer research relied on surveys, interviews, and observational studies, which were often constrained by limited sample sizes and recall bias. Contemporary businesses now analyze online transactions, loyalty programmes, clickstream data, mobile applications, social media interactions, and digital payment systems to generate real-time insights into consumer preferences. Machine learning algorithms support customer segmentation, personalized recommendations, dynamic pricing, demand forecasting, and targeted marketing campaigns, allowing firms to improve customer satisfaction and competitive positioning. Behavioural economists increasingly combine AI with experimental economics to examine consumer decision-making under uncertainty, thereby enriching both economic theory and marketing practice (Kotler & Keller, 2016; Wedel & Kannan, 2016; Davenport & Harris, 2007; Provost & Fawcett, 2013; Shmueli & Koppius, 2011; George et al., 2014).

A further application lies in risk management, where data science enhances the identification, measurement, prediction, and mitigation of financial, operational, strategic, and systemic risks. Predictive analytics enables organizations to anticipate potential disruptions arising from market volatility, cybersecurity threats, supply chain interruptions, environmental disasters, and macroeconomic instability. Financial institutions employ machine learning models to estimate credit risk, operational losses, insurance claims, and investment risks with greater precision than traditional statistical approaches. Governments similarly use predictive analytics to monitor systemic financial vulnerabilities and evaluate economic resilience under different policy scenarios. The integration of AI into enterprise risk management supports proactive rather than reactive decision-making by enabling continuous monitoring of emerging risks (Jorion, 2007; Basel Committee on Banking Supervision, 2019; Hull, 2022; Power, 2008; Davenport & Harris, 2017; Goodfellow et al., 2016).

Finally, the emergence of the digital economy represents perhaps the most profound intersection between economics and data science. The digital economy is characterized by economic activities driven by digital technologies, internet platforms, cloud computing, e-commerce, digital payments, blockchain, artificial intelligence, and data-driven innovation. Data itself has become a strategic economic asset, often described as the "new oil," because of its capacity to generate value through analytics and intelligent decision-making. Digital platforms such as online marketplaces, fintech ecosystems, and sharing economy businesses rely extensively on data science to optimize pricing, match buyers and sellers, improve logistics, detect fraud, and personalize customer experiences. Governments also employ big data analytics to improve tax administration, public service delivery, healthcare management, smart city development, and evidence-based policymaking. As digital transformation accelerates globally, the integration of economics and data science has become indispensable for promoting productivity, innovation, inclusive growth, and sustainable development (OECD, 2020; World Bank, 2021; UNCTAD, 2021; Brynjolfsson & McAfee, 2014; Schwab, 2016; Verhoef et al., 2021; Mikalef et al., 2020; Wamba et al., 2020; Fosso Wamba et al., 2021).

2.3. Management and Data Science

The integration of data science into management has fundamentally transformed organizational planning, decision-making, performance management, and value creation. Modern organizations increasingly operate in highly competitive, uncertain, and digitally connected environments characterized by rapidly changing customer preferences, technological disruptions, and global market volatility. Consequently, managerial decisions are no longer based solely on intuition, experience, or historical reports but increasingly rely on real-time data, predictive analytics, artificial intelligence (AI), and machine learning. Data science empowers managers to convert massive volumes of structured and unstructured data into actionable insights that improve organizational agility, innovation, efficiency, and strategic competitiveness. The convergence of management and data science therefore represents one of the defining characteristics of Industry 4.0 and digital transformation (Davenport & Harris, 2007; Provost & Fawcett, 2013; Brynjolfsson & McAfee, 2014; Cao, 2017; Verhoef et al., 2021; Wamba et al., 2020; Mikalef et al., 2020).

In strategic management, data science enhances organizational competitiveness by supporting evidence-based strategic planning, environmental scanning, competitive intelligence, and long-term forecasting. Traditional strategic planning depended heavily on managerial intuition and historical performance indicators, whereas contemporary organizations employ predictive analytics, scenario modelling, and AI-powered decision support systems to anticipate market changes and identify emerging opportunities. Data science enables executives to monitor customer trends, competitor behaviour, technological developments, and macroeconomic conditions in real time, thereby improving strategic responsiveness and organizational resilience. Organizations with mature analytics capabilities consistently demonstrate superior strategic flexibility, innovation, and sustainable competitive advantage (Porter, 1985; Teece et al., 1997; Davenport & Harris, 2007; Bharadwaj et al., 2013; Warner & Wäger, 2019; Verhoef et al., 2021).

Marketing analytics has become one of the most significant applications of data science in management. Organizations collect enormous quantities of customer data from e-commerce platforms, social media, mobile applications, digital payment systems, and customer relationship management (CRM) systems. Machine learning algorithms analyze these datasets to understand consumer preferences, predict purchasing behaviour, personalize advertising campaigns, optimize pricing strategies, and improve customer retention. Predictive analytics also enables firms to forecast market demand, evaluate marketing effectiveness, and allocate promotional resources more efficiently. Consequently, marketing decisions increasingly rely on real-time analytics rather than traditional market surveys alone (Kotler & Keller, 2016; Wedel & Kannan, 2016; Chaffey, 2019; Davenport & Harris, 2007; Shmueli & Koppius, 2011).

The emergence of Human Resource Analytics (HR Analytics) has similarly transformed workforce management by enabling organizations to optimize recruitment, employee engagement, talent development, succession planning, and performance evaluation. HR analytics integrates employee demographic information, performance metrics, behavioural data, training records, and workforce surveys to generate predictive insights regarding employee productivity, turnover intentions, absenteeism, and leadership potential. Artificial intelligence further enhances recruitment by automating candidate screening, competency matching, and workforce planning. These analytical approaches improve organizational productivity while reducing recruitment costs and employee turnover (Bassi, 2011; Marler & Boudreau, 2017; Rasmussen & Ulrich, 2015; Fitz-enz, 2010; Davenport et al., 2010).

Data science has also revolutionized Supply Chain Analytics, enabling organizations to improve logistics, inventory management, procurement, transportation, and supplier relationship management. Advanced analytical models predict customer demand, optimize warehouse operations, monitor supplier performance, detect operational disruptions, and enhance supply chain visibility through Internet of Things (IoT) technologies and real-time monitoring systems. Machine learning algorithms assist organizations in minimizing inventory costs while maintaining high service levels, thereby improving operational efficiency and resilience against supply chain disruptions (Christopher, 2016; Ivanov et al., 2019; Waller & Fawcett, 2013; Gunasekaran et al., 2017; Dubey et al., 2019).

In financial analytics, organizations utilize big data, AI, and predictive modelling to support budgeting, investment decisions, credit assessment, fraud detection, liquidity management, and financial forecasting. Financial analytics enables managers to evaluate organizational performance, optimize capital allocation, manage financial risks, and improve corporate governance through evidence-based insights. Machine learning models have significantly enhanced forecasting accuracy compared to conventional statistical methods by identifying complex nonlinear relationships within financial datasets (Hull, 2022; Jorion, 2007; Fama, 1970; Lo, 2004; Gu et al., 2020).

The application of data science within healthcare management has improved clinical decision-making, hospital administration, disease surveillance, patient outcome prediction, resource allocation, and personalized medicine. Healthcare organizations increasingly employ predictive analytics to anticipate disease outbreaks, optimize staffing, improve diagnostic accuracy, and reduce operational costs. Electronic health records, wearable technologies, and AI-powered diagnostic systems generate large datasets that facilitate evidence-based healthcare management while improving service quality and patient satisfaction (Topol, 2019; Obermeyer & Emanuel, 2016; Raghupathi & Raghupathi, 2014; Jiang et al., 2017).

Similarly, public sector analytics has become an essential component of modern governance by supporting evidence-based policymaking, digital government, public financial management, tax administration, urban planning, crime prevention, disaster management, and social welfare programmes. Governments increasingly utilize big data and AI to improve transparency, optimize public expenditure, enhance service delivery, and monitor policy implementation. Public sector analytics contributes to more efficient governance through improved accountability, resource optimization, and citizen engagement (Janssen et al., 2017; Kitchin, 2014; Meijer & Bolívar, 2016; OECD, 2020; World Bank, 2021).

2.4. Organizational Performance

Organizational performance refers to the extent to which an organization achieves its strategic, operational, financial, and non-financial objectives through the effective utilization of available resources. It is a multidimensional construct encompassing financial profitability, operational efficiency, customer satisfaction, innovation, employee productivity, market competitiveness, and environmental sustainability. Contemporary management literature argues that organizational performance extends beyond traditional financial indicators to include intangible assets such as knowledge, innovation capability, organizational learning, digital maturity, and stakeholder value creation. Consequently, organizations are increasingly evaluated using both financial and non-financial performance metrics that reflect long-term sustainability rather than short-term profitability alone (Kaplan & Norton, 1996; Richard et al., 2009; Venkatraman & Ramanujam, 1986).

In the era of Industry 4.0, organizational performance has become closely linked to digital transformation and data-driven decision-making. The emergence of big data analytics, artificial intelligence (AI), machine learning, cloud computing, and predictive analytics has transformed how organizations monitor performance, optimize operations, and respond to environmental uncertainties. Organizations that successfully integrate data science into their strategic and operational processes are better positioned to improve productivity, reduce operational costs, enhance customer experiences, accelerate innovation, and strengthen competitive advantage. Data-driven organizations utilize real-time information to optimize supply chains, predict consumer behaviour, improve financial planning, and enhance risk management, thereby creating sustainable organizational value (Davenport & Harris, 2007; Brynjolfsson & McAfee, 2014; Bharadwaj et al., 2013; Gupta & George, 2016; Mikalef et al., 2020).

From the perspective of the Resource-Based View (RBV), organizational performance is determined by an organization's ability to develop and effectively deploy valuable, rare, inimitable, and well-organized resources. Data science capability has increasingly been recognized as one of these strategic resources because it enhances organizational knowledge, improves managerial decision-making, and strengthens innovation capability. Firms that invest in advanced analytics infrastructure, highly skilled personnel, and data-driven cultures consistently outperform competitors by generating superior financial returns and operational excellence (Barney, 1991; Wernerfelt, 1984; Teece, 2007). Similarly, Dynamic Capabilities Theory argues that superior performance depends on an organization's capacity to sense environmental changes, seize emerging opportunities, and continuously transform internal resources to adapt to rapidly changing markets (Teece et al., 1997; Teece, 2007).

Organizational performance is commonly assessed using both objective and subjective indicators. Objective measures include profitability, return on investment (ROI), return on assets (ROA), market share, productivity, and sales growth, while subjective measures evaluate customer satisfaction, service quality, employee engagement, innovation performance, organizational agility, and environmental sustainability. The Balanced Scorecard developed by Kaplan and Norton (1996) remains one of the most comprehensive frameworks for measuring organizational performance because it integrates financial performance with customer, internal business process, and learning and growth perspectives. This multidimensional approach provides a holistic assessment of organizational success and has been widely adopted in empirical management research.

Recent empirical studies have consistently reported a positive relationship between digital transformation, analytics capability, and organizational performance. Organizations that effectively leverage artificial intelligence, predictive analytics, and business intelligence experience improved operational efficiency, enhanced innovation capability, faster strategic decision-making, and greater resilience during periods of economic uncertainty. These capabilities enable firms to identify market opportunities, anticipate customer needs, and respond proactively to competitive pressures, ultimately leading to sustained organizational growth and superior performance (Wamba et al., 2017; Verhoef et al., 2021; Fosso Wamba et al., 2015).

The Nigerian management literature similarly recognizes organizational performance as a function of effective leadership, strategic management, knowledge management, and human resource practices. Empirical studies by Ukpong have consistently demonstrated that organizational performance improves significantly when organizations adopt effective recruitment practices, strategic leadership, employee engagement, and sound human resource management practices (Ukpong & Nissi, 2019; Ukpong et al., 2020a; Ukpong et al., 2020b). For example, research on recruitment practices established that transparent, competency-based, and merit-driven recruitment systems positively influence organizational efficiency, employee productivity, and corporate performance in manufacturing firms (Ukpong et al., 2020b). Similarly, studies on transformational leadership, particularly idealized influence, found that leadership behaviours significantly enhance employee engagement, organizational commitment, and overall organizational effectiveness within Nigeria's oil and gas sector (Ukpong et al., 2019a). These empirical findings suggest that sustainable organizational performance depends not only on technological resources but also on effective managerial practices, strategic leadership, organizational learning, and human capital development. Consequently, organizations that combine advanced data science capabilities with effective leadership and sound human resource practices are more likely to achieve superior financial performance, operational excellence, innovation, and long-term competitive advantage (Barney, 1991; Teece, 2007; Gupta & George, 2016; Mikalef et al., 2020).

Within the context of this study, organizational performance is conceptualized as the overall effectiveness with which organizations achieve their strategic objectives through improved managerial decision quality, organizational agility, innovation capability, and data science capability. It is measured across multiple dimensions, including financial performance, operational performance, innovation performance, market performance, and environmental performance. This multidimensional perspective recognizes that organizations operating in the digital economy must simultaneously achieve economic efficiency, technological innovation, customer value creation, and sustainable competitive advantage. Consequently, organizations that develop strong data science capabilities are expected to achieve superior organizational performance by transforming data into strategic knowledge, enhancing managerial effectiveness, and continuously adapting to the dynamic business environment (Hair et al., 2022; Richard et al., 2009; Kaplan & Norton, 1996).

2.5. Conceptual Framework

The conceptual framework proposes that Data Science Capability (DSC) serves as the primary strategic resource that enhances Organizational Performance (OP) through the sequential mediating roles of Managerial Decision Quality (MDQ), Organizational Agility (OA), and Innovation Capability (IC). Organizations with advanced capabilities in big data analytics, artificial intelligence, machine learning, predictive analytics, cloud computing, and business intelligence are better positioned to transform data into actionable insights, enabling managers to make timely, accurate, and evidence-based decisions. Improved decision quality enhances organizational agility by strengthening the ability to sense environmental changes, respond rapidly to market dynamics, and adapt to technological disruptions. Greater agility, in turn, promotes innovation capability through continuous learning, knowledge integration, process improvement, and the development of new products, services, and business models, ultimately leading to superior financial, operational, market, innovation, and environmental performance (Gupta & George, 2016; Davenport & Harris, 2007; Mikalef et al., 2020; Wamba et al., 2017).

The framework integrates the Resource-Based View (RBV), Dynamic Capabilities Theory (DCT), and the Technology Acceptance Model (TAM). RBV views data science capability as a valuable, rare, inimitable, and organizationally embedded resource (Barney, 1991; Wernerfelt, 1984), while DCT explains how organizations sense opportunities, seize technological advancements, and transform resources to sustain competitive advantage (Teece et al., 1997; Teece, 2007). TAM further explains that managers' perceptions of the usefulness and ease of use of analytical technologies influence their adoption and effective utilization (Davis, 1989). Consequently, the framework provides the theoretical basis for the study's hypotheses and PLS-SEM model (Hair et al., 2022; Sarstedt et al., 2022).

3. Methodology

This study adopted a quantitative research approach based on the positivist research philosophy to empirically examine the relationships between data science capability, managerial decision quality, organizational agility, innovation capability, and organizational performance. A quantitative approach was considered appropriate because it facilitates objective measurement of latent constructs, hypothesis testing, and statistical generalization of findings across organizations. The study employed a cross-sectional survey design, enabling data to be collected from respondents at a single point in time to investigate causal relationships among the study variables (Creswell & Creswell, 2018; Saunders et al., 2019; Hair et al., 2022).

The target population comprised managers, economists, business analysts, data scientists, information technology professionals, and senior executives working in manufacturing, financial services, healthcare, telecommunications, and digital service organizations that have implemented data-driven decision-making systems. Using Cochran's sampling principles and recommendations for structural equation modelling (SEM), a sample size of 520 respondents was determined through stratified random sampling to ensure adequate representation across industries and organizational levels (Cochran, 1977; Krejcie & Morgan, 1970; Hair et al., 2022).

Primary data were collected using a structured questionnaire adapted from previously validated measurement scales. The instrument consisted of six sections covering demographic information, data science capability, managerial decision quality, organizational agility, innovation capability, and organizational performance. All measurement items were assessed using a five-point Likert scale ranging from 1 = Strongly Disagree to 5 = Strongly Agree. Content validity was established through expert review by academics and industry practitioners, while a pilot study involving 40 respondents was conducted to improve clarity and reliability (DeVellis & Thorpe, 2021).

The reliability and validity of the measurement model were assessed using Cronbach's alpha, Composite Reliability (CR), Average Variance Extracted (AVE), indicator loadings, the Fornell–Larcker criterion, and the Heterotrait–Monotrait (HTMT) ratio to establish internal consistency, convergent validity, and discriminant validity (Fornell & Larcker, 1981; Henseler et al., 2015; Hair et al., 2022). Common method bias was evaluated using Harman's single-factor test and full collinearity variance inflation factors (Podsakoff et al., 2003; Kock, 2015).

Data analysis was conducted using IBM SPSS Statistics 29 for descriptive statistics and SmartPLS 4 for Partial Least Squares Structural Equation Modelling (PLS-SEM). Descriptive statistics summarized respondents' characteristics, while inferential analyses examined the proposed hypotheses. Bootstrapping with 5,000 resamples was employed to estimate path coefficients, t-statistics, confidence intervals, mediation effects, effect sizes (f²), predictive relevance (Q²), coefficient of determination (R²), and model fit indices. PLS-SEM was selected because of its suitability for predictive research, complex structural models, and latent variable analysis involving multiple endogenous constructs (Hair et al., 2022; Sarstedt et al., 2022; Ringle et al., 2024). The methodological procedure provides a rigorous empirical framework for evaluating how data science capabilities influence managerial decision-making and organizational performance in the digital economy.

3.1. Research Design

This study adopted a quantitative research design grounded in the positivist research philosophy and implemented through a cross-sectional survey. The quantitative approach was considered appropriate because the study sought to empirically examine the relationships among Data Science Capability (DSC), Managerial Decision Quality (MDQ), Organizational Agility (OA), Innovation Capability (IC), and Organizational Performance (OP) using objective measurement and statistical analysis. Quantitative research facilitates hypothesis testing, measurement of latent constructs, and the generalization of findings across a defined population through rigorous statistical procedures (Creswell & Creswell, 2018; Saunders et al., 2019; Bryman, 2016).

The study was anchored on the positivist philosophy, which assumes that organizational phenomena can be objectively observed, measured, and explained using empirical evidence. Positivism emphasizes deductive reasoning, standardized data collection instruments, and statistical analysis to validate theoretical propositions and establish causal relationships among variables. This philosophical orientation supports the testing of hypotheses derived from the Resource-Based View (RBV), Dynamic Capabilities Theory (DCT), and the Technology Acceptance Model (TAM) while minimizing researcher bias and enhancing the reliability and validity of the findings (Comte, 1853; Crotty, 1998; Bell et al., 2022).

A cross-sectional survey design was employed to collect data from respondents at a single point in time across selected organizations. This design is widely used in management and information systems research because it enables efficient examination of relationships among multiple constructs and supports large-scale data collection from diverse respondents. Furthermore, the design is compatible with Partial Least Squares Structural Equation Modelling (PLS-SEM), which simultaneously evaluates measurement and structural models for predictive research involving complex latent variables (Hair et al., 2022; Sarstedt et al., 2022; Ringle et al., 2024). Consequently, the adopted research design provides a robust framework for investigating how data science capability enhances managerial decision quality, organizational agility, innovation capability, and organizational performance.

3.2. Target Population

The target population for this study comprises senior professionals and decision-makers who are directly involved in organizational strategy, data-driven decision-making, digital transformation, and business analytics. These respondents possess the technical expertise and managerial experience required to provide reliable information on the adoption and impact of data science capabilities within their organizations. The population spans organizations across manufacturing, financial services, healthcare, telecommunications, public administration, and other knowledge-intensive sectors where data science plays a critical role in managerial and economic decision-making.

Table 1 presents the target population of the study, comprising 3,800 senior professionals drawn from manufacturing, financial services, healthcare, telecommunications, public sector institutions, ICT firms, digital enterprises, and research organizations. The respondents include managers, Chief Executive Officers (CEOs), economists, business analysts, data scientists, and Information Technology (IT) managers because they are directly involved in strategic decision-making, digital transformation, and organizational performance management. Their diverse professional backgrounds ensure comprehensive perspectives on the adoption and utilization of data science capabilities across different sectors of the economy. Selecting respondents from multiple industries enhances the representativeness of the sample and improves the generalizability of the findings. The target population therefore provides an appropriate empirical basis for examining how data science capability influences managerial decision quality, organizational agility, innovation capability, and organizational performance (Saunders et al., 2019; Creswell & Creswell, 2018; Hair et al., 2022).

3.3. Sampling Technique and Sample Size

The study employed stratified random sampling to select respondents from the target population of 3,800 professionals comprising managers, Chief Executive Officers (CEOs), economists, business analysts, data scientists, and Information Technology (IT) managers across various industries. Stratified random sampling was considered the most appropriate technique because the target population is heterogeneous, consisting of distinct professional groups with varying roles in organizational decision-making and data science implementation. By dividing the population into homogeneous strata based on professional categories and randomly selecting respondents from each stratum, the technique ensures that every subgroup is adequately represented in the final sample. This approach minimizes sampling bias, improves precision, and enhances the representativeness of the findings (Cochran, 1977; Saunders et al., 2019).

A total sample size of 520 respondents was determined from the population of 3,800 using established sample size determination guidelines proposed by Krejcie and Morgan (1970) and Cochran (1977). The sample size also satisfies the recommendations for Partial Least Squares Structural Equation Modelling (PLS-SEM), which requires an adequately large sample to produce reliable parameter estimates, improve statistical power, and enhance the predictive accuracy of complex structural models (Hair et al., 2022; Sarstedt et al., 2022). A sample of 520 respondents exceeds the minimum threshold required for multivariate analysis, thereby increasing the robustness and generalizability of the study's findings.

The proportional allocation of respondents across the identified strata ensured that each professional category contributed to the study according to its relative size within the population. Following stratification, respondents within each category were selected through simple random sampling, giving every eligible participant an equal probability of selection. This procedure reduced selection bias and strengthened the external validity of the research by ensuring that the sample accurately reflected the characteristics of the broader population. Consequently, the use of stratified random sampling provides a scientifically rigorous basis for examining the influence of data science capability on managerial decision quality, organizational agility, innovation capability, and organizational performance across different organizational contexts.

3.4. Sample Size Determination

The sample size for this study was determined using Cochran's (1977) sample size determination formula for finite populations. Cochran's formula is appropriate because it provides a statistically reliable sample that adequately represents the target population while minimizing sampling error.

The initial sample size for an infinite population is calculated as:

$$n_0={\displaystyle \frac{Z^{2}pq}{e^2}}$$

Where:

• n₀ = Initial sample size
• Z = Standard normal variate at 95% confidence level (1.96)
• p = Estimated proportion of the population possessing the characteristic (0.50)
• q = 1 – p = 0.50
• e = Margin of error (0.05)

Substituting the values:

$$n_0={\displaystyle \frac{{\left(1.96\right)}^2\left(0.50\right)\left(0.50\right)}{{\left(0.05\right)}^2}}$$

$$n_0={\displaystyle \frac{3.8416x0.25}{0.0025}}$$

$$n_0={\displaystyle \frac{0.9604}{0.0025}}=384.16$$

Since the study population is finite (N = 3,800), the finite population correction (FPC) formula was applied:

$$n_{=}{\displaystyle \frac{n_0}{1+\frac{n_0-1}{N}}}$$

Where:

• N = 3,800
• n₀ = 384.16

Substituting the values:

$$n_{=}{\displaystyle \frac{384.16}{1+\frac{384.16-1}{3800}}}$$

$$n_{=}{\displaystyle \frac{384.16}{1+0.1008}}$$

$$n_{=}{\displaystyle \frac{384.16}{1.1008}}$$

$$n=349$$

Therefore, the minimum statistically required sample size is 349 respondents.

However, to improve the statistical power of the analysis, account for possible non-response, incomplete questionnaires, and satisfy the recommended sample size requirements for Partial Least Squares Structural Equation Modelling (PLS-SEM) involving multiple latent variables and complex structural paths, the sample size was increased to 520 respondents. An enlarged sample enhances estimation accuracy, reduces sampling error, improves model stability, and increases the generalizability of the findings (Hair et al., 2022; Sarstedt et al., 2022; Kline, 2023).

Table 2 summarizes the procedure used to determine the sample size for the study. The target population consisted of 3,800 respondents, including managers, CEOs, economists, business analysts, data scientists, and IT managers from selected organizations. Using Cochran's (1977) sample size determination formula at a 95% confidence level, 5% margin of error, and a population proportion of 0.50, the minimum corrected sample size for the finite population was calculated as 349 respondents. However, the final sample was increased to 520 respondents to improve statistical power, compensate for possible non-response and incomplete questionnaires, and satisfy the minimum sample requirements for Partial Least Squares Structural Equation Modelling (PLS-SEM). The larger sample enhances the precision, reliability, robustness, and generalizability of the study findings, while reducing sampling error and increasing confidence in the empirical results (Hair et al., 2022; Sarstedt et al., 2022).

3.5. Research Instrument

The primary instrument for data collection was a structured questionnaire designed to measure the study constructs, namely Data Science Capability, Managerial Decision Quality, Organizational Agility, Innovation Capability, and Organizational Performance. The questionnaire was developed by adapting validated measurement items from previous empirical studies and modifying them to align with the objectives of this research. The instrument comprised two sections: Section A collected respondents' demographic information, while Section B measured the study variables using multiple-item scales.

A 5-point Likert scale was employed to capture respondents' perceptions of each measurement item. The response categories were coded as 1 = Strongly Disagree, 2 = Disagree, 3 = Neutral, 4 = Agree, and 5 = Strongly Agree. The Likert scale was selected because it is simple to administer, facilitates quantitative analysis, and provides reliable measurement of attitudes, perceptions, and organizational practices. It is also widely recommended for behavioural and management research because it improves response consistency and supports advanced multivariate statistical analyses such as Structural Equation Modelling (PLS-SEM) (Likert, 1932; Hair et al., 2022; Saunders et al., 2019).

Before the main survey, the questionnaire was subjected to expert review and a pilot study to establish its content validity, clarity, and reliability. The final instrument demonstrated satisfactory psychometric properties based on Cronbach's Alpha, Composite Reliability (CR), Average Variance Extracted (AVE), and the Heterotrait–Monotrait (HTMT) ratio, indicating that it was appropriate for measuring the latent constructs included in the study.

Table 3 presents the operationalization of the study variables, specifying the constructs, measurement indicators, scales, and sources adopted for empirical assessment. The independent variable, Data Science Capability (DSC), is measured using five indicators that assess big data infrastructure, artificial intelligence adoption, machine learning utilization, employees' analytical competence, and a data-driven organizational culture. The mediating variable, Organizational Agility (OA), comprises four indicators evaluating organizational responsiveness, flexibility, adaptability, and customer orientation. The dependent variable, Organizational Performance (OP), is operationalized using five dimensions: financial, operational, innovation, market, and environmental sustainability performance. Additionally, firm size, industry sector, organizational age, respondents' experience, and digital maturity are included as control variables to minimize extraneous influences on the hypothesized relationships. All latent constructs were measured using a five-point Likert scale, ensuring consistency and comparability across responses. The measurement items were adapted from well-established and validated instruments in previous studies, thereby strengthening the content validity and reliability of the research instrument (Gupta & George, 2016; Kaplan & Norton, 1996; Richard et al., 2009; Tallon & Pinsonneault, 2011; Teece, 2007).

Table 4 presents the statistical criteria used to evaluate the reliability and validity of the measurement model before testing the structural relationships. Cronbach's Alpha (α) and Composite Reliability (CR) assess the internal consistency of the measurement scales, with values of 0.70 or higher indicating that the questionnaire items consistently measure their respective constructs. Average Variance Extracted (AVE) evaluates convergent validity by determining the proportion of variance captured by a construct relative to measurement error; values of 0.50 or above indicate satisfactory convergent validity. Outer factor loadings of 0.70 or greater demonstrate that individual indicators significantly contribute to their latent constructs. Finally, the Heterotrait–Monotrait (HTMT) ratio, with values below 0.85 (strict criterion) or 0.90 (liberal criterion), confirms discriminant validity by ensuring that each construct is empirically distinct from the others. Collectively, these statistical thresholds establish that the measurement model is reliable, valid, and suitable for subsequent Structural Equation Modelling (PLS-SEM) analysis.

Table 5 summarizes the statistical techniques employed to address each research objective and test the proposed hypotheses. The analytical procedures include descriptive statistics, Pearson Product Moment Correlation, reliability and validity assessments, Partial Least Squares Structural Equation Modelling (PLS-SEM), bootstrapping, Multi-Group Analysis (MGA), and Importance–Performance Map Analysis (IPMA). These techniques were selected in accordance with established guidelines for multivariate data analysis and PLS-SEM (Hair et al., 2022; Henseler et al., 2015; Ringle & Sarstedt, 2016).

Table 6 presents the gender distribution of the 520 respondents who participated in the study. The results indicate that 298 respondents (57.3%) were male, while 222 respondents (42.7%) were female. Statistically, the distribution shows a 14.6 percentage-point difference between male and female respondents, indicating a moderate predominance of male participants in the sample. Despite this difference, both genders are adequately represented, suggesting that the sample reflects diverse perspectives regarding data science capability, managerial decision-making, organizational agility, and organizational performance. The inclusion of respondents from both genders enhances the representativeness of the dataset and minimizes gender-related sampling bias. Consequently, the demographic composition provides a reliable foundation for subsequent descriptive, correlational, and Structural Equation Modelling (PLS-SEM) analyses, thereby strengthening the validity and generalizability of the study findings (Hair et al., 2022; Saunders et al., 2019; Creswell & Creswell, 2018).

Table 7 presents the descriptive statistics of the study variables. Data Science Capability recorded the highest mean score (M = 4.28, SD = 0.62), indicating a high level of data science adoption among the sampled organizations. Organizational Performance also recorded a high mean (M = 4.25, SD = 0.67), followed by Decision Quality (M = 4.17, SD = 0.69) and Innovation Capability (M = 4.10, SD = 0.74). The relatively low standard deviations (0.62–0.74) indicate limited variability in respondents' opinions, suggesting a high degree of consistency in their responses. Overall, the findings imply positive perceptions of the study constructs and provide a reliable basis for subsequent inferential analyses (Field, 2018; Hair et al., 2022).

Table 8 presents the reliability and convergent validity results for the study constructs. All constructs recorded Cronbach's Alpha values ranging from 0.90 to 0.92 and Composite Reliability (CR) values between 0.93 and 0.94, exceeding the recommended threshold of 0.70, thereby confirming excellent internal consistency. Similarly, the Average Variance Extracted (AVE) values ranged from 0.72 to 0.76, surpassing the minimum acceptable value of 0.50, which indicates adequate convergent validity. These findings demonstrate that the measurement items consistently represent their respective constructs and possess satisfactory reliability and validity for subsequent structural equation modelling analysis (Fornell & Larcker, 1981; Hair et al., 2022).

Table 9 presents the structural model results obtained through Partial Least Squares Structural Equation Modelling (PLS-SEM). All six hypotheses were statistically significant, with positive path coefficients (β = 0.29–0.71), high t-statistics (7.35–16.44), and p-values of 0.000 (i.e., p < 0.001), indicating strong empirical support for the proposed relationships. Among the paths, H1 recorded the strongest effect (β = 0.71), while H6 showed the weakest but still significant effect (β = 0.29). The R² values reveal that the model explains 61% of the variance in Decision Quality, 57% in Organizational Agility, 49% in Innovation Capability, and 68% in Organizational Performance. These results indicate moderate to substantial predictive power, demonstrating that the proposed model effectively explains organizational performance through data science capability and related mediating constructs (Hair et al., 2022; Sarstedt et al., 2022).

Coefficient of Determination (R²)

Endogenous Construct	R² Value	Interpretation
Decision Quality	0.61	Moderate to substantial explanatory power
Organizational Agility	0.57	Moderate explanatory power
Innovation Capability	0.49	Moderate explanatory power
Organizational Performance	0.68	Substantial explanatory power

Source: Researcher's field survey (2026). Structural model estimated using SmartPLS 4 with bootstrapping (5,000 resamples) following the guidelines of Hair et al. (2022), Sarstedt et al. (2022), and Ringle et al. (2024).

5. Discussion of Findings

The findings of this study demonstrate that data science capability is a critical organizational resource that significantly enhances managerial decision quality, organizational agility, innovation capability, and overall organizational performance across the organizations under review, namely manufacturing firms, financial institutions, healthcare organizations, telecommunications companies, public sector institutions, and digital service enterprises. The structural model revealed that all hypothesized relationships were positive and statistically significant (p < 0.001), indicating that organizations investing in big data infrastructure, artificial intelligence, machine learning, and advanced analytics are more likely to achieve superior strategic and operational outcomes. The strongest relationship (H1: β = 0.71) suggests that data science capability substantially improves managerial decision quality by enabling executives and analysts to make timely, evidence-based, and predictive decisions. This finding corroborates the studies of Gupta and George (2016), Mikalef et al. (2020), and Davenport and Harris (2007), who identified analytics capability as a strategic organizational asset.

The findings further reveal that managers and decision-makers within the manufacturing, banking, healthcare, telecommunications, public administration, and digital services sectors leverage data science to improve organizational agility and innovation capability. Organizations that effectively transform data into actionable intelligence are better positioned to anticipate market changes, optimize operational processes, and respond rapidly to customer demands. These findings support the propositions of Dynamic Capabilities Theory (Teece et al., 1997), particularly the organizational abilities to sense opportunities, seize strategic initiatives, and transform internal capabilities. The moderate-to-high explanatory power of the model (R² = 0.49–0.68) indicates that data science capability explains a substantial proportion of the variance in organizational agility, innovation capability, and performance across these sectors.

The study also established that innovation capability significantly enhances organizational performance, supporting the Resource-Based View (RBV), which argues that valuable, rare, inimitable, and well-organized resources generate sustainable competitive advantage (Barney, 1991). Specifically, the organizations under review reported improvements in financial performance, operational efficiency, service quality, customer satisfaction, and market competitiveness through the adoption of data-driven technologies. These findings are consistent with those of Wamba et al. (2020), Fosso Wamba et al. (2021), and Bharadwaj et al. (2013), who found that analytics capabilities contribute significantly to organizational success.

Furthermore, the findings support the Technology Acceptance Model (TAM) by indicating that managers, economists, business analysts, data scientists, CEOs, and IT managers across the sampled organizations perceive data science technologies as both useful and effective for strategic planning and operational decision-making. Overall, the study demonstrates that organizations operating in manufacturing, financial services, healthcare, telecommunications, the public sector, and digital enterprises can substantially improve their competitiveness by strengthening data science capabilities. The findings provide empirical evidence that investments in analytics, artificial intelligence, and digital transformation are essential drivers of sustainable organizational performance in the era of the digital economy and Industry 4.0.

6. Practical Implications

The findings of this study provide important practical implications for managers, business leaders, policymakers, technology professionals, and organizational stakeholders seeking to harness data science for improved economic and managerial outcomes. The empirical evidence demonstrates that data science capability is a strategic organizational resource that significantly enhances managerial decision quality, organizational agility, innovation capability, and overall organizational performance. Consequently, organizations should prioritize investments in data-driven technologies, analytical infrastructure, and human capital to strengthen their competitive position in the digital economy.

For business organizations, particularly those in the manufacturing, financial services, healthcare, telecommunications, public sector, and digital services sectors, the study highlights the importance of integrating artificial intelligence, machine learning, predictive analytics, and business intelligence into strategic planning and operational processes. Managers should establish enterprise-wide data governance frameworks that ensure data quality, accessibility, security, and ethical use. Investment in cloud computing, advanced analytics platforms, and real-time dashboards will enable executives to make faster and more accurate decisions based on reliable evidence rather than intuition alone.

The study also underscores the need for organizational leaders to develop a strong data-driven culture. This involves encouraging employees at all organizational levels to utilize analytical insights in decision-making while promoting continuous learning in digital competencies. Organizations should invest in regular training programmes, professional certifications, and capacity-building initiatives in data analytics, artificial intelligence, and digital technologies. Strengthening employees' analytical capabilities will improve organizational agility, foster innovation, and enhance responsiveness to dynamic market conditions.

For policymakers and government institutions, the findings suggest that national digital transformation strategies should emphasize investments in digital infrastructure, open data initiatives, artificial intelligence research, and data governance policies. Governments should establish regulatory frameworks that promote responsible data sharing, cybersecurity, privacy protection, and ethical AI adoption while encouraging collaboration between universities, research institutions, and industry. Public investment in digital infrastructure and innovation ecosystems will facilitate broader adoption of data science across both public and private sectors, thereby improving economic productivity and evidence-based policymaking.

The findings further indicate that higher education institutions and professional training organizations should revise academic curricula to incorporate data science, business analytics, artificial intelligence, machine learning, and digital management into economics, business administration, management, engineering, and information technology programmes. Such curriculum reforms will equip future managers and economists with the interdisciplinary competencies required to operate effectively in data-intensive organizational environments.

Finally, organizations should adopt performance measurement systems that continuously evaluate the effectiveness of data science initiatives using key performance indicators relating to decision quality, innovation, operational efficiency, customer satisfaction, and financial performance. Regular monitoring and evaluation will enable organizations to identify capability gaps, optimize resource allocation, and maximize the return on investments in digital technologies. Collectively, these practical implications provide a roadmap for organizations seeking to leverage data science as a strategic capability for sustainable growth, innovation, and competitive advantage in the era of Industry 4.0 and the digital economy.

7. Theoretical Contributions

This study makes several significant theoretical contributions to the literature on data science, economics, and strategic management. First, it extends the Resource-Based View (RBV) by demonstrating that data science capability constitutes a strategic organizational resource that satisfies the VRIO (valuable, rare, inimitable, and organized) criteria for sustainable competitive advantage. The findings provide empirical evidence that investments in big data infrastructure, artificial intelligence, machine learning, and analytics competencies generate superior organizational performance through enhanced decision quality, innovation, and agility, thereby broadening the application of RBV in digital transformation research (Barney, 1991; Wernerfelt, 1984; Gupta & George, 2016).

Second, the study validates Dynamic Capabilities Theory by confirming that organizations possessing strong data science capabilities are better able to sense environmental changes, seize emerging opportunities, and transform organizational resources in response to technological disruption and market uncertainty. These findings reinforce the propositions of Teece et al. (1997) and Teece (2007) regarding the strategic importance of adaptive capabilities in sustaining organizational competitiveness.

Third, the research integrates economics and management by providing a unified empirical framework that explains how data science enhances both economic decision-making and managerial effectiveness. This interdisciplinary perspective bridges a notable gap between economic forecasting, organizational strategy, and digital innovation, supporting calls for greater integration of analytical sciences in management research (Athey, 2018; Brynjolfsson & McAfee, 2014).

Finally, the study develops and empirically validates a new conceptual model linking Data Science Capability → Managerial Decision Quality → Organizational Agility → Innovation Capability → Organizational Performance. This model advances the literature by explaining the mechanisms through which digital capabilities create competitive advantage in the context of Industry 4.0 and the digital economy. Consequently, the study offers a robust theoretical foundation for future empirical research examining digital transformation, analytics capability, and sustainable organizational performance (Mikalef et al., 2020; Wamba et al., 2020; Verhoef et al., 2021).

8. Limitations of the Study and Future Research Directions

Despite its theoretical and practical contributions, this study has several limitations that should be considered when interpreting the findings. First, the study adopted a cross-sectional research design, whereby data were collected at a single point in time. Although this design is appropriate for examining relationships among variables, it does not permit definitive conclusions regarding causality or the evolution of organizational capabilities over time. Future studies should employ longitudinal research designs to investigate how data science capability, organizational agility, innovation capability, and organizational performance evolve as organizations progress through different stages of digital transformation (Creswell & Creswell, 2018; Saunders et al., 2019).

Second, the study relied on self-reported data obtained from managers, Chief Executive Officers (CEOs), economists, business analysts, data scientists, and Information Technology (IT) managers. Although these respondents possess substantial knowledge of organizational practices, self-reported measures may be susceptible to social desirability bias, recall bias, and respondents' subjective perceptions. Future research should combine survey data with objective organizational performance indicators, archival records, financial reports, and enterprise analytics data to improve measurement accuracy and reduce respondent bias (Podsakoff et al., 2003).

Third, the investigation was conducted within a single-country context, which may limit the generalizability of the findings to organizations operating under different institutional, cultural, technological, and regulatory environments. Differences in digital infrastructure, national innovation systems, and economic development may influence the adoption and effectiveness of data science capabilities. Future studies should conduct cross-country comparative analyses involving both developed and emerging economies to validate and extend the proposed conceptual model across diverse contexts (Hofstede, 2001; World Bank, 2021).

Furthermore, although procedural and statistical measures were implemented to minimize common method bias, including the use of validated measurement scales and appropriate statistical diagnostics, the possibility of residual common method variance cannot be entirely eliminated because the data were collected using a single survey instrument. Future researchers should adopt multi-source data collection, obtaining responses from different organizational stakeholders and integrating qualitative methods such as interviews or case studies to enhance methodological rigor (Podsakoff et al., 2003; Kock, 2015).

Beyond these limitations, several avenues for future research emerge. Scholars may examine the mediating and moderating effects of organizational culture, digital leadership, knowledge management capability, cybersecurity readiness, organizational learning, and artificial intelligence maturity on the relationship between data science capability and organizational performance. Future studies could also compare public and private sector organizations, small and medium-sized enterprises (SMEs) versus large corporations, or investigate sector-specific applications in manufacturing, financial services, healthcare, telecommunications, education, and government institutions. Additionally, the use of advanced analytical techniques, such as covariance-based structural equation modelling (CB-SEM), multilevel modelling, longitudinal SEM, machine learning algorithms, and hybrid quantitative–qualitative approaches, would provide deeper insights into the complex mechanisms through which data science influences economic and managerial outcomes. Such investigations would further enrich the theoretical and empirical understanding of data science as a strategic capability in the era of Industry 4.0 and the digital economy.

9. Conclusions

This study investigated the role of data science capability in enhancing managerial decision quality, organizational agility, innovation capability, and organizational performance within the context of the digital economy and Industry 4.0. By integrating perspectives from economics, strategic management, and information systems, the study developed and empirically validated a comprehensive framework that explains how data-driven capabilities contribute to sustainable organizational success. The findings demonstrate that organizations investing in big data infrastructure, artificial intelligence, machine learning, predictive analytics, and digital competencies are better positioned to improve decision-making processes, respond swiftly to environmental changes, stimulate innovation, and achieve superior organizational performance.

The empirical results confirmed that all the proposed hypotheses were statistically significant, indicating that data science capability positively influences managerial decision quality, which subsequently enhances organizational agility, innovation capability, and overall organizational performance. The substantial explanatory power of the structural model further demonstrates that data science is no longer merely a technological resource but a strategic organizational capability that drives value creation and competitive advantage. These findings reinforce the propositions of the Resource-Based View (RBV) by showing that data science capabilities represent valuable, rare, inimitable, and organizationally embedded resources. Likewise, the results validate Dynamic Capabilities Theory, illustrating that organizations capable of sensing market opportunities, seizing technological innovations, and transforming internal processes are more resilient and competitive in dynamic business environments. The findings also support the Technology Acceptance Model (TAM) by highlighting the importance of perceived usefulness and effective utilization of digital technologies in improving managerial outcomes.

From an organizational perspective, the study demonstrates that data science has become an indispensable strategic asset across manufacturing firms, financial institutions, healthcare organizations, telecommunications companies, public sector institutions, and digital service enterprises. Organizations that effectively integrate analytics into strategic planning, operational management, and innovation processes are more likely to enhance productivity, optimize resource utilization, improve customer satisfaction, strengthen financial performance, and sustain long-term competitiveness. Consequently, investments in digital infrastructure, analytical talent, cloud computing, artificial intelligence, and organizational learning should be regarded as strategic priorities rather than discretionary technological expenditures.

The study also contributes to academic knowledge by extending the Resource-Based View, validating Dynamic Capabilities Theory in a digital context, integrating economics with management research, and proposing a novel empirical model linking Data Science Capability, Managerial Decision Quality, Organizational Agility, Innovation Capability, and Organizational Performance. This integrated framework provides a robust foundation for future empirical investigations into digital transformation, analytics capability, and organizational competitiveness.

The increasing complexity of global markets, rapid technological advancement, and the proliferation of data-intensive business environments require organizations to embrace data science as a core strategic capability. Firms that successfully cultivate analytical capabilities, foster data-driven cultures, and integrate intelligent technologies into managerial decision-making will be better equipped to navigate uncertainty, exploit emerging opportunities, and sustain competitive advantage. Therefore, the study concludes that economics and management are increasingly empowered by data science, making analytical capability a decisive factor in achieving innovation, organizational excellence, and sustainable economic development in the twenty-first century.