HUMAN CAPITAL AS A DRIVER OF VALUE CREATION AND COMPETITIVE ADVANTAGE: EMPIRICAL EVIDENCE FROM SERBIAN SMEs
Description
1. Overview and Data Gathering This dataset contains anonymized primary empirical data collected in May 2026 from 30 private small and medium-sized enterprises (SMEs) in the Republic of Serbia. Data was gathered via an online survey containing 38 closed-ended items divided into three segments: firm demographics, adequacy of intellectual capital (IC) management, and performance variables. 2. Research Hypotheses Based on the Resource-Based View (RBV), the study examines how human capital drives organizational success (value creation and competitive advantage) in a transitional economy: * Main Hypothesis (H1): Aggregated human capital has a positive, significant impact on overall organizational performance. * Sub-Hypotheses (H1a, H1b, H1c): Employee skills (HCSkills), training investments (HCTraining), and workplace creativity (HCCreativity) each positively affect performance. 3. Key Findings * Managerial Paradox: 93.3% of managers recognize IC's strategic importance, and 100% claim management knowledge. However, 60.0% lack performance measurement systems to audit these resources. * Macro-Level Impact (H1 Supported): Simple linear regression proves that the aggregated human capital mean score accounts for 38.1% of the performance variance (F(1, 28) = 17.267, p < .001). * The Power of Training (H1b Supported): Multiple regression shows that the three sub-dimensions together explain 56.5% of the total variance (F(3, 26) = 11.272, p < .001). Continuous training investments (HCTraining) is the primary and only statistically significant driver (β = .664, p < .001). Raw expertise (p = .259) and creativity (p = .093) yield non-significant isolated effects. 4. Data Layout and Reuse * Independent Variables: HCSkills, HCTraining, and HCCreativity (3-point intensity scale: 1 = small, 3 = large extent). Their arithmetic mean yields AggregatedHC. * Dependent Variables: 23 metrics across 7 performance fields captured using tailored 5-point scales. Their arithmetic mean yields the composite success metric AggregatedOP. * Usage: Researchers can open the raw .xlsx or .csv file in IBM SPSS Statistics, R, or Python to replicate the baseline OLS diagnostics (Durbin-Watson, VIF) and two-step regression models.
Files
Steps to reproduce
The data was collected in May 2026 via an online survey distributed to a convenience sample of 30 private sector small and medium-sized enterprises (SMEs) in the Republic of Serbia. The measurement instrument comprised 38 closed-ended items tracking firm demographics, intellectual capital management, and organizational performance fields. Statistical analysis and econometric modeling were executed within the IBM SPSS Statistics software environment. Detailed step-by-step procedures for variable aggregation, Cronbach's alpha reliability checks, and two-phase OLS regression models (simple and multiple linear regressions) are explicitly detailed in the 'ReadMe_Instructions' sheet embedded directly inside the attached Excel workbook.
Institutions
- Union UniversityCentral Serbia, Belgrade