아주대학교

목차

1. Introduction 1
1.1. Research Background and Rationale 1
1.2. Research Objectives 2
1.3. Research Questions 4
1.4. Scope of the Study and Dissertation Structure 6
1.4.1. Scope of the Study 6
1.4.2. Dissertation Structure 7
2. Literature Review 9
2.1. Digital Transformation and SMEs 9
2.1.1. Definition and Evolutionary Spectrum of Digital Transformation 9
2.1.2. Strategic Opportunities and Structural Barriers in SME DX 12
2.1.3. The IT Productivity Paradox and the DX Efficiency Paradox 14
2.2. Technology-Organization-Environment (TOE) Framework 16
2.2.1. Theoretical Origins and Structural Dimensions 16
2.2.2. TOE Framework in SME DX Contexts 17
2.3. Resource-Based View (RBV) 18
2.3.1. Theoretical Foundations of the RBV 18
2.3.2. Information Technology Resources and RBV 19
2.3.3. Theoretical Alignment of RBV with Data Envelopment Analysis (DEA) 20
2.4. Dynamic Capability Theory (DCT) 23
2.4.1. Theoretical Origins and Core Concepts of the DCT 23
2.4.2. Dynamic Capabilities in the Context of DX 24
2.4.3. Absorptive Capacity and the Role of DX Training 26
2.5. A ‘TOE–RBV–DCT’ Integrated Theoretical Framework 28
2.5.1. Theoretical Complementarity among TOE, RBV, and DCT 28
2.5.2. Sequential TOE–RBV–DCT Integrated Analytical Framework 29
2.5.3. Academic Position and Contribution as an Integrated Research 31
2.6. Prior Research on Digital Maturity Models and Typologies 33
2.6.1. Evolution of Digital Maturity Models 33
2.6.2. PCA & K-Means-Based Firm Typology Research 33
2.6.3. Policy Implications of Digital Maturity Typologies 35
2.7. Prior Research on Data Envelopment Analysis (DEA) 36
2.7.1. Theoretical Origins and Fundamental Models of DEA 36
2.7.2. The SBM-DEA Model 37
2.7.3. Prior DEA Research on IT and Digital Transformation Efficiency 39
2.8. Prior Research on DEA-Tobit Integrated Analysis 40
2.8.1. The Need for Two-Stage DEA and Tobit Regression 40
2.8.2. Review of Major DEA–Tobit Integrated Studies 42
2.8.3. DEA Homogeneity and Cluster-Specific Analysis in Prior Research 43
2.9. Limitations of Prior Research and Research Gaps 45
2.9.1. Limitations of Existing DX Typology Research in SMEs 45
2.9.2. Limitations of Existing DEA-Based IT Efficiency Research 47
2.9.3. Addressing Research Gaps and Contributions 49
3. Research Design and Methodology 52
3.1. Integrated Research Framework 52
3.1.1. Integration and Rationale of the Two Studies 52
3.1.2. Four-Stages Sequential Integrated Analytical Pipeline 53
3.2. Data and Sample 55
3.2.1. Overview of the SME Informatization Level Survey 55
3.2.2. Characteristics of the 2025 Survey Dataset 55
3.3. Variable Selection and Operationalization 57
3.3.1. TOE-Based Input Variables for PCA and Cluster Analyses 59
3.3.2. RBV-Based DEA Input and Output Variables 61
3.3.3. DCT-Based Variables for Tobit Regression 62
3.4. Research Hypotheses 64
3.4.1. Hypotheses Related to DX Maturity Cluster Profile (Study 1) 64
3.4.2. Hypotheses Related to Cluster-Specific DX Efficiency & Determinants (Study 2) 65
4. Empirical Result 67
4.1. Descriptive Statistics 67
4.2. PCA Result and Interpretation of Principal Components 68
4.2.1. PCA Results 68
4.2.2. Components Interpretation 72
4.3. K-Means Clustering Results 76
4.3.1. Determination of the Optimal Number of Clusters 76
4.3.2. Definition of DX Maturity Groups Based on PCA Scores 78
4.4. K-Means Clustering and Profiling Results 80
4.4.1. Comparison of Key Variables Across Clusters 80
4.4.2. Chi-Square and ANOVA Results and Hypothesis Verification 80
4.5. MNLR Analysis and Moderation Effects 85
4.5.1. Model Fit 86
4.5.2. Main Effects 86
4.5.3. Moderation Effects 88
4.5.4. Subgroup Analysis: Manufacturing vs. Non-Manufacturing 89
4.6. DEA Efficiency Analysis Results 91
4.6.1. Overall Sample Efficiency Analysis 91
4.6.2. Cluster-Specific SBM-DEA Efficiency Comparison 92
4.6.3. Decomposition Analysis of Sources of Inefficiency 94
4.7. Tobit Regression Analysis Results 97
4.7.1. Model Adequacy and Preliminary Assessment 97
4.7.2. Cluster-Specific Tobit Regression Results and Key Findings 98
4.7.3. Robustness Assessment (Integrated Across the Three Clusters) 104
4.8. Summary of Hypothesis Testing Result and Findings 106
4.8.1. Findings and Hypothesis Testing Result (Study 1) 106
4.8.2. Findings and Hypothesis Testing Result (Study 2) 107
4.9. Integrated Insights Derived from the Comparative Analysis 109
5. Discussion 112
5.1. Overview of Integrated Findings Across a Two-Study Framework 112
5.2. DX Maturity Segmentation and Cluster Profiling Results (Study 1) 113
5.2.1. The Role of Firm Resource Scale (H1) 113
5.2.2. DX Education as a Universal Catalyst (H2, H3)) 114
5.2.3. Industry Structure as a Boundary Condition (H4, H6) 114
5.2.4. Firm Age and Organizational Maturity (H5) 115
5.3. Cluster-Specific DX Investment Efficiency (Study 2) 116
5.3.1. Pervasive DX Investment Inefficiency Across the SME Population (H7) 116
5.3.2. Cluster-Specific Heterogeneity of Efficiency Determinants (H8) 117
5.3.3. VRS Efficiency Hierarchy and the Universal Role of Revenue Scale (H9, H12) 119
5.3.4. Heterogeneous Effects of Internal & External Data Management (H9, H12) 121
5.3.5. The Maturity-Contingent Effect of Cloud and Collaboration Solutions (H11) 123
5.4. Synthesis of Integrated Findings 125
5.4.1. The Aspiration-Execution Gap as the Structural Challenge 125
5.4.2. The DX Efficiency Paradox and Maturity-Efficiency Decoupling 126
5.5. Theoretical Contributions 128
5.6. Practical and Policy Implications 132
5.7. Limitations 134
6. Conclusions 136
6.1. Principal Findings 136
Study 1: DX Maturity Typology 136
Study 2: Cluster-Specific DX Investment Efficiency 137
6.2. Concluding Theoretical Contributions 138
6.3. Concluding Practical and Policy Implications 139
6.4. Final Message 140
References 142
Appendix 149