Digital Transformation Leadership and Employees’ Innovative Behavior : A Mediating Model Based on Self-Determination Theory

국문 초록 (Abstract)

디지털 기술이 조직 운영의 모든 측면에 빠르게 융합되면서, 디지털 전환은 글로벌 비즈니스 환경을 재편하는 되돌릴 수 없는 흐름이 되었다(World Economic Forum, 2024). 디지털 전환 리더십(Digital Transformation Leadership, DTL)은 디지털 시대의 독특한 리더십 패러다임으로서 변혁적 리더십의 핵심 차원(도덕적 모범, 비전 제시와 동기부여, 개별적 배려, 카리스마적 리더십)과 디지털 역량 지향(디지털 자원 배분, 디지털 역량 훈련, 데이터 기반 의사결정 지도)을 통합한다. 이는 조직의 디지털 혁신과 지속가능한 경쟁우위의 미시적 토대인 직원 혁신행동(Employees’ Innovative Behavior, EIB)의 중요한 선행요인으로 부상하였다. 자기결정성이론(Self-Determination Theory, SDT)에 기반하여 본 연구는 DTL이 EIB에 영향을 미치는 내재적 메커니즘을 규명하기 위해 다중 매개모형을 구축하였으며, 자율성·유능감·관계성 욕구를 포함하는 기본 심리적 욕구 충족(Basic Psychological Needs Satisfaction, BPNS)을 매개변수로 설정하였다.
본 연구는 디지털 전환의 핵심 3대 산업(제조업, IT 산업, 서비스업)에 종사하는 직원 582명을 대상으로 5점 리커트 척도를 활용한 횡단면 설문조사를 실시하였다. 데이터는 온라인 플랫폼(원쥐안싱, 링크드인)과 오프라인 기업 협력 방식의 결합을 통해 수집되었으며, 유효 응답률은 83.1%였다. 연구의 엄밀성을 확보하기 위해 다음과 같은 방법론적 전략을 적용하였다. 공통방법편의(Common Method Variance, CMV) 통제를 위해 Harman 단일요인 검정과 측정되지 않은 잠재요인(unmeasured latent factor) 방법을 사용하였고, Cronbach’s α, 복합신뢰도(CR), 평균분산추출(AVE), 확인적 요인분석(CFA)을 통해 신뢰도와 타당도를 검증하였다. 매개효과는 PROCESS 매크로를 이용해 Bootstrap 방법(5,000회 표본추출)으로 검정하였으며, 산업 간 이질성은 Mplus 8.3을 활용한 다집단 구조방정식모형(SEM) 분석으로 확인하였다.
분석 결과는 다음과 같다. (1) DTL은 EIB에 유의미한 정(+)의 직접 효과를 보였으며(β=0.342, p<0.001), EIB 변량의 60.3%를 설명하였다. (2) BPNS의 세 차원(자율성, 유능감, 관계성)은 DTL과 EIB 간 관계에서 모두 유의미한 부분 매개 역할을 수행하였다. 매개효과 값은 각각 0.135(전체 효과의 23.8%), 0.158(전체 효과의 27.9%), 0.112(전체 효과의 19.8%)로 나타났으며, 유능감 욕구가 가장 강한 매개요인이고 그다음이 자율성 욕구, 관계성 욕구 순이었다. (3) BPNS의 매개효과는 산업별로 유의미한 이질성을 보였다. 제조업에서는 유능감 욕구가 가장 핵심적인 매개요인(effect=0.165)이었고, IT 산업에서는 자율성 욕구가 가장 큰 영향력을 보였으며(effect=0.152), 서비스업에서는 관계성 욕구가 두 번째로 높게 나타났는데(effect=0.136), 이는 유능감 욕구(effect=0.142)보다 약간 낮은 수준이었다. (4) 추가 분석 결과, BPNS의 세 차원 간 유의미한 연쇄(체인) 매개효과는 확인되지 않았으나, 유능감 욕구와 관계성 욕구 간 상호작용이 자율성 욕구의 매개 경로를 정(+)적으로 조절하는 것으로 나타났다.
본 연구는 다음과 같은 이론적 기여를 제공한다. 첫째, 변혁적 리더십의 4차원 구조와 디지털 역량 지향을 통합함으로써 DTL의 개념적 함의를 확장하고, 기존 연구의 단일 차원적 정의가 지닌 한계를 보완하였다. 둘째, 디지털 맥락에서 현지화된 BPNS 척도의 적용 가능성을 검증하여 SDT의 맥락 간 일반화 가능성에 대한 실증 근거를 제시하였다. 셋째, 매개효과의 산업별 이질성을 규명함으로써 디지털 전환 연구에서의 상황(컨틴전시) 관점을 보완하였다. 실무적으로 본 연구 결과는 기업이 디지털 전환 리더십을 최적화하는 데 있어 구체적 지침을 제공한다. 즉, “욕구 지향과 윤리적 지도”를 갖춘 DTL을 육성하고, 산업 특성에 맞춘 차별화된 혁신 인센티브 메커니즘을 도입하며, 고용관계 및 디지털 관리·모니터링 시스템을 최적화할 필요가 있다.

다국어 초록 (Multilingual Abstract)

With the accelerated integration of digital technology into all aspects of organizational operations, digital transformation has become an irreversible trend reshaping the global business landscape (World Economic Forum, 2024). Digital Transformation Leadership (DTL), as a unique leadership paradigm in the digital era, integrates the core dimensions of transformational leadership (virtuous demonstration, visionary motivation, individualized consideration, charismatic leadership) and digital skill orientation (digital resource allocation, digital skill training, data-driven decision guidance). It has emerged as a critical antecedent of Employees’ Innovative Behavior (EIB)—the micro-foundation of organizational digital innovation and sustainable competitive advantage. Based on Self-Determination Theory (SDT), this study constructs a multiple mediating model to explore the intrinsic mechanism through which DTL influences EIB, with Basic Psychological Needs Satisfaction (BPNS)—including autonomy, competence, and relatedness needs—as the mediating variables.
A cross-sectional questionnaire survey was conducted among 582 employees from three key industries of digital transformation (manufacturing, IT, and service industries) using a 5-point Likert scale. Data were collected through a combination of online platforms (Wenjuanxing, LinkedIn) and offline enterprise collaborations, with a valid response rate of 83.1%. To ensure the rigor of the research, multiple methodological strategies were adopted: Harman’s single-factor test and the unmeasured latent factor method were used to control common method variance (CMV); reliability and validity were verified through Cronbach’s α, composite reliability (CR), average variance extracted (AVE), and confirmatory factor analysis (CFA); the mediating effects were tested using the Bootstrap method (5000 samples) via the PROCESS macro; and industry heterogeneity was analyzed through multi-group structural equation modeling (SEM) using Mplus 8.3.
The results indicate that: (1) DTL has a significantly positive direct effect on EIB (β=0.342, p<0.001), explaining 60.3% of the variance in EIB; (2) All three dimensions of BPNS (autonomy, competence, relatedness) play significant partial mediating roles between DTL and EIB, with the mediating effect values of 0.135 (23.8% of the total effect), 0.158 (27.9% of the total effect), and 0.112 (19.8% of the total effect) respectively. Competence need is the strongest mediator, followed by autonomy need and relatedness need; (3) The mediating effects of BPNS exhibit significant industry heterogeneity: in the manufacturing industry, competence need is the most critical mediator (effect=0.165); in the IT industry, autonomy need takes the leading position (effect=0.152); and in the service industry, relatedness need ranks second (effect=0.136), slightly lower than competence need (effect=0.142); (4) Additional analysis reveals no significant chain mediating effect among the three dimensions of BPNS, but the interaction between competence need and relatedness need positively moderates the mediating path of autonomy need.
This study makes several theoretical contributions: first, it enriches the connotation of DTL by integrating the four-dimensional structure of transformational leadership and digital skill orientation, addressing the limitations of single-dimensional definitions in existing research; second, it validates the applicability of the localized BPNS scale in digital contexts, providing empirical evidence for the cross-context generalizability of SDT; third, it reveals the industry heterogeneity of mediating effects, supplementing the contingency perspective in digital transformation research. Practically, the findings offer targeted guidance for enterprises to optimize digital transformation leadership: cultivating DTL with "need orientation and ethical guidance", implementing industry-differentiated innovation incentive mechanisms, and optimizing employment relationships and digital management monitoring systems.

목차 (Table of Contents)

• 1. Introduction 1
• 1.1 Research Background 1
• 1.2 Research Gaps 3
• 1.3 Theoretical Foundation and Research Objectives 5
• 1.3.1 Theoretical Foundation: Self-Determination Theory (SDT) 5
• 1.3.2 Research Objectives 7
• 1.4 Research Significance 7
• 1.4.1 Theoretical Significance 7
• 1.4.2 Practical Significance 8
• 2. Literature Review and Hypotheses 10
• 2.1 Conceptual Definitions and Operationalization 10
• 2.1.1 Digital Transformation Leadership (DTL) 10
• 2.1.2 Basic Psychological Needs Satisfaction (BPNS) 12
• 2.1.3 Employees’ Innovative Behavior (EIB) 13
• 2.2 Literature Review and Hypothesis Development 14
• 2.2.1 Literature Review 14
• 2.2.2 Direct Effect of DTL on EIB (H1) 19
• 2.2.3 Effect of DTL on BPNS 20
• 2.2.4 Effect of BPNS on EIB 22
• 2.2.5 Mediating Effects of BPNS (H2–H4) 25
• 2.2.6 Additional Hypothesis: Chain Mediating Effect (H5) 26
• 2.2.7 Industry Heterogeneity of Mediating Effects 27
• 3. Research Methodology 29
• 3.1 Sample and Data Collection 29
• 3.1.1 Research Design 29
• 3.1.2 Data Collection Process 30
• 3.1.3 Sample Characteristics 30
• 3.2 Measures 33
• 3.2.1 Digital Transformation Leadership (DTL) 33
• 3.2.2 Basic Psychological Needs Satisfaction (BPNS) 35
• 3.2.3 Employees’ Innovative Behavior (EIB) 36
• 3.3 Research Methodology 38
• 3.4 Control Variables 40
• 3.4.1 Individual-Level Control Variables 40
• 3.4.2 Organizational-Level Control Variables 41
• 3.5 Data Analysis Strategy 42
• 3.5.1 Descriptive Statistics and Correlation Analysis 42
• 3.5.2 Reliability and Validity Test 42
• 3.5.3 Common Method Variance (CMV) Test 43
• 3.5.4 Mediating Effect Test 43
• 3.5.5 Industry Heterogeneity Test 44
• 3.5.6 Robustness Test 44
• 4. Results 45
• 4.1 Common Method Variance (CMV) Test 45
• 4.2 Descriptive Statistics and Correlation Analysis 45
• 4.3 Reliability and Validity Test 47
• 4.3.1 Reliability Test 47
• 4.3.2 Validity Test 48
• 4.4 Hypothesis Testing 49
• 4.4.1 Direct Effect (H1) 49
• 4.4.2 Mediating Effects (H2–H4) 50
• 4.4.3 Chain Mediating Effect Test 51
• 4.4.4 Industry Heterogeneity Test 51
• 4.5 Results 53
• 4.5.1 Individual Effects of Control Variables 53
• 4.5.2 Pairwise Analysis of Sub-dimensions 54
• 4.5.3 Moderating Effect of Digital Literacy 55
• 4.6 Robustness Test 55
• 4.6.1 Alternative Measurement of EIB 55
• 4.6.2 Sample Subset Analysis 56
• 4.6.3 Control Variable Substitution 56
• 5. Discussion 57
• 5.1 Key Findings and Theoretical Explanations 57
• 5.1.1 Direct Effect of DTL on EIB 57
• 5.1.2 Parallel Mediating Effects of BPNS 58
• 5.1.3 Industry Heterogeneity of Mediating Effects 59
• 5.2 Discussion 60
• 5.3 Theoretical Implications 63
• 5.3.1 Enriching the Connotation and Theoretical Framework of DTL 63
• 5.3.2 Validating the Applicability of SDT in the Digital Context 63
• 5.3.3 Revealing the Industry Heterogeneity of Mediating Effects 64
• 5.4 Practical Implications 64
• 5.4.1 Cultivate DTL with “Need Orientation and Key Dimension Focus” 64
• 5.4.2 Implement Industry-Differentiated Innovation Management Strategies 65
• 5.4.3 Optimize the Digital Transformation Support System 66
• 6. Limitations and Future Research 68
• 6.1 Limitations 68
• 6.1.1 Research Design Limitations 68
• 6.1.2 Sample Limitations 68
• 6.1.3 Variable Measurement Limitations 69
• 6.2 Future Research Directions 69
• 6.2.1 Improve Research Design 69
• 6.2.2 Expand Sample Scope 70
• 6.2.3 Deepen Variable Research 71
• 6.2.4 Expand Research Boundary 73
• 7. Conclusion 75
• References 76
• Appendix 1: Factor Loadings of All Items 81
• Appendix 2: Digital Literacy Scale and Reliability/Validity 83
• 국문요지 84