Upper Echelons Theory and the Digital Leap: Expert-Coded Leadership and Organizational Digitalization
Description
This dataset accompanies the study “Upper Echelons Theory and the Digital Leap: Expert-Coded Leadership and Organizational Digitalization.” It provides a unique expert-coded view of how CEO digital expertise, top management team (TMT) diversity, and integration mechanisms influence digital innovation outcomes in leading global technology firms. The dataset covers 12 firm-years (2023–2024) from six representative companies at the forefront of digital transformation: Intel, NVIDIA, Huawei, Tencent, SAP, and ASML. Three independent coders contributed to the evaluation process: (1) the current CEO of a Shenzhen software company (aggressive stance), (2) a retired U.S. semiconductor CEO (conservative stance), and (3) a European scholar specialized in Upper Echelons Theory (moderate stance). This triangulated approach reduces bias and improves construct validity. Variables include: CEO Digital Expertise: five dimensions (human capital depth, track record of digital value creation, boundary spanning with IT leaders, governance of digital/data/AI risk, and centralization stance). TMT Diversity: size, functional heterogeneity (Blau index), gender diversity. Integration Mechanisms: presence of CIO/CDO/CTO roles, digital councils, and ecosystem partnerships. Digital Innovation Outcomes: digital patent families, product/feature releases, digital revenue share, process digitization milestones. Controls: firm age, size, R&D intensity, industry volatility. Contents: The dataset consists of three main workbooks (Coder A, B, C), each containing: Coding_Sheet: ratings per firm-year, including formulas for indices. TMT_Roster: member-level data (local and common names, function category, gender, source URL). Roster_Calcs: auto-calculated measures of TMT size, Blau index, and share female. Rubric & Instructions: coding anchors and coder-specific guidance. Usefulness: This dataset is valuable for scholars in strategic management, leadership, and digital transformation. It offers a replicable, transparent method to operationalize Upper Echelons Theory beyond demographics and surveys, enabling the study of substantive expertise and team integration. Researchers can reuse the coding rubric, adapt the roster approach to other industries, or extend the methodology with automated coding (e.g., NLP, machine learning). The multi-rater structure also allows for intercoder reliability analysis, an uncommon but critical step in leadership research. This dataset directly supports the empirical analysis presented in the associated article, offering open and replicable materials for future comparative, cross-industry, and longitudinal studies.
Files
Steps to reproduce
The dataset is designed for transparent replication of the study “Upper Echelons Theory and the Digital Leap: Expert-Coded Leadership and Organizational Digitalization.” The following steps describe how the coding process and analysis can be reproduced using the files provided: Access the dataset files Download the three Excel workbooks: UET_CoderA_FINAL.xlsx, UET_CoderB_FINAL.xlsx, UET_CoderC_FINAL.xlsx. Each file contains a Coding_Sheet, TMT_Roster, Roster_Calcs, Rubric_Short, and Instructions. Review coding rubric and instructions Open Rubric_Short in each file to see variable definitions, scales, and coding anchors. Instructions detail coder-specific posture (aggressive, conservative, moderate). Inspect TMT_Roster Each firm-year’s top management team (TMT) members are listed with local/common names, function category, gender, and source URLs from annual reports and filings. Roster_Calcs automatically computes TMT size, Blau index, and female share. Check coding sheets Coding_Sheet contains the coders’ ratings for each firm-year, covering: CEO digital expertise (5 dimensions). TMT diversity (size, Blau, gender). Integration mechanisms (roles, councils, ecosystems). Digital innovation outcomes (patents, product releases, digital revenue share, process digitization). Index formulas auto-calculate: CDE_Index (with reverse-coded centralization) and Integration_Index. Reliability assessment Compare Coder A, B, and C values for overlapping firm-years. Compute intercoder reliability (Krippendorff’s α, Cohen’s κ) using statistical software (R, Stata, SPSS, or Python). Prepare the analysis dataset Consolidate coder ratings into a single panel dataset. Average coder scores or use adjudication rules if required. Merge with objective outcomes and control variables (size, R&D, age, industry volatility). Run regression analysis Estimate models: Digital Innovation = TMT Diversity + CEO Expertise + Interaction + Controls. Test for inverted-U moderation (quadratic CEO Expertise term). Test mediation by Integration Index (SEM or regression-based mediation). Apply firm/year fixed effects for robustness. Reproduce figures and tables Generate descriptive tables, correlation matrices, and regression outputs. Create plots for moderation (interaction and inverted-U) and mediation effects. Extend the dataset Researchers can replicate this method for additional firms/industries by following the rubric, filling new TMT_Roster entries, and reapplying the coding process. By following these steps, the coding procedure, reliability tests, and regression analyses reported in the article can be fully replicated and extended.
Institutions
- Universita Ca' Foscari