Beyond Universal AI Models: A Comparative Analysis of Vision-Driven (GCC) vs. Regulation-Driven (EU) Adoption Frameworks

This dataset contains replication materials for the study \"Beyond Universal AI Models: A Comparative Analysis of Vision-Driven (GCC) vs. Regulation-Driven (EU) Adoption Frameworks\". The data supports a mixed-methods approach combining a survey of public sector professionals in the Gulf Cooperation Council (GCC) with a document analysis of National AI Strategies (NASs). The repository includes the following files: 1. AI Adoption Survey (Responses).xlsx: Contains raw, anonymized survey responses (n=203) collected from public sector professionals across six GCC countries: Saudi Arabia, UAE, Qatar, Bahrain, Kuwait, and Oman. The survey measures four key constructs: Technical Infrastructure (TI), Organizational Readiness (OR), Governance Environment (GE), and AI Outcomes (AIO). 2. SmartPLS Report.xlsx: Provides the cleaned dataset and the complete output from the Partial Least Squares Structural Equation Modeling (PLS-SEM) analysis used to validate the study's theoretical model. 3. The Codebook (v2): A detailed protocol for the document analysis of GCC National AI Strategies4. This document outlines the inclusion/exclusion rules and definitions for coding capacity-building references (TI, OR, GE) 5. It also details the methodology for Term-Salience Scoring (TF-IDF) used to quantify the emphasis on specific AI capabilities within the strategies. 4. Supplementary Materials: A compilation of five appendices providing extended statistical diagnostics: - Appendix A: Detailed survey respondent demographics by country. - Appendix B: Item-level descriptive statistics (Mean, Standard Deviation) for all survey constructs. - Appendix C: Clustering diagnostics, including Elbow plots and within-cluster sum of squares, supporting the identification of distinct adoption clusters. - Appendix D: Principal Component Analysis (PCA) details, including KMO values (0.937), Bartlett's Test results (p<.001), and Scree plots confirming the factor solution. - Appendix E: Extended structural model diagnostics, reporting Coefficient of Determination (R^2) and Effect Sizes (f^2) for the predictor paths.