Supplementary Files S1–S5: Data and materials for "From Generative AI Policy to Assessment Governance: A Cross-Institutional Analysis of Policy Implementability in Higher Education"

This dataset contains the five supplementary files supporting a cross-institutional analysis of 80 assessment-facing generative AI (GenAI) policy-ecosystem entries from 79 higher-education institutions, based on official public sources available on or before 11 February 2026. S1 – Source Evidence and Coding Framework: frozen official-source evidence excerpts and URLs; governance-architecture definitions; seven-component implementability rubric; codebook v2.1, decision trees, decision rules and evidence-use protocol. S2 – Rater 1 Full-Corpus Coding: complete independent pre-adjudication coding by Rater 1. S3 – Rater 2 Full-Corpus Coding: complete independent pre-adjudication coding by Rater 2. S4 – Human Adjudication and Reliability: case-level adjudication record, inter-rater reliability (weighted kappa, linearly weighted Gwet’s AC2), governance confusion matrix and ordinal disagreement distributions. S5 – Final Analytic Dataset, Statistical Results and AI QC: final human-adjudicated dataset; descriptive and inferential results; disclosure, robustness and geographical sensitivity analyses; and the protocol, exact prompt and full results (640 checks) of the post-adjudication AI-assisted quality-control audit, which had no coding authority. Each workbook includes a README sheet. Quoted policy excerpts remain the property of the respective institutions and are reproduced for research and verification purposes, with source URLs. Related article: Djenba, F. and Chebout, M.S. (2026) ‘From Generative AI Policy to Assessment Governance: A Cross-Institutional Analysis of Policy Implementability in Higher Education’. Manuscript under review.