Leadership for sustainability in African higher education: protocol, search logs, screening decisions and extraction data for a scoping review, 2010–2025

This deposit supports a scoping review of research on leadership for sustainability in African higher education institutions, 2010 to 2025, with attention to fragile and conflict-affected settings. The review follows the PRISMA-ScR guideline. The protocol was frozen on 2026-10-23 at 15:53:41 UTC, before any search, and its SHA-256 hash is recorded in logs/protocol_freeze.txt . Seven dated amendments (A1 to A7) are logged with their reasons. Contents: the protocol; search logs and the Scopus query strings; OpenAlex bibliographic records and abstracts for the de-duplicated candidate set and the screening set; first-pass screening decisions with reason codes; the author's independent screening of a random 10% sample (84 records) and the agreement statistics; the re-screen under amendment A6; the extraction tables for the 31 included studies; the coding of each study against the four components of a global model of sustainability leadership; the PRISMA-ScR flow diagram; the scripts used to retrieve, de-duplicate and tabulate records; and the text of the manuscript as submitted. How the work was done: the author wrote the protocol and research questions, double-screened the random sample, supplied paywalled full texts and checked the findings. The searches, abstract screening, data charting and drafting were carried out with Claude (Anthropic), and Paperpal was used for language editing, all under the author’s direction Screening agreement between the author and the AI reviewer was 90.5%, with Cohen's kappa of 0.19 (three categories), which led to the rule revision in A6. Not included: Scopus exports and Scopus abstracts, which cannot be redistributed under the Scopus licence (the three Scopus queries are given so others with access can repeat them), and the full texts of included studies, which are under publisher copyright. OpenAlex records are CC0. Limitations: twelve of the 31 included studies were charted from abstracts only; terms were searched in English; Web of Science was not searched; the re-screen after A6 covered flagged records, not the whole set.