GOVERNING QUALITY IN DIGITALLY TRANSFORMING AFRICAN HIGHER EDUCATION

Digital transformation exposes fragmentation in higher-education quality governance: institutional capability, programme assurance, evidence generation and accreditation judgement are frequently considered separately despite their interdependence. This thesis develops an integrated, evidence-traceable framework linking these functions, with Rwanda as the anchor jurisdiction and Kenya, Ghana and South Africa as bounded documentary comparators. It examines the capabilities supporting credible digital provision, the evidence required for programme quality and implementation readiness, the reasoning that justifies accreditation conclusions, and the governance principles requiring jurisdictional adaptation. The study adopts a critical realist orientation and a qualitative, secondary-data, multi-phase framework-development design. Its analytical architecture combines an integrative synthesis of scholarship, Rwanda policy and regulatory document analysis, standards mapping, comparative documentary appraisal and theoretical integration. The analysis distinguishes formal requirements and institutional representations from implementation evidence, actual capability and proposed generative mechanisms. The synthesis differentiates enduring institutional capability from one-time readiness and treats programme implementation readiness as a programme- and time-specific judgement. Documentary completeness alone does not establish enactment. The Evidence-Traceable Accreditation Judgement (ETAJ) model connects normative authority, institutional claims, evidence, verification, evaluative warrants, uncertainty, findings and proportionate regulatory responses. ETAJ operates within the Integrated Evidence-Traceable Quality Governance (IETQG) framework, which links institutional capability and programme assurance to accreditation judgement, monitoring and continuous enhancement. Evidence maturity, implementation alignment and an adaptation protocol specify how the architecture can be interpreted without assuming uniform legal, disciplinary or resource conditions. The principal contribution is an integrated explanation of cross-level quality governance rather than a validated causal model or a compilation of prior publications. Comparative convergence supports the plausibility of selected functional principles, not continental generalisability. Limitations include reliance on public documents, uneven source availability, rapid regulatory and technological change, absence of direct observation and participant data, and the lack of empirical validation or demonstrated inter-rater improvement. Incomplete contemporaneous records of searching, screening, coding and corpus-version control constrain reproducibility. Retrospective analytical appendices clarify these limitations but do not substitute for the original execution records.