FLEX-COMP: A Conceptual Framework for Competency-Based Learning and Assessment Governance in Higher Education

Abstract

Higher education faces growing pressure to offer flexible learning while demonstrating graduate outcomes. However, institutions often develop competency-based education, flexible pathways, blended learning, assessment, learning analytics, and governance as separate initiatives, weakening coherence across curriculum, progression, and quality processes. This conceptual paper develops the FLEX-COMP framework through literature-based conceptual synthesis. FLEX-COMP positions program and course competencies as its organizing core. It proposes modular and stackable learning units that allow variation in pace, place, and sequence while maintaining shared competency expectations. A blended learning ecology supports these pathways, combining face-to-face, synchronous online, and asynchronous learning activities, informed by the Community of Inquiry perspective. Progression is proposed to depend on demonstrated mastery rather than time spent, with projects, simulations, reflective artifacts, and e-portfolios generating evidence of learning. The framework further proposes links between assessment evidence, learning analytics, and governance. With appropriate institutional governance and interpretation, analytics dashboards may support monitoring competency and learner progression across pathways and inform curriculum review, learner support, and institutional decision-making. The framework also identifies the need for governed approaches to distinguish AI-assisted performance from independent demonstrations of competencies; empirical examination is needed to assess the validity, ethics, and feasibility of both. Conceptualized as a concentric model, FLEX-COMP connects competencies, pathways, curriculum design, blended learning, assessment, analytics, and governance. It offers design propositions for institutions considering flexible, competency-driven learning, but its implementation, effectiveness, equity, and applicability across disciplines and contexts remain empirically untested.