Retroalimentación formativa mediada por inteligencia artificial y conciencia metacognitiva: un modelo teórico para las pedagogías activas AI-Driven Formative Feedback and Metacognitive Awareness: A Theoretical Model for Active Pedagogies.

Abstract

The integration of artificial intelligence (AI) systems in educational environments raises fundamental questions about how technologically mediated formative feedback can enhance students' metacognitive awareness. This theoretical-conceptual article proposes the METACOGNITO model, a name that evokes the articulation between metacognition and AI-mediated cognition. The model integrates four theoretical domains: (1) theories of metacognition and self-regulated learning; (2) the principles and mechanisms of AI-mediated formative feedback, including the literature on metacognitive prompts; (3) contemporary frameworks of cognitive engagement and active student-centered methodologies; and (4) AI ethics in education. Through a structured narrative review with framework synthesis, the model describes a five-phase processual cycle (activation, dynamic feedback, metacognitive reflection, regulatory adjustment, and transfer) in which AI agents act as dynamic scaffolds that amplify the learner's internal monitoring and cognitive regulation processes. The original contribution lies in the metacognitive connector construct, a design component distinguished from classical metacognitive prompts by its coupling with the feedback cycle, its multilevel intentionality, and its adaptivity. The model is accompanied by a conceptual map and five testable propositions for future empirical research. Implications for instructional design, teacher training, and educational policy are discussed, emphasizing the need for an ethical, inclusive, and culturally sensitive approach to AI implementation in active pedagogies.