Introduction This systematic review explores how AI-powered Learning Analytics (LA) contribute to the development of metacognitive and socioemotional competencies in educational settings. Method Following the PRISMA guidelines, a total of 161 peer-reviewed articles published between 2013 and 2023 were retrieved from the Scopus database and analyzed. Results The findings reveal a predominant focus on predictive (46%) and prescriptive (28%) analytics, while descriptive (16%) and social-affective (10%) approaches remain significantly underrepresented. This imbalance raises critical concerns regarding the extent to which current LA implementations support higher-order competencies such as self-regulation, reflection, emotional awareness, and collaborative learning. The study identifies four major categories of LA—descriptive, predictive, prescriptive, and social-affective—and examines their pedagogical implications considering learner-centered principles. Discussion Special attention is given to the potential of LA to scaffold metacognitive strategies and foster socioemotional growth, particularly when designed with transparency, learner agency, and emotional sensitivity. Ultimately, the review advocates for a more balanced and human-centered research agenda, calling for the redefinition of educational quality through the integration of holistic learner development in AI-enhanced learning environments.
Pacheco et al. (2025) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: