Learning analytics and data-driven decision-making in educational institutions: a policy perspective

Abstract

The term of learning analytics (LA) is a recently introduced novel paradigm in the field of education that focuses on collecting, processing and analyzing the amount of student generated data to rational support pedagogical or administrative decision-making. While many technologies, including learning dashboards, predictive modeling and early-warning systems have developed at a fast pace, there has been little progress on policy frameworks addressing their ethical and transparent use, and responsible applications. This paper explores the policy challenge of Data Driven Decision Making (DDDM) in educational institutions in particular, as well as the governance, ethical and institutional challenges arising from Implementing Learning Analytics. A Systematic Literature Review was performed for articles published from 2012 to 2025 in the databases: Scopus, IEEE Xplore, SpringerLink and ERIC. The literature reviewed was classified in four categories: conceptual underpinnings of DDDM; institutional integration and training; predictive analysis and early-warning systems; and privacy, ethics, and data protection. The results suggest that learning analytics enhances personalized learning, student support, institution effectiveness, yet at the same time, concerns about privacy and control of data, algorithmic bias, student surveillance, transparency, and educator autonomy have been raised. The paper outlines a four levels policy practical framework in the pedagogical alignment, operational governance, technical governance, ethical and legal protections, and strategic institutional leadership. Robust governance combined with data literacy, informed consent, continuous monitoring of algorithmic fairness, and adapting policies to promote responsible, equitable, and trustworthy implementation of learning analytics is needed for sustainable implementation.