A Critical Review of Using Learning Analytics for Formative Assessment: Progress, Pitfalls and Path Forward

Abstract

Background

While formative assessment is widely regarded as essential for improving teaching and learning, it remains difficult to operationalize due to systemic misalignment with other instructional practices, limited teacher capacity, low feedback quality, inferential uncertainty, domain-general approaches, and validity concerns.

Objectives

This editorial introduces a special issue that critically examines how learning analytics can contribute to advancing formative assessment by addressing persistent challenges in its design and implementation.

Results and Conclusion

The twelve studies featured in this issue demonstrate several innovations such as adaptive feedback, multimodal analytics, predictive modeling, dashboard design, and evidence-centered assessment frameworks. Collectively, these studies demonstrate how learning analytics can enhance formative assessment by personalizing feedback, scaling dialogic feedback, understanding the nature of feedback, improving assessment validity, automating assessment, uncovering deeper learning patterns, and improving assessment alignment with instructional goals. However, the issue also highlights several underexplored gaps, including the limited disciplinary adaptation of analytics tools, a lack of ongoing student involvement in feedback design, insufficient attention to ethical concerns and the physiological and motivational dimensions of assessment, and a limited understanding of the role of emerging technologies, in particular, Generative AI (GenAI). This editorial argues for a more critical, inclusive, and context-sensitive approach to learning analytics in formative assessment—one that centers pedagogy, teacher and student agency, and long-term educational value. The contributions of this special issue lay essential groundwork for future research, policy, and practice aimed at transforming formative assessment through learning analytics.