Research on AI Marking and Assessment in Schools
Understanding the evidence behind AI-assisted assessment, learning analytics and teacher-moderated marking systems.
Why Assessment Research Matters
High-quality feedback is one of the most powerful influences on student learning. Decades of research demonstrate that clear, timely feedback helps students identify misconceptions, refine their understanding, and improve future performance.
Advances in artificial intelligence and learning analytics now offer new opportunities to support assessment processes. When carefully designed and combined with teacher oversight, AI systems can help analyse large volumes of student responses, highlight patterns in performance, and provide structured feedback aligned to assessment criteria.
What the Research Shows
Feedback and Learning
Research consistently shows that feedback is one of the most influential factors affecting student progress. Structured feedback that clearly explains strengths, misconceptions, and next steps can significantly improve student understanding.
Teacher Workload
Providing high-quality feedback requires time. Reducing repetitive marking workload allows teachers to focus on interpreting results, supporting students, and planning targeted instruction.
Learning Analytics
Assessment data becomes most valuable when it reveals common misconceptions, topic-level weaknesses, and question-level performance trends.
Evidence-Informed Design
- Learning Analytics at Multiple Levels
- Human-in-the-Loop AI (Teacher Moderation)
- Diagnosing Misconceptions Through Assessment Data
- Structured Feedback and Metacognition
- Assessment Consistency Through Mark Scheme Alignment
- Explainable AI in Education
- Responsible AI and Data Governance
Academic References
- Bond et al. (2025). AI applications in Initial Teacher Education: A systematic mapping review.
- Dogan et al. (2025). Artificial intelligence professional development: a systematic review.
- Salas-Pilco et al. (2022). Artificial intelligence and learning analytics in teacher education: A systematic review.
- Ledger et al. (2022). Simulation platforms in initial teacher education.
- National Institute of Teaching (2026). NIoT Evidence Toolkit: AI enabled Professional Learning strand.
Contact: team@easyexaminer.com