📄 Full text: arXiv:2604.03926 · local · arXiv:2504.20082 · local · arXiv:2507.04295 · local
Assessment designed to inform ongoing instruction and learning, as opposed to summative evaluation. AI systems can generate, validate, and adapt formative assessment items at scale, though quality varies dramatically across assessment types.
AI-Generated Formative Items
Multiple-Choice Questions (CODE-GEN)
Duan et al. (2026) demonstrate that agentic AI can reliably generate MCQs for coding comprehension when validated across seven pedagogical dimensions. Success rates reach 98.6% for concept alignment and 79.9% for feedback quality—suggesting that AI is strongest on verifiable dimensions and weakest on instructional-judgment dimensions.
Automated Essay Scoring (MASS)
Kamalov et al. (2026) implement a multi-agent framework (MASS) for essay scoring. Preliminary results show improved consistency over stand-alone LLMs, though interpretability of multi-agent scoring decisions remains an open challenge.
Curriculum-Grounded Feedback (LearnLens)
Zhao et al. (2025) present LearnLens, a modular LLM system for science education feedback that addresses three persistent problems in AI formative assessment:
1. Error-aware assessment — captures nuanced reasoning errors rather than surface mistakes 2. Topic-linked memory chains — replaces noisy similarity-based RAG with structured curriculum-grounded retrieval 3. Educator-in-the-loop — teacher customisation and oversight, not full automation
Key differentiator: LearnLens uses a structured, topic-linked memory chain rather than traditional RAG similarity search, improving relevance and reducing noise. This connects to the broader tension in human-in-the-loop-ai: scalable automation with expert validation.
Design Trade-offs
| Dimension | AI Suitability | Human Requirement |
|---|---|---|
| Factual correctness | High | Low |
| Concept alignment | High | Medium |
| Distractor quality | Low | High |
| Feedback depth | Low | High |
| Rubric consistency | Medium | Medium |
Risk: Assessment as Surveillance
Formative assessment systems can shift from learning-support tools to behavior-monitoring infrastructure. The same data streams that enable adaptive tutoring can enable punitive tracking if governance is weak.
Related Pages
- cotal-formative-assessment-scoring-2026 — LLM scoring with ECD-aligned rubrics
- agentic-ai-pedagogical-best-practice-2026 — Agent feedback pre-empting self-assessment
- authentic-products-authenticated-processes-2026 — Process transparency and feedback-use evidence
- correct-answer-trap-ai-tutor — 8 of 8 papers in May 28 scan
- rubric-aware-grading-rec-cbm — 2 of 8 papers in May 28 scan
- llm-sentiment-analysis-education-research — 1 of 8 papers in May 28 scan
- student-misconceptions-conditionals-loops-taxonomy — taxonomy for precise diagnosis within formative assessment cycles
- clara-collaboration-literacy-dashboard — CLARA: An AI-Augmented Analytics Dashboard for Collaboration Literacy
- cyberscholar-genai-writing-feedback — Generative AI Feedback, English Writing and Teacher Rubrics: A Multiple-Case Study of CyberScholar
- structured-llm-feedback-programming — The Effects of Structured LLM-Generated Feedback on Programming Assignment Performance
- lata-ferpa-compliant-local-llm-autograder — Rapid feedback enabling regrading and expanded TA hours
- self-referential-l2-writing-llm-assessment — LLMs as complementary weakness detectors alongside teacher strengths
- short-answer-scoring-quality-degradation — Mid-range degradation threatens nuanced feedback in formative systems
- moodle-ai-tutoring-deep-learning — Uses formative cycles within LMS to promote deep learning
- multimodal-ai-feedback-learning — Zhao et al.: AI multimodal feedback as scalable formative assessment matching educator quality
- ground-truth-reliability-aied — Thomas et al.: formative systems using AI labels need validity evidence beyond inter-rater agreement
- sequenced-ai-feedback-learning — Cao et al. RCT: sequenced AI feedback harmed learning vs. direct feedback despite better engagement
- retrieval-augmented-tutoring-algorithm-kite — KITE's simulated-student evaluation framework for tutoring feedback
- llm-student-misconception-identification — Misconception-specific interventions enabled by LLM diagnosis
- teacher-authored-prompts-student-ai-dialogue — DOK and instructional intent alignment in assessments
- llm-feedback-programming-classroom — LLM-generated feedback as formative assessment in programming
- vismatic-secure-sandbox-cs-education — Process-oriented monitoring for formative assessment in AI era
- foxglove-writing-feedback-experts-llms — Goal-oriented and anchored feedback for formative writing assessment- cross-dataset-bloom-question-classification -- LLMs with tailored prompting generalize better than supervised models for cross-dataset Bloom taxonomy classification
- llm-chatbots-cs-multiple-choice -- ChatGPT answers with explanations do not improve student MCQ performance; GPT-4o/5 outperform smaller models
- ai-assessment-human-tutors -- AI-driven assessment of human tutor training performance correlates with real-li...
- learning-to-prompt-adaptive-tutoring -- Learning to Prompt: Improving Student Engagement with Adaptive LLM-based High-School Tutoring
- confidence-aware-student-drawing-assessment -- Confidence-Aware Automated Assessment of Student-Drawn Scientific Models
- psyscore-essay-scoring-zpd-feedback -- PsyScore: A Psychometrically-Aware Framework for Trait-Adaptive Essay Scoring and ZPD-Scaffolded Feedback
- correct-answer-trap-misconceptions -- Fine-tuned classifiers detect only 57% of hidden student misconceptions, while reasoning models reach 84% but at 8:1 false alarms — a detect-verify-escalate pipeline with diagnostic follow-up questions is proposed.
- adaptive-pretesting-retention -- Adaptive AI pretesting boosts initial learning, but long-term retention over 7 weeks depends on subsequent retrieval practice — learner-directed AI study underperforms structured retrieval.
- machines-misread-pedagogical-quality -- Human-machine disagreements in AI pretest evaluation are systematic; rubric revision has a larger alignment effect than rationale-first evaluation, and the two are complementary.
- critical-engagement-code-completion — measuring critical engagement with AI code completion
- cogtax-cognitive-taxonomy — formative assessment tools
- automated-grading-linux-bash-examinations-large-language-models — Automated Grading of Linux/Bash Examinations Using Large Language Models
- from-answer-generators-to-reasoning-facilitators-ai-tutors — From Answer Generators to Reasoning Facilitators: Designing AI Tutors for Mathem
- evaluating-interactivity-automated-assessment-ai-generated-explorable-explanations — Evaluating Interactivity: Toward Automated Assessment of AI-Generated Explorable
- prompt-coach-agentic-tutor-prompt-engineering — Self-correction oriented tutoring
- aied-unfinished-mission-bypass — AIED's Unfinished Mission
- llm-design-problems-hot-pjbl — HOT assessment method (2026-07-14)
- teaching-feedback-classification-benchmark — Formative feedback (2026-07-14)
- learning-engagement-assistant-lea — agentic RAG+KC tutoring system; first real-student classroom deployment (n=8) and cross-course scalability test
- mllm-scientific-visualization-literacy — assessment instrument for visualization competence
- assessment-team-problem-solving-computing-education — Assessment in Team Problem-Solving Exercises in Computing Education
- automated-formative-assessments-a-level-sciences — Automating the marking of handwritten mock exams enables much higher formative-assessment frequency
- llm-psychometric-calibration-cdp — CDP framework dramatically improves LLM-simulated examinee alignment with human ...
- llm-misconception-difficulty-easy-trap — LLMs systematically underestimate misconception-driven item difficulty ('Easy Tr...
- responsible-assessment-ai-era-stanford-2026 — Stanford convening on extending formative assessment with AI feedback loops
📎 63 other pages tagged formative-assessment
- A Durability and Cross-Language Transfer Benchmark for a Validated Teaching-Feedback Classification Protocol
- Agentic AI and Pedagogical Best Practice: The Tension Between Automation and Learning
- AI Learning Transfer
- AI Peer Feedback Systems
- AI-Generated Lesson Plans in Civic Education
- AI-Integrated Learning Management System for Middle School: A Longitudinal Study of Learning Outcomes
- AIED's Unfinished Mission: Centering Agency and Motivation in the Age of Effortless Bypass
- AISSA: AI-based Student Slides Analysis Tool for Academic Presentations
- Aligning LLM-Simulated and Human Examinees for Psychometric Calibration: A Cognitive Diagnostic Profiling Approach
- Artificial intelligence and feedback in university education: effectiveness and student perceptions
- Assessing the Impact and Underlying Pathways of Sequenced AI Feedback on Student Learning
- Assessment in Team Problem-Solving Exercises in Computing Education
- Authentic Assessment
- Automated Grading of Linux/Bash Examinations Using Large Language Models
- Automated Grading Systems
- Automatic Short Answer Grading with LLMs
- Catching The Correct Answer Trap: Characterising AI Tutor Blind Spots When Analysing Student Reasoning
- CogTax: A Four-Level Cognitive Taxonomy for Command-Line Computing Education
- Collaborative AI Tutoring
- Combating Harms of Generative AI in CS1 with Code Review Interviews and a Flipped Classroom
- Confidence-Aware Automated Assessment of Student-Drawn Scientific Models
- CoTAL: Human-in-the-Loop Prompt Engineering for Generalizable Formative Assessment Scoring and Feedback
- Cross-Dataset Bloom Question Classification: Supervised Models and Prompted LLMs
- DebugTracker: Lightweight Process Evidence for Classroom Debugging
- Do Gains from Generative AI-Enabled Adaptive Pretesting Persist? Evidence from a Retention Study
- Estimating Learners' Skill Acquisition Without Temporal Information
- Evaluating Interactivity: Toward Automated Assessment of AI-Generated Explorable Explanations
- Feedback Loop
- FOXGLOVE: Comparing Goal-Oriented Writing Feedback from Experts and LLMs
- From Answer Generators to Reasoning Facilitators: Designing AI Tutors for Mathematical Reasoning in High-Stakes Environments
- Generative AI Feedback, English Writing and Teacher Rubrics: A Multiple-Case Study of CyberScholar
- How AI Is Changing Teaching Workflows
- Hybrid E-Assessment in Higher Education: Semi-Automated Grading of Paper-Based Written Examinations
- Learning Engagement Assistant (LEA): Cross-Course Scalability and Classroom Evaluation of an Agentic AI Tutoring System
- LLM-assisted sentiment analysis for integrated computational and qualitative mixed methods education research: A case study of students' written reflection assignments
- LLM-Generated Design Problems for Assessing Higher-Order Thinking in Project-Based Learning
- LLM-Generated Feedback in Introductory Programming: A Classroom Study
- Metacognition
- Prober.ai: Gated Inquiry-Based Feedback via LLM-Constrained Personas for Argumentative Writing
- Programming Intelligent Tutoring Systems
- Prompt Coach: An Empirical Evaluation of an Agentic Tutor for Learning Prompt Engineering in Software Development
- PsyScore: A Psychometrically-Aware Framework for Trait-Adaptive Essay Scoring and ZPD-Scaffolded Feedback
- Quality-Conditioned Agreement in Automated Short Answer Scoring: Mid-Range Degradation and the Impact of Task-Specific Adaptation
- REC-CBM: Rubric-Aware Error-Correction Concept Bottleneck Models for Trustworthy Open-Ended Grading
- Responsible Assessment in the AI Era: Key Insights from a Future-Focused Conference
- Retrieval-Augmented Tutoring for Algorithm Tracing and Problem-Solving in AI Education
- Self-Regulated Learning
- Slide Deck Q&A Quality Assurance App: A Multi-Stage Pipeline for Pedagogical Question Generation
- Socratic AI Dialogue
- Student Evaluation of Repeated AI Feedback Across a Semester of Writing
- TeachBench - Evaluating LLM Teaching Ability
- Teacher-Authored Prompts for Configuring Student-AI Dialogue: K-12 Classroom Implementation
- The Correct Answer Trap: Pedagogically-Grounded Detection and Feedback for Hidden Misconceptions
- The Easy Trap: Why LLMs Underestimate Misconception-Driven Difficulty
- The Tutoring Effectiveness Index: Predicting LLM Math Tutor Quality from Four Conversation Signals
- TibetCPR: A Multimodal Tactile Feedback System for CPR Training in High-Altitude Regions
- To Tab or Not to Tab: Measuring Critical Engagement in AI Code Completion Tools Using Behavioral Signals and Attention Checks
- VISMATIC: Secure Containerized Framework for Process-Oriented CS Education Monitoring
- What Does the Credential Still Certify? Cognitive Stewardship for AI-Mediated Education
- What Don't You Understand? Using Large Language Models to Identify and Characterize Student Misconceptions About Challenging Topics
- What Makes Words Hard? Sakura at BEA 2026 Shared Task on Vocabulary Difficulty Prediction
- When AI Is Wrong on Purpose: How Students Respond to Buggy GenAI Code
- Why Machines Misread Pedagogical Quality: Human-Machine Alignment in LLM-Based Pretest Question Evaluation