Best AI for Assessment and Feedback in Education in 2026

Quick Answer: AI for educational assessment generates Hattie-Timperley feedback designed around the three questions (where am I going? how am I going? where to next?) with specific, actionable, goal-referenced responses; Black-Wiliam formative assessment strategy implementation plans including questioning techniques, exit ticket designs, peer assessment protocols, and self-assessment frameworks; Wiggins-McTighe backward design unit planning templates starting from desired results and evidence of understanding before designing learning activities; standards-based grading rubrics that separate academic mastery from compliance behaviors; comment-only feedback protocols that improve student revision behavior; and differentiated formative assessment tools accessible to diverse learners. EduGenius (edugenius.app) helps teachers design high-impact assessment and feedback systems for Grades K-9.

Assessment is the most frequently performed teaching act. Teachers assess constantly — formally and informally; through tests and quizzes; through exit cards and observations; through questions asked in class; through conversations during independent work; through the quality of student writing, projects, and products. Yet the research on how these assessments are designed and how feedback is delivered shows a persistent gap between common practice and what research identifies as most effective.

John Hattie's synthesis of over 800 meta-analyses places feedback among the highest-impact educational interventions (d = 0.70). But Hattie and Timperley's analysis also shows that the form of feedback matters enormously: feedback at the task level (how well did I do this specific task?) is more effective than feedback at the self level (praise or criticism of the student as a person); and feedback that is timely, specific, and informative about how to improve is substantially more effective than feedback that is only evaluative (a grade). These distinctions are not trivial — most teachers provide more evaluative, self-level, and delayed feedback than the research recommends.

The frameworks developed by Hattie and Timperley; Black and Wiliam; Wiggins and McTighe; Sadler; and Guskey represent the most influential research-based contributions to understanding how assessment and feedback can be designed to maximize their impact on student learning.

Research Foundations of Assessment and Feedback

John Hattie and Helen Timperley: The Feedback Model

John Hattie and Helen Timperley (University of Auckland), in "The Power of Feedback" (Review of Educational Research, 2007), produced the most influential research synthesis on feedback effects and developed a model of feedback that has significantly shaped assessment practice:

Three Feedback Questions: Hattie and Timperley organize effective feedback around three questions that correspond to three levels of learning:

1. Feed Up — Where am I going?: The goal, the target, the success criteria — what does "good" look like, and what am I aiming for? Feedback that addresses this question helps students understand what they are working toward. Many students do not have a clear understanding of the learning goal — they know what the assignment is but not what learning it is meant to develop.
2. Feed Back — How am I going?: The gap between the student's current performance and the goal — what is the current level of achievement, and how does it compare to the learning target? Feedback at this level is diagnostic and informative: it tells the student where they stand relative to where they need to be.
3. Feed Forward — Where to next?: The specific strategies, activities, or steps that will move the student toward the goal — what should I do to close the gap? Feed-forward feedback is the most actionable component: it gives the student something specific to do rather than merely informing them that a gap exists.

Four Levels of Feedback: Hattie and Timperley describe four levels at which feedback can be directed:

1. Task level: Feedback about how well the specific task has been completed — correctness; accuracy; completeness. "Your explanation of photosynthesis is accurate but doesn't explain why the light reaction is necessary for the Calvin cycle."
2. Process level: Feedback about the processes and strategies used to complete the task — deeper than task-level feedback and more transferable. "You're checking your arithmetic carefully and that's preventing errors. Now try checking your conceptual set-up at the beginning — is your equation representing the right relationship?"
3. Self-regulation level: Feedback directed at developing metacognitive and self-monitoring capacities — the ability to evaluate one's own work, set learning goals, and manage effort strategically. "What evidence would tell you that your argument is strong? What would a skeptical reader say about this paragraph?"
4. Self level: Personal praise or criticism directed at the student as a person ("You're so smart"; "You should be ashamed"). Research consistently shows this level of feedback is least effective and can actually undermine performance — particularly for students with entity (fixed) theories of intelligence.

Timing of Feedback: Hattie and Timperley also synthesize research on feedback timing. Immediate feedback is most effective for procedural and accuracy tasks; delayed feedback (allowing time for reflection and self-correction attempts first) may be more effective for complex tasks where elaboration and deeper processing are the goal.

Feedback in Context: Crucially, feedback is most effective when students are in a learning context that values effort, strategy, and learning — rather than performance, grades, and comparison with others. In performance-focused contexts, feedback about deficiencies increases anxiety and reduces motivation rather than promoting improvement.

Paul Black and Dylan Wiliam: Assessment for Learning

Paul Black (King's College London) and Dylan Wiliam (now ETS and Vanderbilt University), in "Inside the Black Box: Raising Standards Through Classroom Assessment" (Phi Delta Kappan, 1998) and subsequent work including Assessment for Learning: Putting It into Practice (2003, with Christine Harrison, Clare Lee, Bethan Marshall, and Mary James), produced the most influential research and practice framework for formative assessment:

The Black Box Metaphor: Black and Wiliam describe the classroom as a "black box" — inputs (teaching, curriculum, student effort) go in; outputs (test scores, grades) come out; but what actually happens inside — how teaching affects learning at the moment it is happening — is largely invisible and unmonitored. Formative assessment is the tool for opening the black box: making the learning process visible in real time so that both teachers and students can use that information to adjust.

What Counts as Formative Assessment: Black and Wiliam define formative assessment as any assessment that provides information that teachers and students can use to close the gap between current learning and desired learning goals. Formative assessment is defined not by its format (a quiz; an observation; a discussion) but by its function: is the information used to make teaching and learning decisions?

Five Key Strategies: Black and Wiliam, in their "Developing the Theory of Formative Assessment" (2009, with Nicol and Macfarlane-Dick), identified five key formative assessment strategies supported by the evidence:

1. Clarifying and sharing learning intentions and success criteria: Students learn better when they understand what they are trying to learn and what "good" looks like. This is not just handing students the rubric — it involves students actively engaging with success criteria through examining exemplars, evaluating sample work, and co-constructing criteria.
2. Engineering effective classroom discussions and tasks: Designing questions, discussions, and tasks that reveal the quality of student learning — not just whether students got the right answer but whether they understand why. Effective questioning techniques: longer wait time; no-hands-up (cold calling or random selection); thinking-pair-share; mini-whiteboards; hinge questions (questions specifically designed to reveal whether students have understood a key concept or still hold a misconception).
3. Providing feedback that moves learners forward: Feedback that is task-level or process-level (not self-level); that answers the three feedback questions; that is actionable; and that is received in a context where students have the opportunity and expectation to respond to it.
4. Activating students as instructional resources for one another: Structured peer assessment and peer learning. Students who are one step ahead of their peers often explain concepts more accessibly than teachers who are many steps ahead. Peer assessment also develops students' ability to evaluate work against criteria — a skill that improves their own self-assessment.
5. Activating students as the owners of their own learning: Developing students' capacity for self-assessment — the ability to evaluate their own work against success criteria, identify gaps, and take action. Self-assessment is the most powerful source of feedback because it is always available (unlike teacher feedback, which is intermittent).

The Learning Gap: Black and Wiliam synthesize evidence that formative assessment, implemented with integrity, produces substantial gains in student learning — estimated at 0.4-0.7 standard deviations in their most cited estimates — particularly for lower-achieving students, who benefit most from the gap-closing information that formative assessment provides.

Grant Wiggins and Jay McTighe: Understanding by Design

Grant Wiggins (1950-2015) and Jay McTighe, in Understanding by Design (1998; second edition 2005), developed the "backward design" framework — one of the most widely adopted curriculum planning approaches in K-12 education:

Backward Design: The key insight of Understanding by Design (UbD) is that effective curriculum planning should begin with the end in mind — specifically, with the desired results (what do we want students to understand, know, and be able to do?) — and work backward to identify the evidence that would demonstrate that those results have been achieved, before designing the learning activities that will develop that understanding. This reverses the common "activity-first" planning approach in which teachers plan activities and then assign grades at the end.

Three Stages of Backward Design:

1. Stage 1 — Identify Desired Results: What should students understand, know, and be able to do? Wiggins and McTighe distinguish among knowledge goals (factual knowledge students should have); skill goals (things students should be able to do); and, most importantly, understandings — the transferable, conceptual insights that constitute the core of the discipline and that cannot be directly "covered" but must be developed through experience. Understandings are best framed as "big ideas" and "essential questions."
2. Stage 2 — Determine Evidence of Learning: What evidence would demonstrate that students have achieved the desired results? This stage requires thinking like an assessor — identifying the specific performances, products, and processes that would provide convincing evidence of understanding. Wiggins and McTighe advocate for "authentic assessment" — tasks that ask students to apply knowledge in real-world-like contexts rather than merely recall information.
3. Stage 3 — Plan Learning Experiences and Instruction: Only after identifying desired results and evidence of learning are learning activities designed. The sequence of activities, experiences, and instruction is designed to support students in developing the understanding and skills identified in Stages 1 and 2.

Six Facets of Understanding: Wiggins and McTighe identify six facets of genuine understanding that go beyond surface knowledge recall:

1. Explanation: Students can explain — provide thorough, supported, and justifiable accounts of phenomena, facts, and data.
2. Interpretation: Students can interpret — make meaning; provide revealing historical, artistic, personal narratives; translate, paraphrase.
3. Application: Students can apply — effectively use and adapt knowledge in diverse, real contexts.
4. Perspective: Students can shift perspective — critically consider different points of view; see and hear points of view through critical eyes and ears.
5. Empathy: Students can empathize — find value in what others might find odd or counterintuitive; perceive sensitively.
6. Self-knowledge: Students can self-assess — show metacognitive awareness; perceive the personal style, biases, and projections that shape self-understanding.

Royce Sadler: Formative Assessment and Gap-Closing

Royce Sadler (Griffith University, Australia), in "Formative Assessment and the Design of Instructional Systems" (Instructional Science, 1989), produced one of the most theoretically precise accounts of how formative assessment actually works — and what students need to learn to make self-assessment effective:

The Formative Assessment Triangle: Sadler proposes that effective formative assessment requires students to have three things simultaneously:

1. A concept of the quality that is being aimed at — what does good work look like?
2. The ability to compare their current work to that standard — can they see the gap?
3. A repertoire of strategies for closing the gap — what specific actions can they take?

Self-Assessment as the Goal: Sadler argues that the ultimate goal of formative assessment is to develop in students the ability to assess their own work — to see what the teacher sees; to identify the gap between current and desired performance; and to take action to close it. Until students can do this, they are dependent on the teacher for assessment information — which is always delayed, intermittent, and filtered. The development of self-assessment capacity is therefore the most important long-term goal of assessment.

The Problem with Grades: Sadler identifies a fundamental problem with grades as feedback: grades are evaluative (this is good; this is bad) but not informative (here is what would make it better) and not developmental (here is the strategy that will help you improve). A student who receives a C+ knows their work is somewhat below average but does not know what specifically needs to improve or how to improve it. Comment-only feedback — which Sadler (and Black and Wiliam) advocate for many purposes — provides information without the evaluative closure that grades create.

Thomas Guskey: Standards-Based Grading

Thomas Guskey (University of Kentucky), in Developing Grading and Reporting Systems for Student Learning (2006) and related work, developed the most systematic framework for standards-based grading (SBG):

The Problem with Traditional Grading: Guskey identifies multiple validity threats in traditional percentage-based and letter grading: grades that combine academic achievement with compliance behaviors (completing homework on time; class participation; effort) are reporting multiple, incommensurable things in a single symbol, making interpretation impossible. A student who earns an A partly through completing extra credit has an A that means something different from a student who earns an A purely through demonstrated mastery — yet they are indistinguishable on the report card.

Standards-Based Grading Principles: Guskey advocates for grading practices that: report grade-level standards mastery separately from compliance behaviors (if both are reported, they should be in separate categories); use evidence of most recent and most valid performance rather than averaging performance across the entire learning period (a student who shows mastery in Week 10 should not have that mastery diluted by poor performance in Week 3 when they were still learning); and use rubrics that describe mastery criteria explicitly rather than arbitrary percentage thresholds.

The Grading and Learning Relationship: Guskey's most important practical argument is that grading practices should be aligned with learning goals: if the goal is to develop proficiency in a skill, the grade should reflect proficiency — not effort, completion, or improvement — and students should have opportunities to demonstrate proficiency after additional learning without being permanently penalized for earlier struggles.

AI Applications in Assessment and Feedback

High-Quality Feedback Design System

"Design a complete feedback system for a Grade 7 history classroom — 'Evidence-Based Feedback: Developing Historical Thinkers' — grounded in Hattie and Timperley's three-question feedback model, Black and Wiliam's five formative assessment strategies, and Sadler's self-assessment development framework. This system replaces generic red-pen marking with structured, three-level feedback that develops students' capacity to assess and improve their own historical writing. Feedback System Components: Component 1 — Establishing the Standard (Done Once Per Assignment Type): Before students receive any feedback on their work, they must understand what 'good' looks like. For each major essay type (evidence-based paragraph; comparative essay; document analysis), the teacher provides two sample student essays — one strong, one developing. Students analyze the samples using the success criteria and identify: 'What specifically makes the strong sample stronger? What specifically would improve the developing sample?' This activity develops the first component of Sadler's formative triangle: a concept of quality. Students keep the success criteria and sample analyses in their writer's notebook for reference. Component 2 — Self-Assessment Before Teacher Feedback: Before submitting a draft for teacher feedback, students complete a self-assessment using the success criteria: 'Circle the score you give yourself on each criterion. For each criterion below proficient, write one specific sentence identifying what you would change.' The self-assessment has two functions: it develops metacognitive capacity; and it provides the teacher with information about the student's own perception of their work, which can inform the focus of teacher feedback. Component 3 — Targeted Teacher Feedback (Three Questions): Teacher feedback on each draft addresses the three Hattie-Timperley questions: Feed Up: 'You're working toward [specific learning goal]. The most important criterion for this essay type is [specific criterion].' Feed Back: 'In this essay, you [specific observations about what the student did]. Your use of evidence is [specific evaluation]. Your historical argument is [specific evaluation].' Feed Forward: 'The most important revision to make is [one specific, actionable revision]. To do this, try [specific strategy].' Crucially: ONE primary revision target per piece of feedback. Research shows that students improve more when given one clear target rather than multiple simultaneous targets. Component 4 — Required Revision: Feedback is not the end — revision is expected. Students must complete a revision within 3-5 days addressing the teacher's one primary target. Students submit: revised passage; a two-sentence explanation of what they changed and why. Teacher reviews: 'Did the revision address the target? Did the revision improve the work?' Brief acknowledgment of revision (2-3 sentences) rather than re-grading. Component 5 — Peer Feedback Protocol (Weekly, 10 minutes): Students exchange in-progress work with a designated partner. Using the 'I notice, I wonder, I suggest' protocol: I notice: a specific strength I observe in your writing. I wonder: a question I have as a reader (places of confusion; moments where I wanted more). I suggest: one specific improvement that would strengthen the piece. Peer feedback develops the capacity to see others' work the way a teacher does — and in doing so, develops the capacity to see one's own work more critically. Component 6 — Assessment Conference (Once Per Unit): 10-15 minute individual assessment conference with each student. Teacher and student review the student's portfolio of work from the unit. Questions: 'Which piece are you most proud of? What makes it strong?' 'What is the area where you've grown most this unit?' 'What is your learning goal for next unit?' The conference develops self-regulatory capacity and creates a genuine relationship between assessment and learning planning. Full system with: success criteria templates for each major essay type; self-assessment forms; teacher feedback template; peer feedback protocol; assessment conference guide; rubric for the entire system that describes how assessment information informs grading."

Backward Design Unit Planning

"Design a complete backward design unit plan for Grade 4 Science — 'Where Does Rain Come From? Understanding the Water Cycle' — grounded in Wiggins and McTighe's Understanding by Design three-stage framework and six facets of understanding, and aligned with NGSS Earth's Systems standards. Stage 1 — Desired Results: Standards Addressed: 5-ESS2-1 (Earth's Systems: describe the movement of water on Earth through the water cycle); 5-ESS2-2 (Earth's Systems: describe and graph the amounts and percentages of water in various reservoirs). Big Idea: Water moves continuously through Earth's systems — from the ocean to the atmosphere to the land and back again — in a cycle driven by solar energy and gravity. Essential Questions (open, not answerable with a single correct fact): 'Where does the water in our faucet actually come from — and where does it go?' 'Why does it rain more in some places than others?' 'What would happen if the water cycle slowed down or stopped?' Understanding Goals (by the end of this unit, students will understand that...): Water changes state (liquid → gas → liquid) as it moves through the water cycle. Energy from the sun drives evaporation; gravity drives precipitation and runoff. The water on Earth is the same water that has always been on Earth — recycled over billions of years. Knowledge Goals (students will know): The names of the main stages of the water cycle (evaporation; condensation; precipitation; collection/runoff). The names of the main water reservoirs (oceans; glaciers and ice; groundwater; lakes and rivers; atmosphere). The proportion of Earth's water in each reservoir. Skill Goals (students will be able to): Model the water cycle diagrammatically and explain the movement of water at each stage. Interpret weather maps that show precipitation patterns. Stage 2 — Evidence of Learning: Performance Task (Authentic Assessment): Students receive the following scenario: 'You are a water droplet. You start in the Pacific Ocean. Describe your journey over the next 100 years — where do you go, how do you get there, what form do you take at each stage, what energy is involved?' Students write a first-person narrative from the water droplet's perspective (length: 500-800 words) that accurately describes the water cycle, including at least one complete cycle and at least three different states and reservoirs. The narrative is evaluated on: Scientific accuracy (are the stages and transitions accurate?); Completeness (does it include all major stages?); Energy (does it accurately describe what drives each transition?); Explanation (does the droplet explain WHY each transition happens, not just what happens?). Other Evidence: Water cycle diagram labeled with stages, energy inputs, and direction of movement. Exit ticket: 'Draw a simple sketch showing where you think the water in today's rain came from 100 years ago.' (Reveals whether students understand the recycling concept.) Formative observation notes during investigation activities. Stage 3 — Learning Plan (WHERETO Framework): Where (where is the unit going?): Hook with the driving question 'Where does the water in our faucet come from?' — trace it backward. Hook with dramatic statistic: 97% of Earth's water is in the oceans; 2% is frozen in glaciers; only 1% is available as fresh liquid water. Hook with: 'You might be drinking water that a dinosaur drank 65 million years ago — explain how that's possible.' Hook with: bring in a glass of water and ask students to construct the history of that specific water. Explore: Evaporation investigation (measure how much water evaporates from an open bowl vs. covered bowl in warm vs. cool locations). Condensation observation (cold glass on a humid day; where does the water on the outside come from?). Precipitation modeling (cups of water above warm water producing clouds and rain). Water reservoir data analysis (graph the percentages; calculate how much of Earth's water each reservoir holds). Read: Water cycle explanation text at two levels (differentiated); science trade books ('A Drop Around the World'; 'Water Dance'). Revise: Peer review of water droplet narratives. Exhibit: Final water droplet narratives shared in a 'Water Stories' gallery walk. Full unit with: complete lesson sequence; investigation protocols; data recording sheets; water droplet narrative rubric; exit ticket sequence; differentiated reading materials; home extension activities."

Standards-Based Grading Implementation Guide

"Design a standards-based grading implementation guide for a Grade 6-8 science department — 'Grading for Mastery: A Transition Guide to Standards-Based Assessment' — grounded in Guskey's SBG principles, Black and Wiliam's assessment for learning research, and Hattie's effect size evidence for feedback. This guide supports teachers in transitioning from traditional percentage-based grading to a system that accurately reflects students' mastery of science standards. Phase 1 — Understanding Why: The Case for Change (Professional Development, Day 1). Problem identification: What does a 72% actually mean? (Exercise: Calculate the grade of a student with these scores: 45%, 75%, 85%, 90%, 95% on five assessments throughout the semester. Traditional average: 78%. But what does this tell us about what the student knows NOW? The student's most recent three scores are 85, 90, 95 — clearly developing toward mastery — but the 45% in September is dragging the average down.) Validity problem: Do our grades measure what we say they measure? List everything that goes into a student's grade in your classroom. How much is academic mastery? How much is behavior (on-time submission; participation; effort)? Can a student earn an A through compliance without mastery? Can a student fail through non-compliance despite mastery? Phase 2 — Designing the System: What We Grade and How (Professional Development, Days 2-3). Step 1: Identify the essential learning standards. From the NGSS standards for each grade level, identify 8-12 essential standards per semester — the standards without which students cannot succeed in subsequent learning. These are what grades will report. Step 2: Design proficiency scales. For each essential standard, develop a 4-level proficiency scale: Level 4 (Advanced): Student can apply the standard in novel contexts; make connections to other standards; explain the standard in depth. Level 3 (Proficient): Student meets the standard; demonstrates the understanding or skill described by the standard. Level 2 (Developing): Student demonstrates partial understanding; needs additional support or practice. Level 1 (Beginning): Student demonstrates minimal understanding; needs significant retestation and support. Step 3: Separate academic and compliance grades. Academic grade: based solely on demonstrated mastery of essential standards. Learning habits grade (separate): completed assignments; participation; collaboration; meets deadlines. These are reported separately so that each communicates meaningful information. Step 4: Determine how to calculate the academic grade. Policy: most recent evidence of mastery (not average). If a student demonstrates Level 2 on a standard in October and Level 4 in December, the grade reports Level 4 — the most recent, most accurate evidence of mastery. This means reassessment opportunities are available to all students. Phase 3 — Communicating with Students and Families (Professional Development, Day 4). Sample communication to families: 'This semester, your student's grade will report how well they have mastered the essential science standards for Grade 7. Instead of a single percentage, you will see a score on each standard (1-4, where 3 is proficient). Your student can improve any score by demonstrating growth through additional learning and reassessment. The grade reflects what your student knows and can do, not whether they turned in every assignment on time.' Student communication: 'Your grade tells you where you are right now on each science standard. A score of 2 doesn't mean you've failed — it means you're still developing this skill and we'll work together to get you to 3.' Frequently asked questions from families: Will this hurt college admissions? (Most gradebooks convert SBG to traditional scales for transcripts; mastery-based grades typically reflect higher genuine achievement than average-based grades.) What if my student scored 4 in September but 2 in December? (The most recent evidence is used, so yes, grades can go down — but this is rare and always worth discussing.) Phase 4 — Reassessment Policy and Implementation. Reassessment protocol: Any student who wants to improve a standard score may reassess after: meeting with the teacher to identify what they need to learn; completing additional practice on the standard; the teacher certifying readiness. Reassessment is not automatic re-taking of the same test — it requires evidence of additional learning. Full guide with: PD facilitation guide; proficiency scale templates for each grade level's essential standards; family communication letter; reassessment request form; grading software setup guide; FAQ for teachers, students, and families."

Classroom Scenario: Thanh's Assessment Practice in Cocos (Keeling) Islands

Thanh Nguyen-Islam is a primary school teacher on the Cocos (Keeling) Islands — an Australian territory in the Indian Ocean approximately 2,750 kilometers northwest of Perth, Western Australia, and 970 kilometers southwest of Christmas Island. The Cocos Islands consists of two atolls: Direction Island and Home Island, containing a total of 27 coral islets with a combined land area of approximately 14 square kilometers and a permanent population of approximately 600 people.

The Cocos Malay Community: The population of the Cocos Islands is primarily composed of the Cocos Malay community — descendants of Malay workers brought to the islands by the Clunies-Ross family (a Scottish dynasty that ruled the islands as a private fiefdom from 1827 until Australian sovereignty was established in 1955 and full territorial control was transferred in 1978). The Cocos Malays maintained their own language, religion (Islam), cultural traditions, and community structures through more than a century of effective isolation under the Clunies-Ross estate. The community voted in 1984 to integrate with Australia (rather than independence or free association), and Cocos Malays are Australian citizens. The Cocos Malay language — a variety of Malay with distinct features reflecting the community's history — is the first language of most residents.

Educational Context: The island's school serves the approximately 80-100 school-age children on the islands, with students from the Cocos Malay community making up the majority. The school follows the Australian national curriculum, delivered in English, for a community whose first language is Cocos Malay. The assessment challenges are significant: standardized Australian assessments (including NAPLAN) are conducted in English, a second language for most students; the small school population means that within-classroom diversity in learning needs is substantial; and the school's isolation limits access to specialist assessment support.

Thanh's Assessment Approach: Thanh uses a formative-first assessment philosophy grounded in Black and Wiliam's principles: assessment information is collected primarily for the purpose of informing teaching decisions and providing feedback to students, not for summative grading. He uses frequent low-stakes formative checks (exit tickets; observation; questioning) to track each student's developing understanding; provides comment-only feedback on most drafts (reserving grades for final summative assessments); and builds student self-assessment capacity by regularly asking students to evaluate their own understanding before seeing the teacher's assessment. EduGenius (edugenius.app) helps Thanh generate formative assessment tools adapted to the Cocos Malay community context — including bilingual success criteria that students can use in either Cocos Malay or English; visual rubrics that reduce English language demands while maintaining academic rigor; and feedback templates that ask English questions at appropriately scaffolded language levels.

Internal Resources

• Best AI for Differentiated Instruction and Universal Design for Learning
• Best AI for Project-Based Learning
• Best AI for Growth Mindset and Motivation in Education
• Best AI for Math Problem Solving and Mathematical Reasoning
• Best AI for Special Education and Inclusive Classrooms