How to Strengthen My Assessment with GenAI
You are viewing this module without an account. Sign in to save progress or reflections.
Module Description
This module explores how Generative Artificial Intelligence (GenAI) can expand the way we approach assessment — from a limited, occasional activity into a more continuous and responsive pedagogical process.
It focuses on a central challenge in teaching: visibility. In real classrooms, time and attention are limited, making it difficult to access the thinking processes of all students while learning is taking place.
Through a structured framework of four core assessment needs, you will explore how GenAI can support different stages of assessment — from clarifying what to assess, to designing tools, supporting learning in real time, and identifying patterns across the class.
Throughout the module, GenAI is presented as a thinking partner within the broader field of Artificial Intelligence (AI). It can help expand what becomes visible in learning, while professional judgment remains human-led.
Learning goals
-
Understand how GenAI can support different stages of assessment.
-
Recognise the challenge of visibility in assessment and how it affects teaching decisions.
-
Identify how assessment can focus on students’ thinking processes, not only final outcomes.
-
Explore how GenAI can help reveal patterns in student responses and support more informed decisions.
Introduction
Assessment is not just about results — it is about making learning visible.
Our challenge is to expand and deepen the assessment lens without increasing teacher load.
In a real classroom, we constantly face limits. We can listen closely to only a few students at a time, we often assess final outcomes more than learning processes, and much of our assessment happens after the lesson rather than while learning is still unfolding.
Many assessment practices are shaped by what was possible and available, not necessarily by what was pedagogically ideal.
This module starts from a different question: What becomes possible for assessment when GenAI is available as a pedagogical assistant?
Please watch the short video below before continuing.
Tip: You can turn on subtitles by clicking the CC icon in the video player. To translate subtitles, click the settings (⚙️) icon → Subtitles → Auto-translate, and choose your preferred language.
The framework guiding this module: four assessment needs
This module is structured around four core pedagogical assessment needs. Each unit focuses on one need and explores how GenAI can open up new assessment possibilities that were previously difficult or impossible to realize.
What do I want to assess and why?
From generic goals → to precise, context-specific planning
How do I build the tools?
From manual work → to higher-quality tools, faster
How do I assess while learning happens?
From end-of-unit tests → to formative insight
How do I see patterns across the class?
From reading one-by-one → to seeing the big picture
UNIT 1: Planning Assessment Goals
After exploring the role of assessment in teaching, the next step is to clarify a fundamental question: What exactly do we want to assess — and why?
In this unit, we shift our focus from general intentions to precise planning. Before designing any task, rubric, or test, we pause to define what kind of learning we want to notice and make visible.
In this unit, we will go through a short process:
▸ Reflect on what we usually assess — and why
▸ Examine the difference between outcomes and thinking processes
▸ Learn how to define assessment goals in a more focused and meaningful way
What do I want to assess and why?
From: Generic assessment goals
To: Precise, context-specific evaluation planning
Planning assessment means deciding what kind of learning we want to be able to notice. Without clarity, we tend to focus on what is easiest to assess: finished products, correct answers, visible participation.
How Gen-AI Helps
Gen-AI supports planning by helping us:
▸ Articulate learning goals in observable terms
▸ Identify processes worth assessing (reasoning, strategies, reflection)
▸ Clarify what evidence could reveal those processes
This allows us to plan assessments that target thinking, not only results.
Quick Recipe
Before you create any assessment, clarify what you want to see:
Tell Gen-AI your topic and grade level
Ask: "What thinking processes should I assess, not just final answers?"
Ask: "What misconceptions might students have?"
Use the response to focus your assessment on what matters
Ready-to-copy prompt
I am a [subject] teacher planning an assessment on [topic] for [grade level].
Help me identify:
- The key concepts I should assess
- The thinking processes worth noticing (not just correct answers)
- Common misconceptions students might have
Suggest one way students could help define what success looks like.
Classroom example
Marta, a biology teacher in Bratislava, Slovakia
Marta teaches cell biology to 15-year-olds. She used to assess with a standard test: label the cell parts, explain mitosis. Students memorized and forgot.
What she did differently: Before creating her next assessment, she asked Gen-AI: "What thinking processes should I assess when teaching cell division, beyond memorization?"
What she discovered: The AI suggested she assess students' ability to predict what happens when the process goes wrong - not just describe the steps. It also flagged a common misconception: students often think cells "decide" to divide, as if they have intention.
What changed: Her new assessment asks students to explain what might happen if a cell skipped a phase of mitosis. She also added a reflection question: "What surprised you about how cells work?"
Time invested: 10 minutes with the chatbot before designing the task.
Common challenges and how to address them
| Challenge | What to do |
| Gen-AI suggests too many things to assess | Pick 2-3 that matter most for your lesson. You don't need to assess everything. |
The suggestions feel too generic | Add more context: your students' level, what they already know, what you noticed in previous lessons. |
You're not sure if the AI's suggestions make sense | Trust your experience, or consult with a more experienced teacher. Use the AI as a thinking partner, not as the final word. |
In more detail (optional)
The preparation stage focuses on defining what should be assessed and why. Before any task, rubric, or tool is designed, teachers must clarify learning goals, target competencies, and success criteria.
At this stage, assessment decisions shape the purpose of learning itself: which skills matter, which processes deserve attention, and how assessment aligns with curriculum standards and principles of self-regulated learning (SRL).
Generative AI supports this stage by acting as a consultative thinking partner. Rather than producing ready-made assessments, it helps teachers:
-
Translate broad curricular goals into assessable competencies
-
Surface implicit assumptions
-
Consider how assessment can capture learning processes - not only final products
Pedagogical Roles:
-
Teacher – reflective planner defining learning purpose and assessment intent
-
Students – informed learners aware of goals and expectations
-
GenAI – consultative partner supporting alignment between objectives and evidence
Sample Prompts
The examples below are provided for your reference. They illustrate how prompts can be used in different disciplines to support planning assessment goals.
Below are two sample prompts from different disciplines. You are welcome to explore, adapt, and use them freely — this is not a required task, but an opportunity to experiment.
For STEM teachers:
For Humanities teachers:
UNIT 2: Design Assesment Tools and Techniques
After defining what we want to assess, the next step is to clarify a fundamental question: How do we design tools that actually make that learning visible?
In this unit, we shift our focus from planning to design. Before creating any task, rubric, or tool, we consider how different design choices shape what becomes visible in student learning.
In this unit, we will go through a short process:
- Examine how different assessment tools highlight different aspects of learning
- Explore the difference between evaluating products and revealing thinking processes
- Learn how to design tools that support reflection, reasoning, and explanation
How do I build the assessment tools?
From: Labor-intensive manual rubric design
To: High-quality, personalized assessment criteria
Design determines who gets seen and how. Some tools highlight only final products. Others allow insight into process, reflection, and decision-making.
How Gen-AI Helps
Gen-AI supports design by:
▸ Creating rubrics that describe thinking, not just performance
▸ Generating prompts that invite explanation and reflection
▸ Adapting tools so more learners can engage meaningfully
Design becomes a way to open assessment, not narrow it.
Quick Recipe
Build assessment tools that show thinking, not just answers:
-
Tell Gen-AI what you're assessing
-
Ask for a rubric with 4-5 criteria that describe thinking (not just performance)
-
Ask for one reflection question per criterion
-
Review and adjust to fit your class
Ready-to-copy prompt
I am designing an assessment on [topic] for [grade level].
Create a rubric with 4-5 criteria that evaluate:
- Not just the final answer, but the thinking process
- How students explain their reasoning
- How they use evidence
For each criterion, add one reflection question students can ask themselves.
Classroom Example
Tomáš, a history teacher in Brno, Czech Republic
Tomáš teaches about World War I to 16-year-olds. His old rubric had three levels: "Excellent / Good / Needs Improvement" - but students never understood what made something "excellent."
What he did: He asked Gen-AI to help create a rubric that describes what good historical thinking looks like - not just what a good grade looks like.
The result: His new rubric has criteria like:
-
Uses multiple sources to support an argument (not just "uses sources")
-
Considers why people at the time might have thought differently (not just "shows understanding")
-
Acknowledges what we don't know or can't prove (not just "accurate")
Each criterion has a reflection prompt: "What evidence did you find most convincing? Why?"
What students said: "Now I know what I'm supposed to be thinking about, not just what to write."
Time invested: 15 minutes to create, then small adjustments.
Common challenges and how to address them
Challenge | What to do |
The rubric is too complicated | Ask for "a simpler version with only 3-4 criteria" |
The language doesn't fit your students | Ask to "rewrite this for [age] students in clear, simple language" |
It doesn't match your curriculum | Add specific standards or requirements to your prompt |
In more detail (optional)
The design stage shapes not only how assessment looks, but also who gets seen within it and which aspects of learning are made visible.
Some assessment tools focus primarily on final products, while others allow insight into students' thinking processes, reflection, and decision-making. Design choices play a critical role in determining whether assessment narrows or expands our understanding of student learning.
At this stage, GenAI supports the creation of open, reflective, and inclusive assessment tools. It helps:
-
Design rubrics that describe thinking rather than only performance
-
Generate prompts that invite explanation and reflection
-
Adapt assessment tools so more learners can engage meaningfully
In this sense, design becomes a way to open assessment, not to limit it.
Pedagogical Roles:
-
Teacher – designer and model of reflective, goal-oriented assessment
-
Students – active users of assessment tools for self-monitoring and peer feedback
-
GenAI – creative co-designer aligning criteria with intended learning outcomes
Sample Prompts
The examples below are provided for your reference. They illustrate how prompts can be used in different disciplines to support the design of assessment tools and techniques.
Below are two sample prompts from different disciplines. You are welcome to explore, adapt, and use them freely — this is not a required task, but an opportunity to experiment.
For STEM teachers:
For Humanities teachers:
UNIT 3: Classroom Implementation
After planning what to assess and designing the tools, the next question is: How do we assess while learning is actually happening?
In this unit, we move into the classroom itself. We focus on how assessment can take place during the learning process — while students are thinking, explaining, revising, and making decisions.
In this unit, we will go through a short process:
- Explore what kinds of learning are visible only in the moment
- Understand the difference between summative and formative assessment in practice
- Learn how to capture thinking as it unfolds — not only after it is completed
How do I assess while learning happens?
From: Assessment of learning (summative)
To: Assessment as learning (formative)
Some learning is visible only in the moment: when students explain, revise, hesitate, or change direction. Traditionally, we could capture only fragments of this.
How Gen-AI Helps*
Gen-AI allows:
▸ Students to externalize thinking through guided prompts
▸ Self- and peer-assessment during work
▸ Reflective questioning without stopping the lesson
Assessment becomes distributed, not centralized.
* Suitable only for students allowed working independently with GenAI
Quick Recipe
Use AI as a reflective partner during the activity:
-
Prepare an AI prompt that asks guiding questions aligned with your learning goal
-
During the activity, students interact with the AI to articulate their thinking
-
The AI asks questions like: "What evidence supports this?" or "Why did you choose this approach?"
-
You observe the interactions and provide formative feedback during or after
Ready-to-copy prompt:
You are a reflective learning partner for a student working on [topic/task].
Your role is NOT to give answers. Instead, ask guiding questions that help the student:
- Articulate their thinking process
- Explain the reasoning behind their choices
- Consider evidence that supports or challenges their ideas
- Reflect on alternative approaches
Ask one question at a time. Wait for the student to respond before asking the next question.
Example questions you might ask:
"What made you choose this approach?"
"What evidence supports your thinking?"
"How would you explain this to someone who disagrees?"
"What would change if [alternative scenario]?"
Start by asking the student to describe what they are working on.
Classroom example
Kristaps, a science teacher in Riga, Latvia 🇱🇻Kristaps teaches 15-year-old students. In his class, students were designing an experiment about plant growth. He wanted to see their scientific thinking process - not just the final experiment plan.
What he did differently:While students planned their experiment, each pair worked with an AI “lab partner” that asked reflective questions in real time as they developed their design.
Sample AI–student interaction:AI: “You chose to change the amount of water. Why this variable?” Student: “Because it’s easy to measure.” AI: “That’s practical. What do you predict will happen, and why?” Student: “More water = bigger plant… but wait, can there be too much water?”
What Kristaps discovered:
- He could see students’ reasoning in real time, not just their final answers.
- Quiet students who rarely spoke in class showed sophisticated thinking.
- He identified misconceptions while there was still time to address them.
Time invested:10 minutes to prepare the AI prompt. During class, he walked around observing interactions.
Common challenges and how to address them
Challenge | What to do |
Students ask the AI for answers instead of thinking | Configure the AI prompt to explicitly refuse giving answers. Model the expected interaction first with the whole class. |
The AI asks irrelevant or confusing questions | Be more specific in your prompt about the topic and the type of questions you want. Test it yourself first. |
You can't see all the interactions | Ask students to save or screenshot key moments from their conversation. Review a sample after class. |
Students feel uncomfortable "talking" to AI | Frame it as a thinking tool, not a conversation. |
In more detail (optional)
The implementation stage brings assessment into the learning moment itself. Some of the most meaningful evidence of learning appears while students are thinking, explaining, revising ideas, hesitating, or changing direction. Traditionally, teachers could capture only fragments of this process.
GenAI enables assessment during the learning encounter by serving as a reflective partner for students. Instead of waiting for the end of the unit, students interact with AI that prompts them to articulate their thinking in real-time.
How AI as a Reflective Partner Works:-
The AI asks guiding questions: "What evidence supports your claim?", "Why did you choose this strategy?", "How might another interpretation change your conclusion?"
-
Students externalize their thinking through dialogue with the AI
-
The teacher observes interactions and gathers evidence of reasoning, adaptability, and metacognitive awareness
-
Formative feedback can be provided during or immediately after the activity
Pedagogical Roles:
-
Teacher - Facilitator and observer of thinking-in-action, using evidence for formative guidance
-
Student - Active and reflective learners articulating and refining their thinking during learning
-
GenAI - Adaptive, context-sensitive reflective partner acting as a "cognitive mirror" for learners
The key shift: Assessment moves from something that happens after learning to something that happens during learning. The AI helps make student thinking visible in the moment, so teachers can respond while it still matters.
Sample Prompts
The examples below are provided for your reference. They illustrate how prompts can support assessment during the learning process, while students are still thinking, explaining, and revising their ideas.
Below are sample prompts you can explore and adapt to your own context. You are welcome to use them freely — this is not a required task, but an opportunity to experiment with real-time assessment.
For STEM teachers:
For Humanities teachers:
UNIT 4: Gaining a Learning Overview
After supporting learning in real time, the next question becomes: How do we step back and understand what is happening across the whole class?
In this unit, we shift from individual observations to a broader perspective. We focus on how to identify patterns in student thinking, while still recognizing individual differences.
In this unit, we will go through a short process:
- Explore how multiple student responses can reveal class-level patterns
- Understand the challenge of balancing overview with attention to individual learners
- Learn how to turn scattered evidence into meaningful insights for teaching
How do I see patterns across the class?
From: Reading individual papers one-by-one
To: Seeing the "big picture" of class understanding in real-time
When more learning becomes visible, the challenge shifts: How to see patterns without losing nuance.
How Gen-AI Helps
Gen-AI supports overview by:
▸ Summarizing reflections and observations
▸ Identifying recurring themes and differences
▸ Helping us notice learners who might otherwise remain invisible
This enables more informed pedagogical decisions.
Quick Recipe
See the whole class, not just individual papers:
-
Collect student responses (reflections, exit tickets, short answers)
-
Remove names and paste all responses into Gen-AI
-
Ask for patterns: "What do most students understand? Where is there confusion?"
-
Use insights to plan your next steps
Ready-to-copy prompt
I am a teacher analyzing student responses after a lesson on [topic].
The responses below are anonymous. [paste all responses] Analyze to identify:- Recurring themes or ideas (what do many students understand?)
- Common difficulties or misconceptions
- Signs of progress
- Any surprising or unique responses worth noting
Classroom Example
Jakub, a physics teacher in Kraków, Poland
Jakub teaches physics to 17-year-olds. After a unit on electromagnetic waves, he felt unsure whether students truly understood or just memorized formulas.
What he did: He asked students two questions:
-
"Explain electromagnetic waves to a 10-year-old."
-
"What still confuses you about this topic?"
He collected 31 responses and pasted them (without names) into a chatbot.
What the AI showed him:
-
Most students could explain the basics, but used "wave" and "radiation" interchangeably (a sign of partial understanding)
-
8 students expressed confusion about why different wavelengths behave differently
-
4 students had creative analogies that showed deep understanding - students he hadn't noticed before
What he did next: He created a quick clarification activity on the wave/radiation distinction, and asked one of the "invisible" students to share their analogy with the class.
Time invested: 5 minutes to analyze, saved him from re-teaching the entire unit.Common challenges and how to address them
Challenge | What to do |
The summary is too general | Ask: "Give me specific examples from student responses to support each point" |
You're worried about privacy | Always remove names; use only excerpts; don't paste full essays |
The AI misses something important you noticed | Add your own observation: "I also noticed that several students mentioned X. Include that in the analysis." |
You don't trust the AI's interpretation | Use it as a starting point, not final truth. You know your students better. |
In more detail (optional)
As more learning becomes visible through reflective tasks and real-time interactions, the challenge shifts from collecting evidence to making sense of it without losing nuance.
Teachers need ways to see patterns across learners while still attending to individual voices, experiences, and needs.
Generative AI supports this stage by helping teachers move from scattered reflections and observations to a coherent learning overview. It can:
- Summarize student reflections
- Identify recurring themes and differences
- Surface learners or perspectives that might otherwise remain invisible
This enables more informed, responsive pedagogical decisions - about feedback, pacing, differentiation, and next instructional steps.
Pedagogical Roles:
- Teacher – interpreter of learning data and facilitator of collective reflection
- Students – active participants in self-assessment and meaning-making
- GenAI – analytical partner transforming qualitative data into pedagogical insight
Sample prompts
The examples below are provided for your reference. They illustrate how prompts can help identify patterns across multiple student responses, while still recognizing individual differences.
Below are sample prompts you can explore and adapt to your own context. You are welcome to use them freely — this is not a required task, but an opportunity to experiment with gaining a learning overview.
For STEM teachers:
For Humanities teachers:
UNIT 5: Ethics and Responsible Use in Assessment
After exploring how Gen-AI can support assessment planning, design, implementation, and analysis, an essential question remains: How do we use these tools responsibly, ethically, and in ways that protect our students?
In this unit, we focus on the principles that guide thoughtful and professional use of Gen-AI in assessment. We move beyond “what works” to consider what is appropriate, fair, and educationally meaningful.
In this unit, we will go through a short process:
- Reflect on key ethical risks when using AI in assessment
- Understand the teacher’s role as the final decision-maker
- Learn how to maintain transparency, privacy, and pedagogical responsibility
Quick Checklist (Before using Gen-AI for assessment)
Ask yourself:
-
Have I removed student names and identifying information?
-
Am I using AI to suggest and summarize - not to grade or judge?
-
Will I review the AI output critically before acting on it?
-
Do my students understand how and why AI is being used?
-
Am I still the one making the final decisions?
Preparing Students for Critical Use
When students work with Gen-AI, we don't assume they already know how to use it well. We prepare them explicitly.
In practice, this means we:
- Explain that AI can be helpful but wrong
- Ask students to check AI questions or feedback against their own understanding
- Assess how they respond to AI prompts, not the AI output itself
For example, instead of assessing the final text, we look at:
- How students revise after AI questioning
- Whether their reasoning becomes clearer
- Whether they can explain why they accepted or rejected AI suggestions
A Simple Rule
If Gen-AI:
- Hides thinking → we redesign the task
- Replaces judgment → we step back
- Reduces transparency → we pause and explain
Ethical use is not about limiting innovation. It's about keeping assessment fair, meaningful, and focused on learning - even when new tools are involved.
Unit 6: Assessment in Action
- This unit focuses on assessment in action.
- You will explore how GenAI can support assessment thinking through a short consultation simulator.
- You will then hear from teachers who integrated GenAI into their own assessment practices.
- The goal is to move from understanding the framework to applying it in real teaching situations.
Assessment Consultation Simulator
GenAI in Assessment: Teachers’ Voices
Now that you have explored the assessment framework through the simulator, watch how teachers describe their own experiences using GenAI to support assessment.
Please watch both videos before continuing:
Tip: You can turn on subtitles by clicking the CC icon in the video player. To translate subtitles, click the settings (⚙️) icon → Subtitles → Auto-translate, and choose your preferred language.
Summary
Module Summary
In this module, we explored how Generative Artificial Intelligence (GenAI) can support assessment by helping teachers see learning more clearly.
We focused on one central challenge: visibility. In real classrooms, it is difficult to notice every student’s thinking process in real time. GenAI can help expand the assessment lens, but it does not replace professional judgment.
Throughout the module, we worked with four core assessment needs:
Clarifying what we want to assess and why.
Creating tools that reveal thinking, not only final products.
Supporting assessment while learning is still happening.
Identifying patterns across student responses while keeping teacher interpretation central.
The key idea is that GenAI can serve as a thinking partner: helping to plan, design, ask better questions, and notice patterns. The teacher remains the person who interprets the information, makes decisions, and ensures that assessment stays ethical, meaningful, and focused on learning.