Interactive learning material · M5-T1-MLC3

Reflective Teaching with Artificial Intelligence

From experience to learning, from learning to action

Author: Mária HartyányiiTStudy Hungary Training and Research Centre
Reviewer: Annamária KacsuriTStudy Hungary Training and Research Centre

A teacher reflects on a lesson and takes notes. — The reflective teacher · Source: ChatGPT

Learning objective

The aim of this learning material is to show how artificial intelligence can support teachers in thinking through their classroom experiences, drawing lessons from them and planning their next lessons, so that self-reflection becomes part of their everyday practice and continuous professional development.

By the end of this learning material, you will be able to

Reflect on a teaching situation using Gibbs’ six-step reflective cycle.
Compare notes, learner feedback and other evidence with AI support.
Use AI as a questioning partner while retaining decision-making and responsibility.
Build a small, regular reflection routine and a practical action step.

Learning from experience

Why does reflection so often remain a good intention?

Teaching is one of the most complex professions: subject expertise and pedagogical sensitivity must be combined with excellent presentation, observation and organisational skills. Every lesson is a unique, unrepeatable story that can hold surprises even with the most careful planning. Learners’ motivation, their physical and emotional state, and the mood of the class on the day all influence how the lesson unfolds and how effective it is.

Every lesson offers the teacher new experience and can become a source of continuous development – provided that it is followed by analysis and evaluation, that is, by teacher reflection. Without it, to borrow Stephen D. Brookfield’s idea, ten years of teaching may be nothing more than one year’s distorted experience repeated ten times.

Amid the daily teaching routine, frequent overload and lack of time, it is hard to devote regular time and attention to reflection. This is where artificial intelligence can help.

Routine alone is not enough

Most teachers know that it is worth thinking back over the lessons they have taught. In practice, however, the curriculum repeats itself year after year, tried-and-tested tasks return on their own, and routine is comfortable. Educational researchers warn that experience without reflection remains narrow and yields superficial knowledge at best.

Why is this especially important in VET?

This is particularly important in vocational education and training. Occupations, technologies and workplace expectations change quickly, some learners study in workshops or at companies, and the teacher’s work is divided between theory lessons, practical sessions and cooperation with company partners. What worked well last year may no longer be enough this year.

The steps of reflective teaching

Gibbs’ six-step reflective cycle

Teacher reflection is a cyclical process made up of several steps. Graham Gibbs’ model helps us organise our experience and arrive at a next step that we can try out.

1. Description
“What happened in the lesson?”
Facts: what happened, what did we do, and what was the outcome?
2. Feelings
“What did I feel and think?”
We name the feelings and thoughts associated with the situation separately.
3. Evaluation
“What was good and what was not?”
We consider both what worked well and what caused difficulties.
4. Analysis
“Why did it turn out this way?”
We look for causes, connections and other possible explanations.
5. Conclusion
“What have I learnt from it?”
We identify what we have learnt and what we could have done differently.
6. Action plan
“What will I do next time?”
We choose one small, concrete change that we can try out.

Six steps in a cycle: description, feelings, evaluation, analysis, conclusion and action plan. — Gibbs’ reflective cycle: description, feelings, evaluation, analysis, conclusion and action plan

A detailed description of the model: University of Edinburgh – Gibbs’ reflective cycle.

What we could not do before

Time and an available conversation partner

AI can save time when organising notes and other sources, and can be available as a questioning partner when there is no opportunity for a shared discussion.

Notes after the lesson

Until now: Notes after the lesson were skipped or consisted of just a few words.

Now, with AI: A two-minute voice note on the way home. AI transcribes it, organises it and asks back about what was left out.

Looking back at the lesson

Until now: We have no direct view of our own lessons; even if a video or audio recording is made, watching or listening to it takes hours.

Now, with AI: AI produces a summary of the media sources within minutes: how much the teacher talked, what questions they asked, and where learners joined in.

Discussion with colleagues

Until now: There is rarely time for a longer discussion with colleagues or with in-company trainers coming from external companies.

Now, with AI: AI prepares the conversation: it compiles the lesson observation materials and suggests concrete questions for the discussion.

Learner feedback

Until now: Learner feedback stays in the drawer; there is no time for really detailed processing.

Now, with AI: AI groups the anonymous responses by theme, prepares an analysis and formulates conclusions.

Recurring learning patterns

Until now: Only the average is visible in the results; recurring patterns in the learning process remain hidden.

Now, with AI: AI helps to spot which task type or vocational competence regularly causes difficulties.

Our own teaching experiment

Until now: Designing our own “experiment” to improve the quality of teaching and learning is too big an undertaking alongside daily work.

Now, with AI: AI helps to design the experiment, formulate the research question, prepare a short questionnaire and summarise the results.

The quality of the analysis depends on the sources, the tool’s capabilities and verification. A quick summary is not yet a verified conclusion.

Different sources together

When AI works with text, sound and images

Multimodal AI can process different kinds of sources: text, sound, images and, with a suitable tool, video. Reflection can draw on voice notes, classroom observations, photos of learners’ work, anonymous feedback or aggregated results.

Until now, teachers had to summarise these one by one, relying on their own memory. AI can help us look for connections and contradictions.

For example, the teacher’s notes may describe an engaged class, while the transcript of the recording suggests that the teacher did most of the talking. Or the task learners said they did not understand may also have produced the weakest test results.

Notes, aggregated results, learner feedback and observations feed into a dialogue with AI. — Notes, anonymous learner feedback, aggregated results and observations form the basis of the reflective dialogue

VET offers many different kinds of evidence

A short recording of a practical demonstration, a photo of a workpiece and the learner’s oral explanation together can reveal more than any of them alone. Before processing them, check the institution’s conditions for their use.

Example prompt · Comparing sources

Where do the sources confirm or contradict each other?

Analysing several sources together

“I am attaching an audio recording of part of a lesson, a photo of the learners’ work and the learners’ anonymous feedback. Compare them and show me where they confirm and where they contradict each other. Do not evaluate the lesson and do not give a score. Ask me three questions that help me understand what happened and what would be worth trying next time.”

1

Ask for comparison

Ask about agreements and contradictions, rather than separate summaries.

2

Ask for questions

The prompt asks for three questions to support thinking, without judgement or scoring.

3

Check the source

Check what the AI bases its claims on and what the source does not show.

Use this prompt only with sources you are permitted to process and that the chosen tool can actually handle.

Five minutes at the end of the day

Building reflection into daily work

AI really helps when reflection becomes a regular, daily routine. There is no need to analyse every lesson in detail; a simple rhythm is enough:

1

Five minutes a day

After an important lesson, record a short voice note or write a few sentences about what worked, what did not, and what you observed in the learners. AI organises the thoughts and asks one or two questions.

2

One piece of evidence a week

Add a few anonymous exit cards, a photo of learners’ work handled appropriately, or an authorised lesson extract to your notes, and compare them.

3

One conversation a month

Discuss the month’s experiences with a colleague or the in-company trainer, based on a summary prepared by AI.

Turn a pattern into a practical step

Small, regular steps are worth more than an occasional big analysis. Choose a recurring situation, look for alternative approaches, and try one of them out.

Recurring patterns lead to alternative approaches, a chosen action step and development. — Recurring patterns give rise to alternative approaches; we choose one of them and try it out in small steps

Example prompt · A questioning partner

A critical friend at the end of the day

A self-reflective dialogue about a lesson

“Act as a critical friend to a vocational teacher. Guide me through the steps of the reflective cycle about a lesson I taught recently. Ask one question at a time and wait for my answer. Do not judge or rate, and do not give advice before the analysis stage. Ask for evidence when I make generalisations, point out my assumptions, and offer at least one alternative interpretation. At the end, summarise the patterns you noticed and help me formulate one or two small, concrete action steps. I will not share the names or personal data of learners or colleagues.”

One question at a time

Ask the AI to wait for your answer, request concrete examples, and avoid rushing into advice before the analysis stage.

One or two small steps at the end

Use the summary to decide what to try, when to try it, and how you will know whether it worked.

Professional responsibility

AI asks, the teacher decides

AI can be a good partner, but it is not infallible and tends to agree with whoever is asking. Ask it to point out your assumptions, offer alternative interpretations and ask further questions.

Do not ask it to write the reflection for you or evaluate your work. Self-reflection and decisions remain with the teacher.

Colleagues bring shared institutional experience and knowledge of the situation; learners bring their own experience. AI cannot replace these perspectives.

A teacher reflects on her practice; AI supports her with questions rather than scores, with colleagues and learners involved. — AI can be a questioning partner, but the reflection and the decisions remain with the teacher; colleagues and learners are indispensable

Protecting sources and trust

Do not share the names of learners or colleagues, or recognisable personal details. Recordings require an appropriate legal basis under institutional data-handling rules and prior information; obtain consent where required. Clarify how the material will be processed and deleted in advance.

A transcript used for self-reflection is not performance appraisal. If the purpose changes to evaluating learners or teachers, other, stricter requirements may apply. Follow current institutional and professional guidance.

Ethical use · Traffic lights

What belongs in self-reflection?

These signals show the boundaries of the self-reflection practice presented here. Always follow institutional rules and the conditions for handling data.

Green · Generally suitable

Organising your own voice notes without personal data.
Summarising genuinely anonymous exit cards.
Using the example prompts without recognisable personal details.

Yellow · Subject to conditions

Analysing a lesson transcript with data handling, prior information and deletion procedures agreed in advance.
Video analysis among colleagues under institutional conditions.
Working with simulated learners while recognising the limits of simulation.

Red · Avoid in this practice

Automatically inferring emotions or attention from learners’ faces or voices.
Automatically ranking learners.
Forwarding AI summaries to management without the teacher’s knowledge.

The EU AI Act prohibits certain uses of emotion-recognition AI in educational institutions, with narrow medical and safety exceptions. These traffic lights provide pedagogical guidance; they do not replace a legal assessment of a specific use. European Commission: prohibited AI practices.

Practical examples for teachers

Individual dialogue and shared reflection

Individual exercise · 35–40 minutes

Reflective dialogue after a lesson

Tools: a conversational AI tool, your notes and anonymised evidence, a reflective journal or CPD log.

Choose a recent lesson and collect anonymised notes, feedback and aggregated results.
Use the “A critical friend at the end of the day” prompt to work through Gibbs’ cycle. Give concrete examples.
Evaluate the dialogue itself: which question helped? Where did the AI agree too readily? What alternative explanation was left out?
Write one small action step: what, when, and how will I know whether it worked? Note one question for a colleague as well.

Shared exercise · 45 minutes

Critical friends triad

Tools: the summaries and action steps from the individual exercise, an observation sheet and a timer.

Assign the roles of presenter, questioner and observer. The presenter summarises the situation and planned action step in three minutes.
The questioner asks open questions for eight minutes, initially without giving advice: this practises Brookfield’s colleague lens.
The observer notes which questions differed from the AI’s and what the colleague noticed that the AI did not.
Rotate the roles twice. Finally, discuss what AI adds and what colleagues and learners contribute.

Voluntary and confidential dialogue

Participation and the content of reflections are voluntary and confidential; they must not be used for performance appraisal. In the group, describe learners and colleagues without names or recognisable details.

Detailed instructions and lessons learnt are available in the practical exercises PDF on the credits page.

Takeaways for teachers

Four points to guide your next lesson

Reflection is not a new requirement but an old, often invisible part of good teaching. What has been missing so far is time and a conversation partner who is always available. So the question is no longer whether there is time for reflection, but whether we make use of the opportunity.

Turn experience into learning
Self-reflection is the foundation of teacher development: experience is continuously fed back to improve the quality of teaching and learning.
Build on evidence
Compare notes, appropriately handled recordings, photos and learner feedback. Verify the AI’s interpretation.
Take small, regular steps
Five minutes a day, one piece of evidence a week and one conversation a month can be worth more than an occasional big analysis.
Let AI ask questions; let the teacher decide
Let AI be a partner that supports thinking. The decision and responsibility remain with the teacher; the perspectives of colleagues and learners are indispensable.

Your first step next week

Record a two-minute voice note without personal data after each of two lessons and talk it through using the “A critical friend at the end of the day” prompt. At the end of the week, ask yourself: what did you notice that you had not noticed before?

Self-check quiz

Check your knowledge

Select one answer for each question, then check your answers. Your self-check result appears only on this page.

1What is the appropriate role of AI in teacher self-reflection?To score and rate the teacher’s work.To ask questions and offer other interpretations, while the teacher makes the decisions.To write the reflection for the teacher.

2Which step completes Gibbs’ reflective cycle?Naming feelings.Describing the events of the lesson.Action plan: what will I try next time?

3What should we do if the teacher’s notes and learner feedback contradict each other?Compare the sources and look for several possible explanations.Automatically accept the AI’s interpretation.Leave learner feedback out of the analysis.

4Which rhythm can help embed reflection in everyday work?One major analysis of all lessons once a year.Five minutes a day, one piece of evidence a week, one conversation a month.Reflect only when a lesson goes noticeably badly.

5Which statement about processing a lesson recording is correct?It is enough to remove names from the filename.It can be uploaded to any public tool for self-reflection.Institutional data handling, prior information and deletion procedures must be clarified in advance.

6What do colleagues add to reflection prepared with AI?Knowledge of the context, shared experience and questions that differ from the AI’s.They simply repeat the AI’s summary.They automatically confirm that the AI’s conclusion is correct.

Sources

References and further reading

References for this learning material

Brookfield, S. D. (1995): Becoming a Critically Reflective Teacher. San Francisco: Jossey-Bass, p. 7. The idea used in this material is cited in Hanson (2011), p. 295.
OECD (2010): PISA 2009 Results: What Makes a School Successful? Resources, Policies and Practices (Volume IV). OECD Publishing.
Wu, Y. (2022): How to Be a Reflective Teacher? Journal of Contemporary Educational Research, 6(6), 50–54.
Hanson, K. (2011): ‘Reflect’ – is this too much to ask? Reflective Practice, 12(3), 293–304.
Schön, D. A. (1983): The Reflective Practitioner: How Professionals Think in Action. New York: Basic Books.
Gibbs, G. (1988): Learning by Doing: A Guide to Teaching and Learning Methods. Oxford: Further Education Unit, Oxford Polytechnic.
UNESCO (2024): AI Competency Framework for Teachers. Paris: UNESCO.
European Commission (2022): Ethical guidelines on the use of artificial intelligence (AI) and data in teaching and learning for educators. Luxembourg: Publications Office of the European Union.

Reflection Toolkit

Tools and responsible use

VITA · Credits

Credits and copyright

Author: Mária HartyányiiTStudy Hungary Training and Research Centre
Reviewer: Annamária KacsuriTStudy Hungary Training and Research Centre
Project identifier: 2025-1-LT01-KA220-VET-000362456
Learning material: M5-T1-MLC3

Attachments

Download the learning materialM5-T1-MLC3-TEXT-EN.pdf Download practical exercisesM5-T1-MLC3-EXC-EN.pdf

This work may be reused and adapted under the terms of the Creative Commons Attribution-ShareAlike 4.0 International License (CC BY-SA 4.0).

Funded by the European Union. Views and opinions expressed are those of the author(s) only and do not necessarily reflect those of the European Union or the National Agency.