By John Bigelow
The Padagogy Wheel provides educators with a simple way of thinking about a complicated problem: how do we choose technology based on what we want students to learn, rather than choosing a technology and then trying to find something educational to do with it?
It is a distinction that matters in 2026.
Generative AI can produce explanations, summaries, questions, feedback and plausible answers to almost anything. Inevitably, this raises the question: where does AI fit on the Padagogy Wheel?
Perhaps the better question is: should it fit anywhere in particular?
The value of the wheel is not that an application must belong permanently beside one learning activity. It is better viewed as a reflective planning aid connecting learning objectives, cognitive processes and technology choices. Generative AI makes that distinction important because the same tool can support different kinds of thinking — or remove the thinking altogether.
Start With The Thinking
The revised Bloom’s Taxonomy separates the knowledge students are expected to acquire from the cognitive processes they use: Remember, Understand, Apply, Analyse, Evaluate and Create. The framework can help educators align objectives, learning activities and assessment (Krathwohl, 2002).
That provides a useful starting point for AI.
If students need to Remember, AI might generate practice questions or flashcards from teacher-approved material. However, having AI answer every factual question rather defeats the exercise.
At Understand, students might ask an AI system to explain a concept in different ways, then compare those explanations with course materials and identify inaccuracies or omissions.
At Apply, students could use AI to generate scenarios requiring them to apply a principle, procedure or formula. The important learning occurs not because AI produced the scenario, but because the student must decide how existing knowledge applies to it.
At Analyse, students might compare AI-generated explanations, identify assumptions, separate evidence from assertion or examine why different prompts produce different answers.
At Evaluate, educators can turn one of AI’s weaknesses into a learning opportunity. Rather than asking AI for the “correct” answer, give students an AI-generated response and require them to verify it, critique its reasoning, identify bias or unsupported claims and defend their judgement using evidence.
Finally, at Create, AI might act as a brainstorming partner, critic or feedback mechanism while students design an argument, presentation, solution or original work. The question is whether students remain responsible for the intellectual decisions underpinning the finished product.
This distinction matters because evidence concerning generative AI and learning is encouraging, but not a licence to abandon instructional design. A systematic review and meta-analysis found positive effects associated with ChatGPT interventions across academic performance, affective-motivational states and higher-order-thinking propensities, while also finding reduced mental effort and no significant effect on self-efficacy. Importantly, much of the evidence was concentrated in higher education and language learning, making claims across every classroom difficult to justify (Deng et al., 2024).
SAMR Is Not A Ladder
SAMR provides a useful lens, provided educators resist the temptation to treat Substitution, Augmentation, Modification and Redefinition as rungs on a ladder to educational enlightenment.
Hamilton, Rosenberg and Akcaoglu (2016) caution that SAMR can oversimplify technology integration, neglect context, focus on products rather than processes and create the impression that Redefinition is inherently superior. It is not.
Sometimes substitution is precisely what is required. If AI helps provide timely feedback within an existing activity, there is little merit in redesigning the entire exercise simply so it can be labelled “transformative”.
Conversely, AI may enable activities that alter the learning process. Students might interrogate competing AI-generated positions, progressively refine solutions through feedback or examine how generated outputs change as assumptions and instructions change.
But technology doing something impressive is not the same as students learning something important.
Adding AI Without Removing The Learner
Research identifies opportunities for generative AI in personalised assistance, explanation, feedback and content creation. It also identifies familiar problems: inaccurate information, bias, privacy concerns, unequal access, academic integrity, over-reliance and the need for clear guidance, human oversight and AI literacy (Kasneci et al., 2023; Ogunleye et al., 2024).
For schools in particular, AI literacy therefore cannot consist merely of teaching students how to write better prompts. Research into K–12 AI education points towards authentic contexts, collaborative learning, teacher involvement, deliberate scaffolding and multiple forms of evidence (Liu & Zhong, 2024).
So perhaps the 2026 Padagogy Wheel does not need another ring labelled “AI”. Instead, AI might be better viewed as a layer sitting across the existing wheel, accompanied by a handful of questions.
What do I want students to learn? What thinking must they actually do to learn it? What role, if any, should AI play? How will students verify its output? How will they make their use of AI transparent? And, most importantly, if I remove the technology from this activity, can I still explain the learning objective?
If the answer to that final question is no, then perhaps it is time to go around the wheel again.
References
- Deng, R., Jiang, M., Yu, X., Lu, Y., & Liu, S. (2024). “Does ChatGPT enhance student learning? A systematic review and meta-analysis of experimental studies.” Computers & Education, article 105224. https://doi.org/10.1016/j.compedu.2024.105224
- Hamilton, E. R., Rosenberg, J. M., & Akcaoglu, M. (2016). “The Substitution Augmentation Modification Redefinition (SAMR) Model: a Critical Review and Suggestions for its Use.” TechTrends, 60(5), 433–441. https://doi.org/10.1007/s11528-016-0091-y
- Kasneci, E. et al. (2023). “ChatGPT for good? On opportunities and challenges of large language models for education.” Learning and Individual Differences, 103, 102274. https://doi.org/10.1016/j.lindif.2023.102274
- Krathwohl, D. R. (2002). “A Revision of Bloom’s Taxonomy: An Overview.” Theory Into Practice, 41(4), 212–218. https://doi.org/10.1207/s15430421tip4104_2
- Liu, X., & Zhong, B. (2024). “A systematic review on how educators teach AI in K-12 education.” Educational Research Review, 45, 100642. https://doi.org/10.1016/j.edurev.2024.100642
- Ogunleye, B., Zakariyyah, K. I., Ajao, O., Olayinka, O., & Sharma, H. (2024). “A Systematic Review of Generative AI for Teaching and Learning Practice.” Education Sciences, 14(6), 636. https://doi.org/10.3390/educsci14060636
Latest posts by John Bigelow (see all)
- The Padagogy Wheel in 2026: Using Bloom’s Taxonomy, SAMR and AI to Design Better Learning - September 9, 2026
- Competition Brings STEM Skills to Life - July 27, 2021
- Making Technology Safe - February 20, 2020
You must be logged in to post a comment.
There are no comments
Add yours