AI Insights #6

This week's AI Insights newsletter is focused on a single topic: the role of artificial intelligence in the classroom, drawing insights from a five-week experiment with Year 9 students. In previous editions, I've shared my weekly reflections on the project; now, I’m presenting the results of a student survey. Their voices are key in understanding AI's impact on learning.

Does AI have a place in the classroom?

As AI continues to make waves, its potential to enhance education is increasingly in the spotlight. To explore this potential, we conducted a five-week project with Year 9 students using an AI chatbot named Canvy to develop their Canva skills. The experiment aimed to test whether AI could serve as an effective learning partner, offering personalised guidance while promoting student autonomy.

Blending Structure with Autonomy

This experiment departed from the traditional ‘one-size-fits-all’ classroom setup. With Canvy at the helm, students could take charge of their learning paths.

Here’s how it worked:

1. Students began by self-assessing their Canva skills.
2. They input this information into Canvy.
3. The chatbot generated a four-week learning plan tailored to each student.
4. Canvy provided weekly guidance, directed students to relevant resources, and offered feedback.

This approach allowed each student to follow a unique learning path, with the AI offering structured support that, in theory, would free me to engage with students on a deeper level. The real question, however, is whether theory matched reality.

Key Findings

Strengths of the AI Learning Partner

1. Structured Learning Plans: The learning plan provided by Canvy received high approval from students in two key areas: Usefulness: 80% of students agreed or strongly agreed that Canvy's learning plan was helpful. Clarity: 70% of students agreed that Canvy clearly outlined what was expected of them each week. This strong positive response reinforced the value of structured guidance. One student noted, "It was helpful to see a plan laid out for me; it made it easier to understand what was expected each week."
2. Helpful Feedback: 60% of students agreed or strongly agreed that the feedback provided by Canvy was helpful in completing their project. Many students appreciated how Canvy provided guidance on their work. One student commented, "The feedback provided by Canvy was helpful because it broke down what I needed to do."
3. Improved Time Management and Autonomy: Over half of the students (53.3%) found it easier to manage their time, and 54.7% reported working more independently than usual. Although these percentages don't scream "revolution," they do suggest that AI can push students toward taking more responsibility for their learning.
4. Interest in Future Use: 65.3% of students agreed or strongly agreed that they would be interested in using AI tools like Canvy for future projects, suggesting a positive reception to this type of technology in education.
5. Enhanced Teacher Support: The greatest benefit was perhaps the AI's ability to handle routine guidance, freeing me to engage more deeply with students on a one-on-one basis. This shift allowed for more meaningful interactions, particularly for those struggling with specific project aspects.

Areas for Improvement

1. Mixed Feedback Reception: While 60% found Canvy's feedback helpful, 40% were neutral or disagreed. Some students (10.6%) found the suggestions repetitive or lacking depth, with comments like: "It just repeated the same suggestions for different tasks." “It didn’t seem to take into account what I’d already changed, so it felt like I wasn’t making progress.” From my observations during the lessons, those who were frustrated often made minimal adjustments to their work, which prompted Canvy to repeat the same feedback. In many cases, the repeated suggestions were justified, but students seemed less forgiving of this from the AI than they would be if similar advice came from me.
2. Time Management Challenges: While 53.3% of students felt that Canvy made time management easier, experiences varied across the group. About 29.33% were neutral on its impact, 16% disagreed, and 1.33% strongly disagreed. Despite these mixed reviews, every student managed to complete their project on time. Even those who missed a few classes caught up by using Canvy and Canva in their own time, showing that the AI's structure still played a role in keeping them on track.
3. Diverse Preferences: When asked about preferring traditional teaching methods, student responses were notably divided:

• 37.3% disagreed or strongly disagreed, indicating a preference for the AI-led approach.
• 32% agreed or strongly agreed, favouring traditional teaching methods where the teacher leads step by step.
• 30.7% were neutral. This split underscores the importance of a balanced approach that combines AI-driven learning with traditional teaching methods.

High Quality of Student Work

Despite some challenges with the feedback process, I have started marking the students work and the quality of the projects is very high. More information on this next week, as I’ll be comparing my human-marked grades with AI-generated grades soon to see how AI's consistency measures up. Still, these preliminary results hint that AI could be a valuable tool for skills-based learning, particularly in projects with clear outcomes.

AI as a Complement, Not a Replacement

Ultimately, the Canvy experiment highlighted AI’s potential as a supportive learning partner, especially for tasks that lend themselves to structured guidance. The AI’s ability to automate routine aspects of teaching – creating individualised rubrics, generating learning plans, and providing basic feedback – was undeniably useful.

Yet, the results were a mixed bag. Roughly 29% of students were indifferent to AI-led learning, and a similar number did not see much impact on their time management. These divided responses reveal that while AI brings clear benefits to some, it is not a panacea for all. Its role in education must be carefully calibrated, combining AI’s efficiency with the indispensable qualities of human instruction.

Whether AI can replicate this effectiveness in more abstract or theory-heavy subjects remains an open question.

The Verdict: A Promising Partnership

So, does AI have a place in the classroom? Based on our experience, the answer is yes...but with caveats. Canvy proved itself valuable in providing structure and fostering a measure of autonomy, particularly for students who thrived on a more self-paced approach. But the experiment also reminded us that AI cannot replace the irreplaceable. It cannot read a room, sense when a student is struggling emotionally, or inject humour into a dry topic.

As we continue to explore the potential of AI in education, the real challenge will be to strike the right balance. We should harness AI to allow us to try new approaches. By giving students hands-on experience with AI, we can help them understand both its benefits and limitations, preparing them for a world where technology plays an ever-growing role.

Ta-ra Duck!

As I wrap up this week’s AI Insights, I’m looking forward to a trip to the UK and speaking at the BRILLIANT festival on Tuesday (15th October). It’s going to be an exciting day of connecting with fellow educators, sharing experiences, and exploring the future of AI in education. I can’t wait to catch up with everyone and bring back some fresh insights.

Safe travels if you’re heading to the festival too, and hope to see you there!