Throughout my career I’ve worked in fields that rely heavily on clear explanation, problem solving, and practical application. From electrical installation and engineering, to analysing school data, and teaching and leading Science across multiple schools. One thing has remained clear, people learn best when complex ideas are broken down clearly, explained at the right level, and revisited in a way that feels natural rather than intimidating.
That principle is at the heart of my research project. This blog will document the development and evaluation of a new digital system designed to support students learning Key Stage 3 Science. Over the coming weeks I’ll be building a Voice Learning Assistant (VLA), a system that aims to bring together conversational AI and clear diagrams to give students the kind of accessible explanations that are often difficult to achieve regularly enough in the classroom.
Research Focus
My research sits where education, data, and computer science meet. Drawing on my experience teaching Science and my ongoing M.Sc. in Computer Science, I’m exploring how voice enabled AI can support students in understanding scientific concepts.
The focus is on developing a VLA for KS3 Science; a system that talks with students, asks questions, responds to their thinking, and uses pre-drawn diagrams to help clarify concepts. It’s a practical idea that uses modern technology to provide the sort of responsive, diagnostic conversation a student might get from a conversation with a teacher.
Why it Matters
In classrooms, especially when teaching abstract scientific ideas, text heavy or standard teacher explanations often leave some students behind as everyone progresses at their own pace (Greenberg, Gross and Wright, 2014). Furthermore, some learners grasp ideas much more effectively through spoken explanation or by working verbally through a problem. I’ve seen this first hand in schools, when a student can talk something through they often understand it more quickly and deeply.
Similarly, in technical fields like electrical installation when working with apprentices I've found conversation, questioning, and guided explanation are essential. You don’t teach circuit protection or diagnostic reasoning by handing someone a textbook or talking them through a PowerPoint; you explain, demonstrate, question, check, and revisit. Science in schools is no different and they key is willing active participation (Frey and Fisher, n.d.).
The VLA aims to bring that principle into everyday homework and independent study.
What it Does
The VLA behaves like a calm, knowledgeable tutor who interacts through spoken conversation. Students can:
-
- Ask questions aloud
- Receive spoken explanations
- View accurate, pre-drawn diagrams
- Work through misconceptions step by step
- Learn at their own pace
It’s not a replacement for quality first teaching, it’s a support tool that helps students access the curriculum more easily between lessons.
The combination of conversational explanation and high quality visuals follows well established cognitive principles, similar to how diagrams and schematics assist understanding in engineering and electrical work. Clear and informative diagrams, paired with clear verbal explanation, help make abstract ideas more concrete.
Execution
The research will be carried out with a group of KS3 students using a simple, structured approach.
Pre-test
Students complete a short science assessment and a survey about their usual homework experience.
Two-Week Trial Using the VLA
Students complete homework tasks through spoken interaction with the system and the VLA logs their engagement, questions, and progress.
Post-test
Students complete the same assessment again so any change in understanding can be measured.
Student Reflections
Surveys and focus groups will give insight into usability, clarity, and how the tool feels to use.
This mixed methods design provides both data and student perspective which is essential in evaluating whether the approach genuinely supports learning (Ahmed, Pereira and Jane, 2024).
What I Aim to Learn
If the system meets its objectives, I expect to see:
-
- Increased engagement with homework
- Better conceptual understanding of key Science ideas
- Stronger verbal reasoning around scientific concepts
- Improved confidence for students who find text heavy tasks difficult
If the VLA works as intended, it could provide a scalable way for students to receive clear explanations and greater personalised support, especially outside lesson time.
What Comes Next
This blog will follow the full journey from creating the diagrams, structuring the system and designing the conversational logic to testing the prototype and analysing the results.
My aim is to make the process transparent and share both the successes and the challenges. Whether the tool delivers a significant impact or simply offers insights for future work, I want this to be a useful and honest account of the development process.
Thank you for reading the first entry and please check back for the next post.