Date
Publisher
arXiv
The integration of artificial intelligence (AI) into education presents new opportunities for supporting learning processes. This study investigates the impact of AI-assisted versus traditional Excel-based data analysis on both learning outcomes and emotional-motivational responses in a physics education context. A custom GPT-based chatbot, ExperiMentor, was developed to support student teachers in analyzing experimental data from thread and spring pendulum experiments. Fifty student teachers were randomly assigned to either the AI or Excel group, with both groups completing identical tasks in a guided setting. Learning progress was measured using pre- and post-tests, while emotional and motivational variables were assessed through structured surveys. Both groups demonstrated significant learning gains, with no statistically significant differences found between them in terms of cognitive performance. However, the AI group reported substantially higher levels of engagement, enjoyment, and perceived method effectiveness compared to the Excel group. These findings suggest that interactive AI tools may enhance the affective dimensions of learning, even when cognitive outcomes remain comparable to traditional methods. The results underscore the importance of integrating AI not as a replacement for instructional design, but as a supportive element within pedagogical frameworks. Future research should explore long-term retention effects, the role of learner diversity, and comparisons with other forms of pedagogical support.
Study at a glance
- Question: How does AI-assisted data analysis compare to spreadsheet-based analysis in physics learning outcomes and students' emotional experiences?
- Setting and data: University of Cologne, Germany; 50 student teachers (35 bachelor's, 15 master's); June to August 2024.
- Method: Randomized controlled trial with pre- and post-tests and surveys.
Why it matters
Although artificial intelligence tools show promise for education, empirical evidence on their learning impact remains inconclusive, highlighting the need to understand how AI-assisted data analysis compares to traditional tools in physics instruction.
Key findings
- Both groups showed significant learning gains, but learning outcomes did not differ significantly between students using the AI chatbot and those using Excel.
- Student teachers using the custom AI tool reported significantly higher positive emotional learning experiences, motivation, and perceived method effectiveness than those using Excel.
- Students in the AI group experienced significantly lower levels of negative emotions, including stress and frustration, compared to students in the Excel group.
Summary drafted with AI from the full paper and may contain errors. See the full paper for details.
What is the application?
Who is the user?
Who age?
Why use AI?
