The main goal of this study was to evaluate the impact of using ChatGPT as a virtual mentor on K-12 students learning science when teachers are unavailable.
Methods: The study involved a systematic assessment of ChatGPT's performance in K-12 chemistry and physics, followed by a quasi-experimental analysis with real 15-16-year-old students. Students in the experimental group used ChatGPT to correct homework and solve doubts, while the control group did not. Students' grades and perceptions of AI were monitored before and after the intervention.
Key Findings:
- ChatGPT demonstrated high performance (97% accuracy) in answering K-12 chemistry and physics questions.
- Students using ChatGPT showed a 30% improvement in grades, compared to minimal improvement in the control group.
- 70% of students reported a positive perception of ChatGPT as an educational tool after using it for two terms.
- ChatGPT provided both functional and emotional support to students, enhancing their learning experience.
Implications: The findings suggest that ChatGPT can be a valuable tool to complement traditional teaching methods, providing personalized, time-independent learning support for K-12 students in science education.
Limitations: The study was limited by a small sample size (one class) and potential biases such as the Hawthorne effect and John Henry effect.
Future Directions: Future research should involve larger, more diverse student populations and explore the long-term impacts of AI-assisted learning in various educational contexts.
Title and Authors: "Empowering Education 4.0: Impact of Using ChatGPT as a Virtual Mentor on K-12 Students Learning Science" by Rafael Castañeda, Laura Mercadé, Víctor J. Gómez, Teresa Mengual, Francisco Javier Díaz-Fernández, Miguel Sinusia Lozano, Juan Navarro Arenas, Ángela Barreda, Maribel Gómez-Gómez, Elena Pinilla Cienfuegos, and David Ortiz de Zárate.
Published on: September 17, 2024
Published by: Preprints.org