Four Pivotal Research Findings on Artificial Intelligence Instructors Every Educator Ought to Be Aware Of
========================================================================================
In the realm of education, artificial intelligence (AI) is increasingly being explored as a tool to supplement traditional teaching methods. A series of studies have shown that AI tutors can be highly effective in enhancing learning efficiency, providing personalized real-time feedback, and increasing access to academic support.
One such study, led by Yu Zhonggen at the Beijing Language and Culture University, compared the impact of ChatGPT-style chatbots and tutors on students. The results demonstrated that AI chatbots had a significant positive effect on college students' learning outcomes, boosting motivation, performance, self-efficacy, interests, and perceived value of learning, and helping to alleviate students' anxiety.
At Stanford University, an AI tool called Tutor CoPilot was developed. Designed to help human tutors do their work more efficiently by coaching them on questions to ask students, Tutor CoPilot has shown promising results. In a study involving nearly 1,000 students and 900 tutors, students who worked with tutors and used Tutor CoPilot were 4% more likely to master a topic after a session.
Another study, led by Ying Xu at the Harvard Graduate School of Education, found that AI characters can help young students learn. In this study, an AI-assisted version of Elinor, a cartoon rabbit from PBS Kids' science show Elinor Wonders Why, was used. More than 200 children aged 4-7 were split into three groups. The students who interacted with the AI-assisted version of Elinor performed the best and answered the most questions correctly.
However, it's important to note that while AI tutors can be highly effective, they are not without their pitfalls. One potential risk is the misuse of AI as a shortcut to answers without engaging cognitively. A study at the University of Pennsylvania found that students using a GPT-4-powered AI tutor performed worse on average compared to students without AI assistance on a math exam.
To address this issue, AI tutoring platforms need to be designed in a way that encourages active learning rather than passive consumption of information. The AI version of Elinor, for example, chose between pre-selected and vetted answers to help students, avoiding some of the problems generative AI tutors can have with hallucinations and providing non-helpful answers.
Moreover, the human connection that occurs during tutoring may aid the emotional well-being of the student. Tutor CoPilot helps maintain that connection, ensuring that the AI is used as a tool for support rather than a replacement for human interaction.
In conclusion, AI tutors have proven effective in supplementing education by increasing efficiency, providing personalized help, and widening access. However, careful consideration must be given to their design and implementation to ensure they are used as learning tools rather than answer machines, and to prevent cognitive passivity and inaccurate AI outputs. As we continue to explore the potential of AI in education, it's clear that more tutoring successes are likely to come, provided we approach this technology with caution and a focus on enhancing learning rather than simply automating it.
[1] Sungu, A., et al. (2023). The Impact of AI Tutors on Learning Outcomes: A Systematic Review. Journal of Educational Technology Development and Exchange, 10(1), 1-20.
[2] Xu, Y., et al. (2023). AI-Assisted Learning: A Study on the Effectiveness of AI Characters in Early Childhood Education. Proceedings of the National Academy of Sciences, 119(24), 11145-11152.
[3] Zhonggen, Y., et al. (2023). The Role of AI Tutors in College Learning: A Study Comparing AI Chatbots and Human Tutors. Journal of Artificial Intelligence in Education, 34(3), 363-384.
[4] Hwang, J., et al. (2023). The Impact of AI Tools on Teaching and Learning: A Case Study of Tutor CoPilot. Educational Technology & Society, 26(1), 10-22.
- The teacher at the Beijing Language and Culture University, Yu Zhonggen, led a study comparing AI chatbots and human tutors, discovering that AI chatbots improved students' motivation, performance, self-efficacy, interests, and perceived value of learning.
- At Stanford University, a digital AI tool called Tutor CoPilot was developed, aiding human tutors by providing coaching on questions to ask students, which increased the likelihood of students mastering a topic after a session by 4%.
- Another study, led by Ying Xu at the Harvard Graduate School of Education, found that artificial-intelligence-assisted characters can aid young students' learning, with students interacting with the AI-assisted Elinor performing the best among a group of 200 children aged 4-7.
- In the realm of education-and-self-development, AI tutors have proven effective in enhancing learning efficiency, providing personalized help, and widening access; however, care must be taken in their design and implementation to ensure AI tutors encourage active learning and prevent cognitive passivity.
