Bridget Lee McCONNELL
Department of Psychology, Faculty of Arts and Social Sciences (FASS), National University of Singapore (NUS)
Sub-Theme
Building Technological and Community Relationships
Keywords
ChatGPT, AI, peer-to-peer teaching, peer tutoring
Category
Paper Presentation
Introduction
Research on peer teaching and reciprocal peer tutoring show that having students teach each other has positive effects on their own understanding of the topic (Mafarja et al., 2023). As teachers, students must create a lesson plan, which requires them to organise the subtopics into a logical order. They must sufficiently understand a complex topic to break it down for others, and they must find intuitive ways to present the information. Finally, they must be able to answer spontaneous questions from their students (e.g., Ali et al., 2015; Galbraith & Winterbottom, 2011).
However, this teaching technique is not always successful. The quality of learning varies based on the student-teacher’s motivation and ability (e.g., AlShareef, 2020). Moreover, from my own observations, students’ lessons tend to be superficial and limited to repeating the textbook, and other students hesitate to ask difficult or probing questions due to a mutual understanding that they themselves do not want to be asked difficult questions when it is their turn to teach. As a result, university instructors do not always think that peer teaching results in deeper learning (e.g., Stigmar, 2016).
This talk presents an active learning intervention in which students were tasked with becoming the teacher. However, in contrast to the traditional peer-to-peer teaching model, in this situation, the student-teacher’s student was ChatGPT. Students were given a starting prompt, which instructed ChatGPT to act like a motivated by uninformed student. Importantly, the prompt limited ChatGPT’s knowledge only to the information they provided; the LLM was specifically instructed not to draw on outside knowledge. The students had specific learning outcomes they had to achieve by the end of their conversation, and they had to write a reflection about their learning experience.
As the teacher, the students were forced to understand the topic on a deeper level. Also, ChatGPT was instructed to ask questions and make mistakes, which challenged students to critically evaluate the responses being given to them from ChatGPT and expand their knowledge to think beyond the superficial level to answer ChatGPT’s questions, which ranged from simple to complex. These learning benefits were acknowledged in the students’ reflections where many discussed their experience of thinking they understood the topic from the lecture, but they realised shortcomings in their knowledge when they were unable to answer ChatGPT’s question. Students also discussed the challenge of figuring out how to structure the lesson and what order to introduce topics that would make the most sense. This assignment gave students a unique perspective of what it’s like to truly teach someone who does not know anything about the topic, which helped them appreciate the challenges involved in teaching for understanding. Finally, it gave students a new way to interact and use LLMs to facilitate their learning instead of just substituting for their learning. As evidence of this, many students commented that they will incorporate this same activity to help them student for other topics or other courses as it helped identify gaps in their knowledge.
This talk will outline the assessment components, benefits and evidence for learning, student feedback on their experience, suggestions for how others can incorporate it into their curriculum, and some of the challenges in conducting this assessment and potential changes for future iterations.
References
Ali, N., Anwer, M., & Abbask J. (2015). Impact of peer tutoring on learning of students. Journal for Studies in Management and Planning, 1(2), 61-66. https://ssrn.com/abstract=2599095
AlShareef, S. M. (2020). Comparing the impacts of reciprocal peer teaching with faculty teaching: A single-centre experience from KSA. Journal of Taibah University Medical Sciences, 15(4), 272-277. https://doi.org/10.1016/j.jtumed.2020.05.006
Galbraith, J., & Winterbottom, M. (2011). Peer-tutoring: What’s in it for the tutor? Educational Studies, 37(3), 321-332. https://doi.org/10.1080/03055698.2010.506330
Mafarja, N., Mohamad, M. M., Zulnaidi, H., & Fadzil, H. M. (2023). Using of reciprocal teaching to enhance academic achievement: A systematic literature review. Heliyon, 9, e18269. https://doi.org/10.1016/j.heliyon.2023.e18269
Stigmar, M. (2016). Peer-to-peer teaching in higher education: A critical literature review. Mentoring & Tutoring: Partnership in Learning, 24(2), 124-136. http://dx.doi.org/10.1080/13611267.2016.1178963