Identifying Tutoring Strategies for an AI Chatbot: Participatory Co-Design With Math Teachers

• In Event: AERA Roundtable Session Saturday 1:30 pm Four Seasons Ballroom 1
  In Roundtable Session: AI and Technology Integration in Education (Table 2)

Abstract

This study used a participatory co-design with middle school math teachers to explore tutoring strategies an AI chatbot can use to support struggling students and examine teachers’ perspectives on the efficacy of tutoring dialogues generated by GPT4. We conducted interviews with six diverse 7th grade math teachers to learn about their teaching philosophies and tutoring strategies. Teachers then wrote math tutoring dialogues in a chatbot style. We analyzed tutoring strategies teachers used and iteratively improved the prompt for GPT4 to generate tutoring dialogues, which were mostly deemed as very helpful for students. This study identified key tutoring strategies AI chatbots can use to support struggling students to learn math and demonstrated the potential of employing Generative AI to create chatbots.

Authors

• Li Cheng, University of North Texas

• Natalie Brezack, WestEd

• Eamon Worden, Worcester Polytechnic Institute

• Wynnie Chan, WestEd

• Jenna Grady, WestEd

• Mingyu Feng, WestEd

• Cristina Heffernan, The ASSISTments Foundation

• Neil T. Heffernan, Worcester Polytechnic Institute