Harnessing AI to Develop Students’ Statistical Thinking: Project Page

Project Contributors

In alphabetical order:

Undergraduate RAs:

  • Bhuvan Kala (Phase 1: task design, interviewing, transcription, qualitative coding, structural equation modeling)
  • Julianna Drew (Phase 1: task design, interviewing, transcription, qualitative coding, classroom resources design)
  • Sanjana Gongati (Phase 1: task design, interviewing, transcription, qualitative coding)
  • Preetha Manjunath (Phase 2a: transcription)
  • Jasmin Zepeda (Phase 2a: transcription)
  • Pavitra Shankar (Phase 2b: UI/UX survey design, exploratory data analysis)
  • Rayhan Driver (Phase 2b: classroom resources design)

Introduction to the Project

In October 2022, OpenAI’s ChatGPT3.0 took the world by storm. Since then, AI has seeped into seemingly every aspect of our personal and professional lives. As with any new powerful tool, AI has the potential to transform the way we live and work, but wielded inappropriately and it can also harm individuals and society.

We believe that AI tools can help students learn, especially statistics and data science, but may need guardrails to ensure students’ don’t outsource thinking (and therefore their learning) to the AI tool. Our project draws on design-based educational research methods to investigate how AI tools can best be incorporated into statistics and data science education.

Current Status

We have primarily focused thus far on students’ creation and interpretation of multivariable data visualizations, and how AI tools can support both procedural fluency and conceptual understanding. We have established that:

  • there is something of a minimum level of conceptual understanding students’ must have for effective utilization of the AI tool
  • utilizing the AI tool appears to support the development of students’ conceptual understanding
  • exposure to the AI tool appears to support the development of students’ interests and goals towards statistics and data science

Our upcoming work will focus jointly on:

  • design, evaluation, and implementation of classroom activities
  • testing and comparison of AI-based activities to non-AI-based activities
  • evaluation of impacts of the activities on learning outcomes and associates

Publications

Rao, V. N. V., Jeng, A., Drew, J., Kala, B., & Gongati, S. (2026). Students’ Statistical Thinking When Using Generative AI: A Descriptive Case Study. Journal of Statistics and Data Science Education, 1–12. https://doi.org/10.1080/26939169.2025.2596684

Presentations

Kala, B., Drew, J.M., & Rao, V.N.V. (2025, July). Rtutor.AI Labs: Incorporating AI into Introductory Statistics [Poster Presentation]. United States Conference on Teaching Statistics (USCOTS), Ames, Iowa, USA.

Gongati, S., Jeng, A., Drew, J.M., Kala, B., & Rao, V.N.V. (2025, July). Students’ statistical thinking when using generative AI [Poster Presentation]. United States Conference on Teaching Statistics (USCOTS), Ames, Iowa, USA.

Kala, B., Bandi, L., & Rao, V.N.V. (2025, July). Can using AI tools to do statistics broaden participation in STEM? [Poster Presentation]. Consortium for the Advancement of Undergraduate Statistics Education (CAUSE) Research Satellite, Ames, Iowa, USA.

Kala, B., Drew, J., Gongati, S., & Rao, V.N.V. (2024, May). Computational Thinking with Data [Oral Presentation]. University of Illnois Urbana-Champaign Undergraduate Research Experience in Statistics (URES) Symposium, Urbana, Illinois, USA.

Manuscripts in Preparation

Rao, V.N.V., & Ge, X. (in preparation). Rtutor: Marrying R to ChatGPT to support the development of students’ statistical thinking.

Rao, V.N.V., Bandi, L., & Kala, B. (in preparation). Using AI tools to do statistics can broaden participation in STEM.

Reyes-Denis, T., Rao, V.N.V., & Jeng, A. (in preparation). R vs. AI: A comparative descriptive case study of students’ statistical thinking.

Funding

This work has been generously supported by a grant from the University of Illinois Center for Social and Behavioral Sciences. In particular, we would like to thank Kaylee Lukacena for support on this project.