A pedagogical model of generative AI tutoring in informatics education and its effects on problem-solving quality and academic independence

Raushan Sambetova; Lyazzat Zhaidakbayeva; Raushan Sambetova; Lyazzat Zhaidakbayeva

doi:10.3934/steme.2026037

STEM Education

2026, Volume 6, Issue 5: 927-951. doi: 10.3934/steme.2026037

Previous Article Next Article

Research article

A pedagogical model of generative AI tutoring in informatics education and its effects on problem-solving quality and academic independence

Raushan Sambetova ^{1
,
,},
Lyazzat Zhaidakbayeva ^2,

1.
International University of Tourism and Hospitality, Turkistan, 161200, Kazakhstan
2.
Mukhtar Auezov South Kazakhstan University, Shymkent, 160000, Kazakhstan

Academic Editor: Feng-Kuang Chiang

Received: 21 February 2026 Revised: 13 July 2026 Accepted: 22 July 2026 Published: 03 September 2026

Generative AI systems are increasingly adopted as on-demand tutors in informatics and introductory programming, yet evidence remains mixed regarding whether performance gains come at the cost of reduced academic independence and heightened over-reliance. Resolving this issue requires moving beyond open-ended AI assistance and toward tutoring systems where pedagogical safeguards are built into the design from the beginning, rather than added later as an afterthought. In this study, we proposed and evaluated an Independence-by-Design pedagogical model for generative AI tutoring that targets improvements in problem-solving quality while preserving learners' academic independence through scaffolded, guardrail-based assistance. The intervention operationalizes tutoring policies such as staged hints, explain-first prompting, and answer-suppression aligned with a programming problem-solving rubric, and it was evaluated in a controlled classroom study comparing a guardrail-based tutor condition with a control condition that used standard instructional resources without AI assistance. Primary outcomes included problem-solving quality, assessed via rubric-based scoring of correctness, reasoning, and code quality, and academic independence, quantified using an Academic Independence Index that triangulates self-regulated learning measures, help-seeking behaviors, and interaction trace features (e.g., hint depth, revision cycles, and direct-answer requests). Effects were estimated using regression models with prior achievement covariates and reported with standardized effect sizes. Results indicated that the guardrail-based generative AI tutor improves problem-solving quality relative to baseline support while maintaining or increasing academic independence indicators, and interaction analyses suggest that moderate hint usage and explain-first behavior are associated with the strongest outcomes. These findings support the responsible deployment of generative AI tutors in informatics education by demonstrating that learning gains can be achieved without amplifying dependency when pedagogical guardrails are embedded in the tutoring design. In practical terms, we give educators a transferable design approach for using generative AI in programming courses. This approach includes staged hints, requiring students to attempt a solution before receiving help, avoiding direct code answers, and prompting students to explain their thinking first. Together, these features help AI support students' reasoning rather than replace it.
- generative artificial intelligence,
- large language models,
- intelligent tutoring systems,
- computer science education,
- problem-solving performance,
- academic independence,
- self-regulated learning
Citation: Raushan Sambetova, Lyazzat Zhaidakbayeva. A pedagogical model of generative AI tutoring in informatics education and its effects on problem-solving quality and academic independence[J]. STEM Education, 2026, 6(5): 927-951. doi: 10.3934/steme.2026037

Related Papers:

Abstract

Generative AI systems are increasingly adopted as on-demand tutors in informatics and introductory programming, yet evidence remains mixed regarding whether performance gains come at the cost of reduced academic independence and heightened over-reliance. Resolving this issue requires moving beyond open-ended AI assistance and toward tutoring systems where pedagogical safeguards are built into the design from the beginning, rather than added later as an afterthought. In this study, we proposed and evaluated an Independence-by-Design pedagogical model for generative AI tutoring that targets improvements in problem-solving quality while preserving learners' academic independence through scaffolded, guardrail-based assistance. The intervention operationalizes tutoring policies such as staged hints, explain-first prompting, and answer-suppression aligned with a programming problem-solving rubric, and it was evaluated in a controlled classroom study comparing a guardrail-based tutor condition with a control condition that used standard instructional resources without AI assistance. Primary outcomes included problem-solving quality, assessed via rubric-based scoring of correctness, reasoning, and code quality, and academic independence, quantified using an Academic Independence Index that triangulates self-regulated learning measures, help-seeking behaviors, and interaction trace features (e.g., hint depth, revision cycles, and direct-answer requests). Effects were estimated using regression models with prior achievement covariates and reported with standardized effect sizes. Results indicated that the guardrail-based generative AI tutor improves problem-solving quality relative to baseline support while maintaining or increasing academic independence indicators, and interaction analyses suggest that moderate hint usage and explain-first behavior are associated with the strongest outcomes. These findings support the responsible deployment of generative AI tutors in informatics education by demonstrating that learning gains can be achieved without amplifying dependency when pedagogical guardrails are embedded in the tutoring design. In practical terms, we give educators a transferable design approach for using generative AI in programming courses. This approach includes staged hints, requiring students to attempt a solution before receiving help, avoiding direct code answers, and prompting students to explain their thinking first. Together, these features help AI support students' reasoning rather than replace it.

References

[1]	Anders, A.D. and Speltz, E.D., Developing generative AI literacies through self-regulated learning: A human-centered approach. Computers & Education: Artificial Intelligence, 2025, 9: 100482. https://doi.org/10.1016/j.caeai.2025.100482 doi: 10.1016/j.caeai.2025.100482
[2]	Lai, C.-H. and Lin, C.-Y., Analysis of learning behaviors and outcomes for students with different knowledge levels: A case study of intelligent tutoring system for coding and learning (ITS-CAL). Applied Sciences, 2025, 15(4): 1922. https://doi.org/10.3390/app15041922 doi: 10.3390/app15041922
[3]	Ministry of Education of the Republic of Kazakhstan & Ministry of Digital Development, Innovation and Aerospace Industry. (2025). Conceptual Framework for the Implementation of Artificial Intelligence in Secondary, Technical and Vocational, and Post-Secondary Education for 2024–2029 (Joint Order No. 221, 18 September 2025). Astana, Kazakhstan. Retrieved from: [https://prg.kz]
[4]	Yan, Y.M., Chen, C.Q., Hu, Y.B. and Ye, X.D., LLM-based collaborative programming: Impact on students' computational thinking and self-efficacy. Humanities and Social Sciences Communications, 2025, 12(1): 149. https://doi.org/10.1057/s41599-025-04471-1 doi: 10.1057/s41599-025-04471-1
[5]	Hou, C., Zhu, G., Sudarshan, V., Lim, F.S. and Ong, Y.S., Measuring undergraduate students' reliance on generative AI during problem-solving: Scale development and validation. Computers & Education, 2025,234: 105329. https://doi.org/10.1016/j.compedu.2025.105329 doi: 10.1016/j.compedu.2025.105329
[6]	Stojanov, A., Liu, Q. and Koh, J.H.L., University students' self-reported reliance on ChatGPT for learning: A latent profile analysis. Computers & Education: Artificial Intelligence, 2024, 6: 100243. https://doi.org/10.1016/j.caeai.2024.100243 doi: 10.1016/j.caeai.2024.100243
[7]	Létourneau, A., Deslandes Martineau, M., Charland, P., Karran, J.A., Boasen, J. and Léger, P.M., A systematic review of AI-driven intelligent tutoring systems (ITS) in K–12 education. npj Science of Learning, 2025, 10(1): 29. https://doi.org/10.1038/s41539-025-00320-7 doi: 10.1038/s41539-025-00320-7
[8]	Domino, M. and Shaer, C.A., Using a wider digital ecosystem to improve self-regulated learning. Frontiers in Education, 2025, 10: 1487344. https://doi.org/10.3389/feduc.2025.1487344 doi: 10.3389/feduc.2025.1487344
[9]	Bastani, H., Bastani, O., Sungu, A., Ge, H., Kabakcı, Ö. and Mariman, R., Generative AI without guardrails can harm learning: Evidence from high school mathematics. Proceedings of the National Academy of Sciences, 2025,122(26): e2422633122. https://doi.org/10.1073/pnas.2422633122 doi: 10.1073/pnas.2422633122
[10]	Perikos, I., Grivokostopoulou, F. and Hatzilygeroudis, I., Assistance and feedback mechanism in an intelligent tutoring system. International Journal of Artificial Intelligence in Education, 2017, 27(3): 475‒514. https://doi.org/10.1007/s40593-017-0139-y doi: 10.1007/s40593-017-0139-y
[11]	Chen, D., Zhang, Y., Luo, H., Gao, Z., Yu, L. and Lin, Y., Exploring the impact of scaffolding in programming on students' computational thinking: Evidence from a three-level meta-analysis. Journal of Educational Computing Research, 2026. https://doi.org/10.1177/07356331251386618 doi: 10.1177/07356331251386618
[12]	Yan, L., Martinez-Maldonado, R., Jin, Y., Echeverria, V., Milesi, M., Fan, J., et al., The effects of generative AI agents and scaffolding on enhancing students' comprehension of visual learning analytics. Computers & Education, 2025,234: 105322. https://doi.org/10.1016/j.compedu.2025.105322 doi: 10.1016/j.compedu.2025.105322
[13]	Hou, C., Zhu, G., Liu, Y., Sudarshan, V., Chong, J.L.L., Zhang, F.Y., et al., The effects of critical thinking intervention on reliance behaviors, problem-solving quality, and creativity during human-Generative AI collaborative learning. Computers & Education, 2026,247: 105576. https://doi.org/10.1016/j.compedu.2026.105576 doi: 10.1016/j.compedu.2026.105576
[14]	Johansen, M.O., Eliassen, S. and Jeno, L.M., The bright and dark side of autonomy: How autonomy support and thwarting relate to student motivation and academic functioning. Frontiers in Education, 2023, 8: 1153647. https://doi.org/10.3389/feduc.2023.1153647 doi: 10.3389/feduc.2023.1153647
[15]	Bassner, P., Lenk-Ostendorf, B., Beinstingel, R., Wasner, T. and Krusche, S., Less stress, better scores, same learning: The dissociation of performance and learning in AI-supported programming education. Computers & Education: Artificial Intelligence, 2025, 10: 100537. https://doi.org/10.1016/j.caeai.2025.100537 doi: 10.1016/j.caeai.2025.100537
[16]	Aruğaslan, E., Examining the relationship of academic dishonesty with academic procrastination, and time management in distance education. Heliyon, 2024, 10(19): e38827. https://doi.org/10.1016/j.heliyon.2024.e38827 doi: 10.1016/j.heliyon.2024.e38827
[17]	Brunton, R.J., Rhazzafe, S., Moodley, R., Kuhn, S., Caraffini, F., Wilford, S., et al, Using generative artificial intelligence to enhance the performance of disadvantaged students in secondary education. Social Sciences & Humanities Open, 2025, 12: 102110. https://doi.org/10.1016/j.ssaho.2025.102110 doi: 10.1016/j.ssaho.2025.102110
[18]	Deci, E.L. and Ryan, R.M., The "what" and "why" of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 2000, 11(4): 227–268. https://doi.org/10.1207/S15327965PLI1104_01 doi: 10.1207/S15327965PLI1104_01
[19]	Zimmerman, B.J., Becoming a self-regulated learner: An overview. Theory Into Practice, 2002, 41(2): 64–70. https://doi.org/10.1207/s15430421tip4102_2 doi: 10.1207/s15430421tip4102_2

Authors' biographies Raushan Sambetova is a staff member at the International University of Tourism and Hospitality in Turkistan, Kazakhstan. She completed her bachelor's degree in Informatics at the International Kazakh-Turkish University (2002–2007). She then obtained a master's degree (2015–2017) and completed her doctoral studies (2018–2021) in Informatics at M. Auezov South Kazakhstan State University. She is currently conducting research on the development of a pedagogical model for using generative AI in informatics education and examining its effects on the quality of problem-solving and students' academic independence. Raushan Sambetova's research interests lie at the intersection of artificial intelligence and education, with a focus on the safe and effective integration of generative AI tools into the learning process to enhance educational outcomes without compromising students' academic autonomy; Lyazzat Zhaidakbayeva is the Head of the Department of Informatics and a Research Fellow at M. Auezov South Kazakhstan Research University. She holds the degree of Candidate of Sciences (2010) and began her research career as a research applicant in 2000.
In 1996, she graduated from Khoja Akhmet Yassawi International Kazakh-Turkish University.
Throughout her professional career, she has held academic and leadership positions at several leading universities. She is an expert in educational programs at South Kazakhstan Pedagogical University. In 2024, she completed professional development at Lancaster University under the Bolashak International Scholarship Program.
In 2026, she was awarded the Certificate of Honor by the Ministry of Science and Higher Education of the Republic of Kazakhstan in recognition of her contributions to higher education and scientific research.
Her research interests include computer science education, STEM education, artificial intelligence in education, digital technologies, and educational innovation. She has published numerous scientific papers in high-impact international journals indexed in Scopus and Web of Science

Reader Comments

Your name:*

Email:*
© 2026 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)