Document Type : Original Research Paper-English Issue
Authors
1 Department of Mathematics, Faculty of Science, Shahid Rajaee Teacher Training University, Tehran, Iran
2 Center for Cognitive Science, Institute for Convergence Science and Technology (ICST), Sharif University of Technology, Tehran, Iran ; School of Cognitive Sciences, Institute for Research in Fundamental Sciences (IPM), Tehran, Iran
3 Farhangian University, Bushehr, Iran
Abstract
Background and Objectives: Mathematical modelling has increasingly become a central component of mathematics education, as it helps students connect abstract mathematical concepts to real-world situations. Despite its importance, many elementary pre-service teachers (PSTs) face challenges when attempting to design or solve modelling problems, partly due to their limited exposure to such tasks during their teacher education programs. With the rapid development of AI (AI), new opportunities have emerged to compare human problem-solving approaches with machine-generated solutions. The purpose of this study is to examine how elementary PSTs and an AI system perform on mathematical modelling tasks. Specifically, the study seeks to understand the strengths and weaknesses of PSTs’ modelling competencies, while also exploring the potential contributions of AI as a supportive tool in mathematics education. The objective is not only to evaluate performance but also to address the broader significance of integrating AI into teacher education, highlighting both opportunities and risks.
Methods: A convenience sample of 50 PSTs participated in the study. Participants were asked to complete a researcher-developed test that included mathematical modelling tasks designed to evaluate different aspects of modelling competence. The face and content validity of the instrument were reviewed and confirmed by two experts in mathematics education, ensuring the appropriateness and relevance of the items. Data were collected in written form and analyzed using both quantitative and qualitative approaches. The PSTs’ responses were examined based on the competency framework, which provides structured criteria for assessing modelling skills such as understanding the problem, using mathematics, interpreting results, and validating solutions. AI-generated responses were also collected for the same tasks, and these were compared with the human responses in terms of strategy, accuracy, creativity, and solution quality. The study followed a qualitative design with deductive (directed) content analysis, which allowed the researchers to systematically code and interpret the data while drawing comparisons across human and AI performance.
Findings: The findings revealed several important patterns. Many PSTs demonstrated weaknesses in mathematical modelling, particularly in the stages of validation and interpretation. Their answers were often incomplete or lacked sufficient reasoning. However, the participants performed relatively better in the criterion of 'use of mathematics', indicating that they could carry out mathematical procedures but struggled to connect them meaningfully to real-world contexts. No statistically significant differences were found between participants based on gender or academic background, suggesting that the difficulties with modelling are widespread among elementary PSTs regardless of their demographic characteristics. In contrast, the AI system produced solutions that were generally more coherent and well-structured, particularly in terms of validation and creativity. The AI was able to provide step-by-step explanations and generate multiple approaches to solving the same problem, which could serve as a model for PSTs. Nevertheless, the AI responses were not flawless, and in some cases, over-simplifications or assumptions were made without sufficient justification.
Conclusion: The study concludes that while PSTs have some foundational skills in mathematical modelling, their overall competencies are insufficient for solving complex modelling tasks independently. This finding points to a need for teacher education programs to place greater emphasis on developing modelling skills through practice, reflection, and guided instruction. At the same time, the study highlights the potential of AI as a valuable educational resource. AI can accelerate the learning process by providing worked examples, alternative solutions, and creative perspectives on modelling tasks. However, the findings also caution against over-reliance on AI. If PSTs depend too heavily on AI-generated solutions without critical evaluation, their own problem-solving and reasoning abilities may be undermined. Therefore, AI should be integrated into teacher education thoughtfully, with clear guidelines for its use as a supportive rather than substitutive tool. By balancing human learning with technological support, teacher educators can better prepare future teachers to engage their students in meaningful mathematical modelling activities.
Keywords
Main Subjects
COPYRIGHTS
© 2026 The Author(s). This is an open-access article distributed under the terms and conditions of the Creative Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) (https://creativecommons.org/licenses/by-nc/4.0/)
', './data/jte/coversheet/cover_en.jpg'))
Send comment about this article