Investigating the use of artificial intelligence supported tools by preservice middle school mathematics teachers in the numbers and quantities theme

Abstract

Technology and artificial intelligence (AI) have increasingly influenced teaching and learning processes. With their integration into education, AI-based tools have begun to alleviate instructional challenges faced by teachers and preservice teachers, including lesson planning, activity design, assessment, and mathematical problem posing. One of the central themes of the middle school mathematics curriculum, Numbers and Quantities, has long been emphasized in mathematics education research due to its conceptual complexity and instructional significance. The use of AI tools within this theme is therefore expected to support instruction and instructional design. Accordingly, this study aimed to examine the use of two widely adopted AI tools—ChatGPT and Gemini—by preservice middle school mathematics teachers. The study focused on how these tools were used in relation to the Grade 6 learning outcome defined in the Turkish Century Education Model: solving real-life problems that require the four basic operations involving fractions, decimals, and percentages. A qualitative case study design was employed with 45 fourth-year preservice teachers enrolled in an elementary mathematics teacher education program at a public university. Data were collected through problem-posing and problem-solving tasks and analyzed using descriptive analysis. The findings showed that most problems constructed independently by preservice teachers did not fully meet the specified learning outcome, whereas a lower proportion of misaligned problems was observed in those generated using ChatGPT and Gemini. However, preservice teachers experienced difficulties in writing effective prompts, and AI-generated problems frequently relied on repetitive contexts, outdated numerical values, or incomplete data. These results emphasize the need for critical evaluation and validation of AI-generated outputs for instructional use.

Ähnliche Arbeiten