Development of Mathematics Learning Design Using AI-Based Deep Learning Approach to Enhance Student’s Conceptual Understanding

Source abstract

This study aims to develop a deep learning-based mathematics learning design integrated with artificial intelligence for fractions, intended to improve third-year pupils’ conceptual understanding, as well as to test its validity, practicality and effectiveness. A research and development approach was employed, following five stages: analysis, design, development, implementation and evaluation. The instructional design, developed using ChatGPT integrated with the Polypad application, was validated by three experts, whilst its effectiveness was tested on 22 Year 3 pupils using a single-group pre-test–post-test procedure analysed with the Wilcoxon Signed Ranks test, and its practicality was measured through a pupil feedback questionnaire. The findings indicate that the instructional design was highly valid according to the experts’ assessment, and that the pre-test and post-test instruments were both valid and reliable. There was a significant increase between pre-test and post-test scores, indicating that the design was effective in enhancing pupils’ conceptual understanding of fractions. Student feedback also indicated that the design was practical to use. These findings suggest that a systematically designed learning approach integrated with artificial intelligence can meaningfully support the conceptual learning of fractions in early primary school mathematics education.