Exploration of Students' Computational Thinking on Geometric Transformations Through AI-Assisted Geogebra
TL;DR
The results show that students demonstrated strong abilities in decomposition and pattern recognition, while abstraction and algorithmic thinking remained the main challenges, especially in complex transformations.
Abstract
Computational thinking (CT) is a key competence in 21st-century mathematics learning, particularly in geometry, which involves spatial reasoning and problem solving. This study aims to explore how students’ CT develop when solving geometric transformation problems through artificial intelligence (AI)-assisted GeoGebra learning. The study employed an Educational Design Research (EDR) approach involving 18 junior high school ninth-grade students. The learning design followed a hypothetical learning trajectory with three increasingly complex challenges: translation with fixed orientation, translation with varied positions, and a combination of translation and rotation. Data were collected from student artifacts, including worksheets, AI prompts, GeoGebra commands and screenshots, teacher observations, and exit tickets, and analyzed qualitatively using deductive coding based on four CT components. The results show that students demonstrated strong abilities in decomposition and pattern recognition, while abstraction and algorithmic thinking remained the main challenges, especially in complex transformations. GeoGebra and AI primarily functioned as tools for reflection and debugging, rather than as direct solution providers. This study highlights the importance of scaffolded, challenge-based learning to support CT development in geometry. However, the findings are limited to a single learning session with a small sample size.