Enhancing Computational Thinking Through STEAM-AI Learning

Computational Thinking (CT) is an essential competency in twenty-first-century education, especially in interdisciplinary learning based on Science, Technology, Engineering, Arts, and Mathematics (STEAM). This study aims to analyze the development of CT skills in prospective mathematics teachers through STEAM learning enhanced by Artificial Intelligence (AI) technology and to identify personal and instructional factors that influence CT achievement. The research used mixed methods with a quasi-experimental design and a reflective qualitative study involving 67 students from one of state university in Indonesia. The Teacher Computational Thinking Scale (TCTS) instrument was used to measure five dimensions of CT, and qualitative data were collected through reflective journals and in-depth interviews. The results showed that STEAM-AI learning significantly improved CT skills in all dimensions: problem-solving, algorithmic thinking, creativity, critical thinking, and collaboration. Regression analysis identified AI literacy, digital experience, and project participation as the main predictors of CT improvement. Qualitative findings highlighted the transformation of students’ perceptions of mathematics learning, the strengthening of reflective capacity, and the emergence of professional awareness as CT-based learning designers. This study contributes to the development of computational pedagogy in teacher education and offers innovative curriculum recommendations that integrate AI-based STEAM activities reflectively and contextually.