Machine learning and educational robotics, an implementation in initial university teacher training and for practicing teachers in primary education

Sara Redondo-Duarte¹Daniel Pattier-Bocos²Adrian Neubauer¹Jose-Manuel Sáez López^1*

• ¹National University of Distance Education (UNED), Madrid, Spain
• ²Universidad Complutense de Madrid, Madrid, Spain

Programming and robotics-based training programs have been shown to enhance computational thinking and self-efficacy, yet limited attention has been paid to preparing future teachers to effectively implement these methodologies in educational practice. This study analyzes the impact of an educational intervention based on visual block programming, educational robotics, and machine learning on the initial training of pre-service teachers and in-service primary school teachers. A quasi-experimental design was employed. The sample consisted of 170 participants: 83 students enrolled in a Primary Education degree program and 87 in-service primary school teachers. The instructional procedure was implemented through hands-on activities with micro:bit, Maqueen robots, and introductory machine learning concepts. Data were collected using a coding, robotics, and machine learning knowledge test, along with several validated Likert-type scales to assess attitudes toward the curricular integration of these technologies. For the first dimension, Student's t-tests and linear regression analyses were conducted, while correlation analyses and nonparametric tests were applied to the second dimension. The findings revealed significant improvements in the acquisition of basic computational concepts (sequences, loops, and conditionals) and in the understanding of machine learning, with university students outperforming practicing teachers. Comparative tests indicated a greater self-perception of technological competence among university students, particularly in block-based programming and the use of game engines for educational purposes. The results suggest that the structured integration of robotics and machine learning appears to constitute a viable and effective strategy for enhancing teacher training, promoting active methodologies, and fostering an interdisciplinary approach in primary education.