Developing computational thinking in teacher education: a brief experience with coding, robotics, and machine learning
TL;DR
It is suggested that well-structured microinterventions may support higher self-reported knowledge of programming and machine learning, greater familiarity with block-based programming and game engines, and more positive perceptions of active learning in teacher education.
Abstract
This study examines post-intervention self-reported outcomes associated with a short intervention designed to introduce coding, educational robotics, and introductory concepts of machine learning into initial teacher education for Primary Education. The proposal is grounded in Papert’s principles and the Creative and Computational Problem Solving model, combining active learning methodologies with accessible technologies. A quasi-experimental design with a non-equivalent control group was employed. A total of 144 first-year students from three Spanish universities participated (105 in the experimental group and 39 in the control group). The intervention consisted of a two-hour workshop with hands-on stations and technological challenges, designed as a feasible microintervention within regular teacher education contexts. Assessment was carried out through an ad hoc questionnaire focused on three dimensions: knowledge of programming and machine learning, block-based programming and game engines, and active learning, all assessed through self-reported perceptions. The results showed statistically significant post-intervention differences in favor of the experimental group, with a small effect size for knowledge of programming and machine learning and large effect sizes for block-based programming and game engines and active learning. Item-level analyses revealed that favorable overall perceptions coexisted with lower self-reported familiarity in specific technical areas, particularly machine learning, prior experience with block-based programming, and programming in game engines. Three exploratory response profiles were also identified: reluctant, intermediate, and enthusiastic. The findings suggest that well-structured microinterventions may support higher self-reported knowledge of programming and machine learning, greater familiarity with block-based programming and game engines, and more positive perceptions of active learning in teacher education.