Computational thinking and global competence in higher education: insights from machine learning analysis

Introduction Universities face the challenge of preparing increasingly diverse students to solve complex problems in global environments. This study analyzes key indicators related to global competency development (GI), including understanding environmental challenges, promoting sustainability, and valuing intercultural perspectives as well as computational thinking (CT) competence for analytical, synthetic and problem-solving skills. Method Verified if there is a relation between both competences. Data was collected from international users of an open-access platform oriented towards CT development. Responses from pretests, posttests, and user feedback were processed through supervised machine learning algorithms using scaling and balancing techniques to ensure robustness. A five-fold cross-validation confirmed consistent metrics. The Random Forest model and Logistic Regression algorithm were applied. Results suggest that (a) global and intercultural competencies can be accurately predicted based on individual skills such as reflection, adaptability, and prior experience (b) CT-High can be related to problem decomposition, solution execution, and strategic reasoning are key to this prediction and (c) although the country variable provides contextual insights; it does not replace personal competencies as a determinant of success, and nonlinear relationship between these competencies. Discussion The findings highlight the potential of global, technology-mediated interactions to cultivate critical, creative, and globally engaged citizens.