machine-learning
Filtering by topic machine-learning(2)Clear all filters
- PaperComputers and Education: Artificial Intelligence16 Jun 2026
Fostering machine learning literacy in senior primary education: Evaluating a structured pedagogical course design
Siu Cheung Kong, Qiaoyi Wang
A structured 6-8 hour machine learning course for senior primary students (age 11) significantly improved their understanding of supervised and reinforcement learning, including algorithms like k-nearest neighbours and artificial neural networks. The pedagogical approach combined guided worksheets, hands-on robot training, and real-time algorithm visualization, leading to enhanced conceptual learning and student engagement.
Original abstract
Current K–12 artificial intelligence (AI) literacy education emphasizes tool usage over fundamental concepts, yet AI literacy requires grasping how and why AI works – critical for an AI-driven society. This highlights the need for machine learning (ML) education for young learners. We designed and evaluated a six-to 8-h ML course for 752 senior primary students (average age 11.36) across seven Hong Kong primary schools. Pre- and post-test results showed significant improvement in ML understanding, with a Wilcoxon effect size of 0.55. Students comprehended supervised learning and reinforcement learning, including algorithms such as k-nearest neighbours and artificial neural networks, via training robots in competitive circuit tasks and real-time algorithm visualization. Thematic analysis of student interviews revealed that our structured pedagogical approach — blending guided worksheets, hands-on activities, and iterative refinement of data processing, parameter adjustment, and model training within the robots' software — enhanced students’ conceptual learning and engagement. Surprisingly, they developed initial reflections on distinguishing between AI and human learning. These findings suggest the feasibility and promise of teaching fundamental ML concepts to senior primary students through a structured course design. The study contributes to future research and practice in fostering ML literacy among young learners, providing actionable insights for educators to support and allocate resources.
- PaperJournal of Learning Analytics25 Feb 2026
Learning Analytics to Uncover Ethnic Bias in Educational Texts
Josmario Albuquerque, Bart Rienties, Martin Hlosta, Wayne Holmes
Used learning analytics and machine learning to identify ethnic bias in text-based educational materials, finding that social science content correlated with perceived bias. Support vector machines and random forest classifiers achieved consistent F1-scores of 0.71 and 0.70, while naive Bayes showed highest precision. This work provides a foundation for automated bias detection to support fairer educational technologies.
Original abstract
Online learning platforms have expanded access to education but also raise concerns about biased content, particularly in text-based learning materials such as textbooks, lesson plans, and course excerpts. Such biases can perpetuate discrimination, can harm student outcomes, and can often be difficult to detect, as identification typically relies on time-consuming human review. Learning analytics (LA) can enhance this process by supporting human reviewers through automated detection, offering a scalable solution while retaining human judgment for nuanced evaluations. Accordingly, this LA study explores two research questions: RQ1: Which features might support the identification of ethnic bias in text-based online learning materials? and RQ2: Which classification approaches might be suitable for identifying ethnic bias in text-based online learning materials? First, we identified features signalling potential ethnic bias (presence or absence) in textual content using a dataset (N = 345) labelled by 193 students from diverse ethnic backgrounds. Then, we evaluated multiple machine learning (ML) models for their effectiveness in bias classification. The results suggest significant correlations between perceived bias and content from social sciences. Additionally, through bootstrap analysis, support vector machines and random forest classifiers showed consistent performance in bias identification (with F1-scores of 0.71 and 0.70 on the test set, respectively). In contrast, the naive Bayes (NB) model demonstrated the highest precision (0.75 on the test set). We discuss these findings and their implications for LA, emphasizing the importance of quality and inclusive educational tools. As an initial step toward automated bias classification in education, this study provides a foundation for spotting ethnic bias in learning content, supporting fairer technologies for more inclusive learning environments.