Hong Kong-French Collaboration in Machine Learning Education for Junior Secondary Students: A Mixed-Methods Study on Teacher Development

Abstract

This exploratory mixed-methods study investigates the
effectiveness of a teacher development (TD) workshop,
collaboratively designed by Hong Kong and French educators,
to enhance machine learning (ML) literacy among French
secondary teachers. The workshop, conducted with 11
participants from diverse schools in the
Auvergne-Rhône-Alpes region, utilised the Technological
Pedagogical Content Knowledge (TPACK) framework and
hands-on activities involving AlphAI learning robots. Over
a six-hour in-person session, teachers engaged in
robot-assisted activities, reflective discussions, and
scaffolded resources aligned with core ML concepts,
including supervised learning using k-nearest neighbours
(KNN) and artificial neural networks (ANN). Quantitative
analyses revealed statistically significant improvements in
teachers' ML conceptual understanding (p < 0.01),
particularly on challenging topics such as overfitting,
alongside marked gains in self-efficacy across TPACK
domains (TCK, CK, PCK, and TPACK; all p < 0.01).
Qualitative data from reflections and interviews
highlighted teachers' cognitive growth, increased
confidence in integrating robots into ML instruction, and
affective engagement with the workshop's experiential
design. Participants also identified practical challenges,
including time constraints and resource management for
classroom implementation. The study underscores the value
of contextualised, technology-enhanced TD workshops that
blend theoretical frameworks with hands-on tools to build
teacher capacity for ML education. Findings suggest that
adaptable TD models can support scalable teacher
preparation for AI technologies in junior secondary
settings. Limitations of this study include the small
sample size, which warrants cautious interpretation; we
shall conduct larger cohorts of TD in France in the future.