Virtual Booth

As computational thinking (CT) gains prominence in primary education, there is an increasing need to personalize learning experiences in block-based programming environments like Scratch. This innovation constructs and proposes a data-driven learner typology framework to support differentiated instruction based on Scratch projects. Twelve student-created Scratch projects were analyzed using the Scratch Evaluation Framework. The hierarchical clustering analysis was conducted to identify patterns of CT strength. The findings revealed three distinct learner profiles: (i) Logic-Dominant, (ii) Synchronization-Dominant, and (iii) Interaction-Dominant, each exhibiting unique patterns of CT concept mastery and associated CT practices such as abstraction, decomposition, and iterative design. This learner typology model holds strong potential as the foundation for a diagnostic toolkit or adaptive learning platform for coding instruction.