Assessing multidimensional student engagement in studio-based design learning: a traceable multimodal framework for...

Student engagement is a central construct in educational psychology, but it is difficult to assess in studio-based design learning because participation, emotion, cognitive effort, self-regulation, and design-process activity are distributed across classroom interaction, learning records, written reflection, and design artifacts.

This study presents PD-EviMEA as a proof-of-concept and modular feasibility framework that organizes such traces into teacher-reviewable evidence for engagement assessment. It represents visual, behavioral, textual, and design-process information as learning evidence, estimates modality reliability and availability, and links each diagnostic statement to explicit evidence units rather than treating model outputs as self-sufficient judgments.

Across several public datasets, the results provide preliminary modular evidence that multimodal traces can support engagement recognition, that learning logs and design-process sequences can function as contextualized indicators of behavioral and cognitive engagement, and that evidence-constrained feedback can preserve alignment with human scoring criteria while adding traceability.

Because the datasets were not collected from a single product design studio and several analyses rely on proxy tasks, the findings should be interpreted as public-dataset-based feasibility evidence rather than validation of a deployable classroom assessment system. The study clarifies how multidimensional engagement may be operationalized in studio-based design contexts while preserving teacher agency, construct caution, and transparency in technology-supported assessment.