Designing Multimodal Writing Assessments for Beginner Chinese Learners

Abstract

This classroom-based study investigates how structured multimodal writing tasks can support beginner learners of Chinese. Four tasks—vocabulary matching, image-supported sentence writing, gap-fill completion, and sentence reordering—were designed to gradually increase cognitive load and reduce external support. Ten first-year undergraduates completed these tasks at the end of a semester-long Chinese elective. Data sources included written outputs, Padlet logs, student reflections, and interviews. Tasks that combined visual cues with sentence-level scaffolds led to more confident and accurate writing. In contrast, grammar-heavy tasks with minimal support resulted in lower accuracy and increased cognitive overload. Drawing on cognitive load theory and dual coding theory, the study proposes a three-phase model: perceptual support, linguistic scaffolding, and output assessment. A rubric assessing language accuracy, content coherence, and peer interaction demonstrated strong inter-rater reliability (κ = .82). The findings offer practical insights into how multimodal tools can be used to scaffold sentence-level writing and provide low-tech assessment alternatives for under-resourced classrooms. The model can be adapted across various proficiency levels and teaching contexts.