Proceedings of the Annual Meeting of the Cognitive Science Society

This study examined whether learner-generated multimodal videos enhance L2 word retention beyond multimodal exposure alone and beyond traditional teacher-led instruction. To isolate production from exposure, 164 Chinese university EFL learners were taught 39 unfamiliar words in one of three conditions: video creation (each learner produced two short videos integrating visual, auditory, linguistic, bodily, and affective resources), video viewing (simply watched the same videos), or traditional instruction (teacher-led explanations and examples). Retention was assessed with immediate and delayed post-tests, with previously known items excluded before instruction. Both video-based conditions outperformed traditional instruction, with the largest gains in the video creation condition. The creation advantage extended beyond self-authored items: Creators also outperformed viewers on words encountered only through subsequent in-class viewing, suggesting that creating videos alters attention to and use of multimodal cues during later viewing. Brief learner-generated multimodal production therefore supports deeper encoding and more durable lexical learning than exposure alone.