Modality matters: Eye-tracking evidence of attentional engagement in second language feedback processing

Abstract

While recasts are recognized as effective for L2 learning, their efficacy in synchronous computer-mediated communication (SCMC), particularly videoconferencing, remains underexplored. Videoconferencing enables multimodal feedback, which according to Mayer's (2014) Cognitive Theory of Multimedia Learning, can enhance processing and retention by combining visual and auditory input. Yet, empirical comparisons of multimodal versus unimodal feedback are limited (Jung & Révész, 2024), and little is known about how feedback modality affects learner attention and learning outcomes in video-based interaction. To fill these gaps, this study compared the effects of unimodal (oral or written) and multimodal (oral plus written) recasts on grammatical development and learner attention, with attention tracked via eye-tracking. Two questions guided the study: (1) How does recast modality affect L2 grammatical development during videoconferencing tasks? (2) How does recast modality affect visual attention, and does the relationship between attention and grammar learning vary across modalities? The study employed a pretest–posttest–delayed posttest design, with 120 Korean learners of English assigned to four conditions: oral, written, multimodal, or no recasts. They completed three videoconferencing tasks targeting the third person singular and present continuous forms. Written recasts were presented as real-time captions. Attention was recorded during the treatments, and development was assessed through oral, written, and spontaneous oral production tasks. Mixed-effects analyses revealed that the multimodal group significantly outperformed the others, displaying greater attention to the target forms. The implications of the findings will be discussed for inclusive feedback strategies tailored to diverse learners in online language education.