Human- or AI-delivered praise? The moderating role of praise publicity in synchronous video-based instruction
摘要
Unlike prerecorded video lectures, synchronous video-based instruction allows real-time teacher–student and student–student interaction. With the rapid development of artificial intelligence (AI), AI systems can also deliver praise with high frequency and immediacy. However, little is known about how praise source and publicity jointly shape learning outcomes in synchronous video-based instruction. Grounded in a dual-pathway framework (i.e., social–motivational and cognitive–behavioral processes), this study examined the interaction effect of praise source (human vs. AI) and publicity (public vs. private) on immediate learning in a short-duration EFL vocabulary task during in synchronous video-based instruction. Data were collected from 220 high school EFL students who completed a synchronous video-based instruction integrating a human assistant and an AI assistant into translation tasks, using scales, voice recording, and an immediate learning performance test. The results revealed a differentiated pattern of effects across conditions. Public human-delivered praise was more effective in enhancing learners’ motivation and response accuracy, whereas private AI-delivered praise better supported response length, response accuracy, and immediate learning performance. Overall, the findings indicate that the effects of praise are context-dependent and jointly shaped by source and publicity. Human-delivered praise is more effective in enhancing motivational outcomes in public contexts, whereas AI-delivered praise is more effective in supporting cognitive engagement and immediate performance in private contexts. These findings highlight the importance of considering both praise source and publicity in designing instructional feedback for short-duration EFL vocabulary-learning tasks in synchronous video-based instruction.