Research Trends in Mapping Engagement in AI-Supported Multimodal Language Learning: A Network Analysis of Psychological Factors
TL;DR
It is suggested that engagement in AI-supported multimodal language learning is structurally associated not only with positive learning experiences but also with learners’ adaptive beliefs, perceived capability, and technological competence.
Abstract
As artificial intelligence (AI) increasingly reshapes language learning through multimodal environments involving text, audio, and images, recent research has called for more dynamic and system-oriented approaches to understanding learner engagement. Against this background, the present study examined the relationships among AI literacy, growth mindset, self-efficacy, flow, and engagement in AI-supported multimodal language learning. Guided by Sociocultural theory (SCT), data were collected from a sample of 776 participants and analyzed using network analysis. The results revealed a fully connected and entirely positive network structure, suggesting that these variables function as closely related components of a broader learner-related psychological system. Among the five nodes, engagement emerged as the most central position, as indicated by the highest values of strength, closeness, betweenness, and expected influence. In addition, growth mindset, self-efficacy, and AI literacy showed comparatively stronger connections with engagement, whereas flow was positively associated with engagement but to a lesser extent. These findings suggest that engagement in AI-supported multimodal language learning is structurally associated not only with positive learning experiences but also with learners’ adaptive beliefs, perceived capability, and technological competence. The study contributes to the research trends in understanding AI engagement by offering a novel network-based perspective on engagement in contemporary AI-mediated language learning.