AI Affordances and EFL Learners' Speaking Engagement: The Moderating Roles of Gender and Learner Type

Abstract

Contextualised in the AI–supported English-speaking learning, this study examined the roles of AI affordances in influencing EFL learners' emotional, cognitive, and behavioural speaking engagement, and explored the moderating roles of gender and learner types (on-campus vs. on-job) in influencing AI-supported English-speaking engagement. Data collected from 332 Chinese EFL learners (159 on-campus and 173 on-job learners) were analysed by using structural equation modelling. Results indicated that Chinese EFL learners perceived AI affordances to be significant in influencing their emotional, cognitive and behavioural engagement in practicing their spoken English. The results from the PLS-SEM model revealed that AI affordances accounted for 54.7%, 52.4% and 56.0% of the variance in emotional engagement, cognitive engagement and behavioural engagement, respectively. Learner type was not found to significantly moderate the relationships between AI affordances and speaking engagement. Gender was found to be a significant moderator for the AI affordances–behavioural engagement and AI affordance–cognitive engagement relationships. These findings enrich existing literature about AI–empowered speaking engagement and provide practical implications for English teachers to design effective speaking-teaching models.