Unveiling the effects of GenAI on motivation of learning English as a second/foreign language: Evidence from meta-analysis

Abstract

Despite growing interest in the use of generative artificial intelligence (GenAI) in second language education, its impact on learners' motivation in English as a second or foreign language (ESL/EFL) contexts remains unclear due to inconsistent findings and limited theory-based synthesis. This meta-analysis synthesized 47 effect sizes from 24 empirical studies (N = 7075) to examine the effects of GenAI-supported interventions on ESL/EFL learner motivation. A three-level model revealed a moderate and statistically significant overall effect (g = 0.611, p < .001). Stronger effects were observed for motivation types such as the Ideal L2 Self, L2 Learning Experience, and self-efficacy, while intrinsic, extrinsic, and Ought-to L2 Self motivation showed non-significant effects. Moderator analyses based on the usefulness framework identified three significant contextual factors: communication format , use context, and ESL/EFL learning setting. Specifically, multimodal formats combining text and voice yielded stronger effects than single-mode formats, in-class applications were more effective than after-class use, and GenAI produced larger motivational gains in EFL settings than in ESL ones. These findings offer theory-informed insights and practical guidance for developing GenAI-supported language learning interventions that align with learners’ motivational profiles and instructional contexts.