Impact of Artificial intelligence tools on learning motivation in English instruction: a network meta-analysis

Abstract The rapid expansion of English-Medium Instruction (EMI) programs and the growing demand for English language proficiency have introduced significant motivational challenges for EFL learners who face linguistic barriers. Artificial intelligence (AI) tools offer a potential solution by providing personalized scaffolding and interactive support. This study conducts a network meta-analysis (NMA) of 16 empirical studies (N = 1,923) to compare the effectiveness of three AI interventions, including Generative AI Chatbots, AI Writing Assistants, and AI Language Learning Applications, on the learning motivation of K-12 and university EFL students in English instruction contexts. Results indicate that all AI tools significantly outperform traditional instruction. AI Language Learning Applications (g = 0.907, k = 2) showed the largest effect, followed by Generative AI Chatbots (g = 0.824, k = 11) and AI Writing Assistants (g = 0.692, k = 3). These findings suggest that interactive and adaptive AI tools are particularly effective in fostering motivation by fulfilling students' basic psychological needs of autonomy and competence. The study provides evidence-based recommendations for the strategic integration of AI into English instruction curricula to optimize student engagement and academic success.