Comparative Effectiveness of Artificial Intelligence Tools in Improving English Language Learning Outcomes

Abstract

Artificial intelligence (AI) has emerged as an important technology for enhancing English language education by providing personalized learning experiences, adaptive feedback, and continuous learner support. This study evaluated the comparative effectiveness of four widely used AI tools—ChatGPT, Grammarly, Duolingo, and ELSA Speak—in improving English language learning outcomes. A quantitative, comparative, cross-sectional design was employed using a publicly available dataset comprising 15,000 learner records. The analysis examined learning gain, language error reduction, learner satisfaction, and AI usage behavior through descriptive and comparative statistical methods. The findings revealed that all AI tools positively contributed to English language learning by improving learner performance and reducing language errors. Duolingo achieved the highest average learning gain, while ChatGPT demonstrated the greatest reduction in language errors. Learner satisfaction remained consistently high across all platforms, with only minor differences observed. Furthermore, learners who used AI tools more frequently showed slightly higher learning gains than less frequent users, indicating that regular engagement enhances educational outcomes. Overall, the results suggest that although individual AI platforms possess distinct instructional strengths, each effectively supports language acquisition through personalized feedback and adaptive learning. The study provides empirical evidence to support the integration of AI-assisted technologies into English language education and offers practical insights for educators and institutions seeking evidence-based strategies to improve language learning. Future research should investigate long-term learning outcomes and evaluate emerging AI technologies across diverse educational settings.