AI-Enhanced Blended Learning in EFL Higher Education: Effects on Learner Autonomy and Language Performance in the Kazakhstani Context

Abstract Book of the 11th International Conference on Advanced Research in Teaching and Education

Year: 2026

[PDF]

AI-Enhanced Blended Learning in EFL Higher Education: Effects on Learner Autonomy and Language Performance in the Kazakhstani Context

Guldana Seidaliyeva, Uldana Baidabek

ABSTRACT:

The rapid advancement of artificial intelligence (AI) has transformed higher education, creating new opportunities for personalized learning and adaptive feedback which are integrated into blended learning environments, yet empirical evidence from Central Asian higher education remains scarce, regarding learner autonomy and separate dimensions of language performance complexity, accuracy, and fluency (CAF).
This study examined the effects of AI-enhanced blended learning on learner autonomy and on speaking and writing performance among 60 first-year EFL students at Abai Kazakh National Pedagogical University in Kazakhstan. A mixed-methods quasi-experimental design compared an experimental group receiving AI-supported blended instruction (ChatGPT, Grammarly, QuillBot) with a control group receiving conventional blended instruction over twelve weeks.
Quantitative data included pre- and post-tests rated with an analytic CAF framework and a learner autonomy questionnaire; qualitative data comprised classroom observations and semi-structured interviews. ANCOVA results, controlling for pre-test scores, showed that the AI-enhanced group achieved significantly higher post-test autonomy and overall language performance. Disaggregated analyses indicated large gains in accuracy and fluency, reflected in fewer grammatical and lexical errors and more continuous oral and written production, while improvements in complexity were positive but comparatively modest.
Qualitative findings revealed that AI-supported environments fostered self-regulation, independent learning, and increased language use, but also occasional tendencies toward reliance on AI-generated phrasing. The study concludes that AI can serve as an effective pedagogical mediator to enhance accuracy, fluency, and autonomy when embedded in purposeful, teacher-guided blended learning, while higher-order complexity still requires targeted instruction and rich communicative practice.

Keywords: Artificial Intelligence; Blended Learning; Learner Autonomy; English As A Foreign Language; ChatGPT