Evaluating the Impact of an AI-Powered Blended Learning Platform on Students’ Business English Performance: A MANCOVA Approach

Background and Aim: The integration of artificial intelligence (AI) in education has significantly transformed language learning, particularly in blended learning environments. AI-powered platforms such as FLIT provide personalized learning experiences, real-time feedback, and data-driven insights, making them valuable tools for Business English instruction. However, factors such as gender and geographical background may influence student performance in these AI-driven environments. This study examines the impact of gender and geographical background on Business English proficiency (listening, speaking, reading, and writing) within an FLIT-based blended learning framework, and it explores the potential interaction between these factors. Materials and Methods: Using a quasi-experimental design with pre-test and post-test assessments, 240 English major students from a science and technology university in northeastern China participated in ten weeks of FLIT-based instruction. Pre-tests and post-tests were administered using the Cambridge Business English Certificates (BEC) exam. Data were analyzed using ANCOVA to assess the main effects of gender and geographical background, and MANCOVA to investigate interaction effects. Results: Results indicate that geographical background significantly influences Business English proficiency across all four language skills. For example, ANCOVA revealed that geographical background had a significant effect on reading performance (F (3,232) = 17.31, p < 0.001, partial η² ≈ 0.19), with urban students scoring approximately 10% higher than their rural counterparts. In contrast, gender did not exhibit a statistically significant effect (all p > 0.05), and no interaction effect between gender and geographical background was observed. Conclusion: The findings underscore the need to address regional disparities in AI-powered language learning environments. Practically, these results suggest that targeted interventions—such as enhanced digital literacy programs and increased allocation of educational resources to rural areas—are essential for bridging the performance gap. Policy-wise, investments in digital infrastructure and tailored educational technology are recommended to ensure equitable learning opportunities across diverse regions. Future research should investigate the long-term impacts of AI-powered instruction and consider additional learner variables such as motivation and learning styles.