




































BECE. Newsletter 

Vol.21, No.2, 2025, BECE 2051 

 

 

 

 

The Role of Generative Artificial Intelligence in the 

Teaching and Learning of English as Foreign Lan-

guage: An Empirical Study of GenAI-Based Feedback 

in Secondary English Writing 

By Liu, M., Zou, W., Zong, M., & Yang, B. 
Correspondence to: Bing Yang, Hubei Performance Evaluation Information 

Management Research Center, China. E-mail: editorial-office@bonoi.org   

HIS STUDY is an evaluation of the effect of GenAI-based feedback on 

English writing performance of secondary school students as English-as-

foreign-language (EFL) learners, based on experimental data from 201 sev-

enth graders from four classes at J Middle School in Wuhan, Hubei Province, 

China. Using the GenAI software “Doubao,” which is built on the Skylark 

model, as the experimental platform, the researchers compared the English 

writing scores of the experimental group (trained with GenAI-based feed-

back) and the control group (trained with teacher feedback) across four writ-

ing assignments.  

Research Findings: 

• GenAI-based feedback was more effective than teacher feed-

back in improving the writing’s relevance to the topic in the students with 

ordinary writing skills, while neither GenAI-based feedback nor teacher 

feedback showed significant improvement effect on this aspect in the high-

performing students. 

• Compared to teacher feedback, GenAI-based feedback led to 

greater improvement in grammatical accuracy in English writing. 

• GenAI-based feedback was more effective in improving 

wording appropriateness of students’ English writing than teacher feedback. 

• GenAI-based feedback significantly enhanced the logic of 

narrarive in students’ English writing, whereas teacher feedback showed no 

measurable effect in this regard. 

The research results imply that the introduction of a real-time feed-

back mechanism is beneficial for students being promptly informed of their 

performance and deficits in English writing in terms of relevance, grammar, 

wording, and logic, enabling targeted revisions and improvements; that 

GenAI-based feedback is more suitable for beginners in English writing; and 

that GenAI-based feedback has the potential to narrow the gap between stu-

dents with different levels of English writing proficiency. 

 

T 

mailto:editorial-office@bonoi.org


 

Vol.21, No.2, 2025, BECE 2052 

Source: China Educational Technology, 2025; 2025(06): 70-79. 


