An Empirical Study on the Utilization of the Large Language Model, in English Education

The study investigated the impact of a Large Language Model, on English learning among 20 undergraduate students enrolled in an English course in 2023. Data were collected through a survey that asked students about their opinions on using ChatGPT. The survey included 20 structured questions and also asked for open-ended responses to explore the advantages and disadvantages of using the large language model for English learning. In addition, the survey examined students’ perceptions of the importance, necessity, difficulty, and interest of various English skills, asking them to rank them. Paired samples t-tests were used to analyze significant changes before and after the intervention. The results indicated significant improvements in learning effectiveness, a decrease in anxiety and stress, increased motivation and confidence, and enhanced participation. The students reported that using ChatGPT led to improvements in their vocabulary, grammar, reading, and writing skills. They also noted a slight improvement in their communication skills. Interestingly, the students perceived a decline in their listening skills after the intervention. However, they expressed a greater willingness to continue using ChatGPT for future learning, appreciating its accessibility, flexibility, and the enriching learning experience it provided. Lastly, the study identified challenges, including a lack of support for in-depth grammar, conversational practice, and exam preparation, as well as a potential overreliance on the tool. Therefore, the use of ChatGPT requires appropriate application based on the learning objectives.