Abstract
This research explores the use of ChatGPT to enhance argumentative essay writing skills for undergraduate English as a Foreign Language (EFL) learners and develop pedagogical guidelines for instructors. Argumentative writing is especially difficult for novice undergraduate EFL learners as they must deal with grammar, vocabulary, and stylistic problems while developing a good argument structure. A one-year Design-Based Research (DBR) study was designed to develop an EFL argumentative chatbot assistant within ChatGPT. The process began with a comprehensive literature review followed by expert input and student feedback to embed an argumentative writing GPT assistant in a writing course guided by Bloom’s taxonomy and seamless language learning. A key feature of the study was the use of ChatGPT prompts for smart questioning and clarification, followed by constructive feedback. This process enabled students to develop their ideas and deepen their understanding of the writing process. The study consisted of two main phases with 24 students and included pre-post writing tests, and chat analysis to understand improvements in writing. Surveys, interviews, and observations were also used to help understand student perceptions and problems. The results demonstrated improvements in students’ grammatical accuracy, organization, argument development, and counterargument skills. Overall, the students found the learning experience to be valuable and helpful, resulting in perceived motivation, and changes in writing behavior. However, students feared they would become over-reliant on AI for writing and struggled to assess the quality of the AI feedback. The research led to pedagogical guidelines for integrating ChatGPT into undergraduate EFL writing instruction.
Presenters
Neil BarrettAssociate Professor, Center for Bilingual Education, Southern Taiwan University of Science and Technology, Tainan, Taiwan
Details
Presentation Type
Paper Presentation in a Themed Session
Theme
KEYWORDS
ChatGPT, EFL learners, Argumentative writing, Seamless language learning, Bloom’s taxonomy