본문 바로가기
  • Home
Basic Search

Researching Academic Writing Practices and Pedagogy with ChatGPT – A Case Study of H University's “Logical Thinking and Writing” Course

  • Korean Language & Literature
  • 2024, (128), pp.265-302
  • DOI : 10.21793/koreall.2024.128.265
  • Publisher : Korean Language & Literature
  • Research Area : Humanities > Korean Language and Literature
  • Received : September 30, 2024
  • Accepted : November 19, 2024
  • Published : November 30, 2024

lee mi na 1

1서울대학교 국어교육연구소

Accredited

ABSTRACT

The advent of generative AI technologies, including ChatGPT, has brought transformative changes to education, necessitating an evolution in pedagogical approaches to integrate these tools effectively. Recognizing the potential of AI, universities and governmental bodies have begun implementing policies that embrace generative AI while safeguarding learners' rights to utilize such technologies. In this context, the debate surrounding AI and its role in writing instruction has intensified, prompting both a reevaluation of traditional writing pedagogy and exploration into innovative uses of these tools. University-level writing education primarily aims to cultivate critical thinking and academic inquiry, which are essential for producing true knowledge through academic writing. Given this objective, there is a growing need to investigate how tools like ChatGPT can be practically integrated into the academic writing process without compromising the development of independent thinking skills. This study examines this question through an empirical investigation of students enrolled in Logical Thinking and Writing, a required liberal arts course at H University in Seoul, South Korea. The course was designed to incorporate ChatGPT into the academic writing process, with a focus on understanding students' experiences and perceptions. Findings reveal that students found ChatGPT most beneficial during the initial stages of topic selection and organizing their writing direction. However, the tool was perceived as less effective during the drafting phase due to the awkward phrasing and errors in its generated content. Furthermore, students highlighted challenges related to ChatGPT's reliance on unverifiable sources, raising concerns about the authenticity and reliability of its outputs. Other limitations included hallucinations, abstract or generic language, and biases stemming from its training data. Many learners also struggled to craft effective prompts, underscoring the need for targeted instruction on prompt design and strategic questioning. This study emphasizes the importance of fostering critical thinking skills to evaluate AI-generated content and ensuring that students learn to ethically and effectively utilize generative AI. Additionally, the findings highlight the need to develop students' proficiency in prompt engineering, enabling more productive interactions with AI tools. By exploring both the possibilities and limitations of generative AI in academic writing, this study provides valuable insights into the future direction of writing education in the AI era. This research contributes to the growing body of literature on generative AI in education and underscores the need for an adaptive approach to writing pedagogy that prepares students to collaborate effectively with emerging technologies.

Statistics

  • 72 Viewed
  • 0 Downloaded

Tools

Issue List

Citation status

* References for papers published after 2023 are currently being built.

Basic Search Advanced Search