The impact of ChatGPT on students' academic writing

Abstract

A mere two months following the release of the Chat Generative Pre-Trained Transformer (ChatGPT) in late November 2022, the chatbot reached 100 million monthly active users by January 2023 (CHOW, 2023 ) Given the novelty of this topic and its burgeoning usage, research in this field appears imperative. This study aims to examine ChatGPT's impact on education more thoroughly. The impact of ChatGPT on students' academic writing is a novel and intriguing thesis topic that focuses on four unique research questions. First, how do students use ChatGPT in their academic writing? How frequently do they use it, and for what purposes? This question holds practical implications for understanding the role of AI in education. Additionally, it investigates students' perceived differences in writing improvement when using ChatGPT compared to writing without it. Moreover, it explores the psychological experiences of students utilizing ChatGPT, including their confidence and anxiety levels when incorporating the tool into their academic writing. Finally, the fourth research question that this thesis will address is how students perceive the limitations of ChatGPT in academic writing. Ethical concerns represent another crucial aspect to be addressed in this paper. While the project's primary focus is to investigate ChatGPT's impact on students' academic writing, it is essential to acknowledge the comprehensive nature of this study. However, it is also necessary to recognize the limitations imposed by factors such as sample size, time restrictions, and the need for the study's findings to encompass various student populations fully. Consequently, the perceptions and experiences contributing to this study's results may differ from those of other studies. ii The study employs a mixed-methods approach, gathering data through surveys administered to students across different educational levels. Additionally, the methodology involves analyzing literature reviews from previously published authentic articles. The study's results reveal several notable findings which need to be investigated furthermore. Keywords: Natural Language Processing (NLP), ChatGPT, Academic writing, Machine learning in education, Writing assistance tools, Language model, Artificial intelligence, GPT-3.