An Insight on the Use of Generative AI Tools by Computer Science Students in Higher Education

Abstract

<p xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" class="first" dir="auto" id="d327913e105">Since the release of Generative Artificial Intelligence (GenAI) tools, particularly in the domain of natural language processing, such as ChatGPT, Gemini, Copilot, and Llama, there have been concerns about students using such tools in their courses of study. In this paper, we provide insights on how Computer Science (CS) students at both undergraduate and postgraduate levels have been using GenAI tools alongside their modules in university. The insights are based on a designed quantitative combined with qualitative research which aims to gather information on their GenAI tool usage in terms of frequency, purpose, tool type, and its potential impact on their academic performance. We draw the heatmaps between the year of study, purpose (e.g. brainstorming/idea generation, coding and troubleshooting help, report writing), and frequency of GenAI tools usage. The paper provides additional insights on the relationship between the year of study and frequency of GenAI tools usage, and discusses the effect our findings may have on students learning through qualitative research. Furthermore, we provide insights on students' ethical perception and academic integrity for using such tools while providing recommendations for university staff to shape their future programmes, policies and guidance on the use of GenAI tools. Our analysis reveals that CS students primarily use GenAI as an assistive tool rather than for academic misconduct.

Ähnliche Arbeiten