Vol. 19 / 2024 – International Conference on Virtual Learning
Analyzing the uses and perceptions of computer science students towards generative AI tools
Alisha HASAN, Bogdan SIMION, Florin POP
With the evolution of generative AI (GenAI) tools based on Large Language Models (LLMs), stakeholders in Computer Science education have sought to leverage opportunities and mitigate risks that these tools may entail in the educational process. Thus, it is crucial to understand students' uses and perceptions of GenAI tools in their learning. Our work captures how computer science students use GenAI tools that produce code or text. Additionally, we investigate student perceptions in terms of important factors such as perceived usefulness, correctness, reliability, accessibility, trust in responses, motivation to learn, confidence building, potential dangers, and perceived implications on the software development industry. After surveying students from three institutions, we conducted a mixed-methods analysis and noted that student uses of GenAI tools are typically educational, with few students using them for purposes detrimental to their learning such as generating solutions. Additionally, students are more aware of the potential for misleading incorrect responses from GenAI tools and the pitfalls of over-reliance on such tools than educators might expect. Overall, our findings are useful for computer science educators to be mindful of students’ uses and perspectives on GenAI, to better guide students towards positive learning outcomes and foster good learning habits.
Keywords:
LLMs,
Student Perceptions,
Generative AI,
CS Education
CITE THIS PAPER AS:
Alisha HASAN,
Bogdan SIMION,
Florin POP,
"Analyzing the uses and perceptions of computer science students towards generative AI tools",
International Conference on Virtual Learning,
ISSN 2971-9291, ISSN-L 1844-8933,
vol. 19,
pp. 363-372,
2024.
https://doi.org/10.58503/icvl-v19y202430