Dies ist eine Übersichtsseite mit Metadaten zu dieser wissenschaftlichen Arbeit. Der vollständige Artikel ist beim Verlag verfügbar.
Challenges faced by students using ChatGPT: A qualitative study at private universities in Malaysia
0
Zitationen
1
Autoren
2025
Jahr
Abstract
The rising presence of artificial intelligence (AI) tools such as ChatGPT is shifting the way students are involved with academic content. While many students value its capability to quickly assist with assignments, create content, and support study efforts, the practice of using ChatGPT also presents various challenges in educational settings. The purpose of this study is to explore the problems students face when depending on ChatGPT for academic work. Using data collected from 441 students at two private universities in Malaysia, a thematic analysis method was employed to identify recurring themes. The findings reveal seven central themes: concerns about accuracy and reliability, technical access issues, difficulties in crafting effective prompts, limitations in subject-specific understanding, language and communication barriers, questions of academic integrity, and emotional or cognitive effects. The results also indicate that although students recognize the tool’s usefulness, there are significant issues such as distortion of information, overdependence, ethical risks, and inadequate access. The implications of this research emphasize the importance of equipping students with improved digital literacy skills, establishing clear strategies for ethical AI use, and ensuring AI tools are adaptable to diverse academic and linguistic needs. These insights aim to promote a more responsible and balanced integration of AI in higher education, supporting students in leveraging these technologies effectively while mitigating associated risks.
Ähnliche Arbeiten
Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI
2019 · 8.663 Zit.
Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead
2019 · 8.576 Zit.
High-performance medicine: the convergence of human and artificial intelligence
2018 · 8.091 Zit.
BioBERT: a pre-trained biomedical language representation model for biomedical text mining
2019 · 6.859 Zit.
Proceedings of the 19th International Joint Conference on Artificial Intelligence
2005 · 5.781 Zit.