Dies ist eine Übersichtsseite mit Metadaten zu dieser wissenschaftlichen Arbeit. Der vollständige Artikel ist beim Verlag verfügbar.
Student Use, Performance and Perceptions of ChatGPT on College Writing Assignments
9
Zitationen
2
Autoren
2024
Jahr
Abstract
ChatGPT is a generative chatbot that conversationally responds to prompts using content mined from the internet. In a very short time, ChatGPT has disrupted many industries, including education. However, the risks and dangers of this technology remain to be seen; for example, students can use ChatGPT dishonestly on writing assignments. Education currently lacks a clear understanding of students’ perspectives and intentions around this technology. In this study, undergraduate students from a comprehensive university were briefed on ChatGPT, given guidelines for its use, encouraged to use this tool on their written assignments and surveyed for their opinions on ChatGPT. We found that a small number of students chose to use ChatGPT and there was no grade difference between students using and not using this tool. Additionally, thematic analysis of student survey responses revealed that while some students viewed ChatGPT as an idea generator and improvement tool for written assignments, others saw it as a way to create submissions for class assignments, potentially infringing on academic integrity. The majority of students felt that ChatGPT should be allowed in college classes and intend to use ChatGPT on future assignments. Finally, most students share misconceptions about ChatGPT, including that it is a source of information and a search engine. Given the rapid rates of adoption of this technology, and its enormous potential for integration across sectors, it is incumbent on educators to deliberately guide and train our students with this tool, while keeping our students’ perspectives and intentions in mind.
Ähnliche Arbeiten
Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI
2019 · 8.534 Zit.
Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead
2019 · 8.423 Zit.
High-performance medicine: the convergence of human and artificial intelligence
2018 · 7.917 Zit.
Proceedings of the 19th International Joint Conference on Artificial Intelligence
2005 · 5.781 Zit.
Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI)
2018 · 5.582 Zit.