What is it about?
This research delves into the implications of generative AI, particularly ChatGPT, in the educational sector. ChatGPT, developed by OpenAI, can generate human-like text and has various applications such as language translation, creating content, and coding. The study focuses on its impacts in educational assessments, highlighting both the advantages (like providing instant feedback) and concerns (such as academic integrity and plagiarism). To address these challenges, the research suggests rethinking online assessment practices. A survey was conducted to understand both educators' and students' viewpoints on assessment adaptations in light of AI tools.
Featured Image
Photo by Levart_Photographer on Unsplash
Why is it important?
The increasing use of generative AI in education can revolutionize learning methods, but it also brings forth ethical dilemmas and practical challenges, especially in assessments. Ensuring academic integrity becomes complex when AI tools, like ChatGPT, can produce assignments or answer questions that might be indistinguishable from a student's genuine effort. Moreover, a mismatch was found to exist between educators' and students' perspectives on adapted assessment prompts, with some students feeling the modifications might suppress creativity. This discrepancy emphasizes the need for balanced assessment designs that prioritize learning and creativity while deterring cheating. Preparing students for an AI-prevalent future involves not just teaching them about AI but equipping them to understand its broader societal implications and how to use it ethically.
Perspectives
This is an interesting comparison between students and educators but also across two different countries, Australia and United States. This area is rapidly changing so we have an opportunity here to understand the current thinking.
Elaine Huber
University of Sydney
Read the Original
This page is a summary of: Educator and Student Perspectives on the Impact of Generative AI on Assessments in Higher Education, July 2023, ACM (Association for Computing Machinery),
DOI: 10.1145/3573051.3596191.
You can read the full text:
Contributors
The following have contributed to this page
