Abstract
In recent years, and particularly since the launch of ChatGPT in December 2022, the potential impact of Artificial Intelligence Large Language Models (LLMs) such as ChatGPT or Google Bard on students’ academic output has both energised and terrified academic staff worldwide. Questions continue to rage over whether such assistive technology should be banned or embraced, and in the UK at least, the former appears to be the more common reaction, with a third of the elite Russel Group universities prohibiting use of LLMs, with many more considering doing so (Wood, 2023).
Despite staff concerns over the threat of LLMs to academic integrity, it seems clear that students are keen to employ the technology, with 53% of students surveyed in the USA admitting to using ChatGPT to help with homework assignments (Westfall, 2023). In the view of this researcher, without a reliable method of detecting AI-supported output, its widespread use is inevitable, and universities should take a proactive approach in harnessing the affordances of LLMs.
This presentation will report on a study implemented at a UK university in the Spring semester of 2023, in a module which helps students for whom English is not a first language develop their proficiency in English for Academic Purposes. Over the 12 weeks of the module, as he does every semester, the researcher worked with the students to develop research topics, introduced elements of academic English writing such as citating and referencing, topic sentences, paragraph structure, hedging, and signposting language.
This trimester, however, student production of their essay was supported through the use of ChatGPT. As will be described in the presentation, students were instructed in the use of prompts, and upon receiving the output from ChatGPT, whole class discussions and evaluations were held to discuss the merits of the pieces of text, and how they could be further refined by the students themselves, given their own knowledge of the research topics.
Overall, a majority of students reported that use of ChatGPT had improved the quality of the essay they had produced and also felt that the experience left them better equipped to produce future academic output without the use of LLMs. Nevertheless, they also sympathised with academic staff concerns of the threat to academic integrity posed by LLMs.
The presentation will outline the process implemented during the semester, detail feedback from the students on their experience of using ChatGPT and their perceptions on its impact on their academic writing skills. The presentation will conclude by posing and answering some questions regarding the ethical position of implementing this structured use of chatGPT as an academic writing resource.
Despite staff concerns over the threat of LLMs to academic integrity, it seems clear that students are keen to employ the technology, with 53% of students surveyed in the USA admitting to using ChatGPT to help with homework assignments (Westfall, 2023). In the view of this researcher, without a reliable method of detecting AI-supported output, its widespread use is inevitable, and universities should take a proactive approach in harnessing the affordances of LLMs.
This presentation will report on a study implemented at a UK university in the Spring semester of 2023, in a module which helps students for whom English is not a first language develop their proficiency in English for Academic Purposes. Over the 12 weeks of the module, as he does every semester, the researcher worked with the students to develop research topics, introduced elements of academic English writing such as citating and referencing, topic sentences, paragraph structure, hedging, and signposting language.
This trimester, however, student production of their essay was supported through the use of ChatGPT. As will be described in the presentation, students were instructed in the use of prompts, and upon receiving the output from ChatGPT, whole class discussions and evaluations were held to discuss the merits of the pieces of text, and how they could be further refined by the students themselves, given their own knowledge of the research topics.
Overall, a majority of students reported that use of ChatGPT had improved the quality of the essay they had produced and also felt that the experience left them better equipped to produce future academic output without the use of LLMs. Nevertheless, they also sympathised with academic staff concerns of the threat to academic integrity posed by LLMs.
The presentation will outline the process implemented during the semester, detail feedback from the students on their experience of using ChatGPT and their perceptions on its impact on their academic writing skills. The presentation will conclude by posing and answering some questions regarding the ethical position of implementing this structured use of chatGPT as an academic writing resource.
Original language | English |
---|---|
Publication status | Published - 13 Jul 2023 |
Event | Technology-Enhanced Language Learning International Conference - Online, United States Duration: 13 Jul 2023 → 14 Jul 2023 https://www.utpb.edu/academics/colleges/education/telic2023 |
Conference
Conference | Technology-Enhanced Language Learning International Conference |
---|---|
Abbreviated title | TELIC2023 |
Country/Territory | United States |
Period | 13/07/23 → 14/07/23 |
Internet address |