Recent technological advances in artificial intelligence (AI) have enabled high school students to provide coherent and informative answers to everyday questions or educational activities at the touch of a button (Kasneci et al., 2023; Yan et al., 2023).
The easy access to generative AI raises ethical and practical questions for educators. Turnitin reports that one in ten college student submissions contains AI-generated text (Turnitin, 2024). While it is widely accepted in the education community that secondary school students also interact with large language model (LLM) chatbots, the understanding of their motivations for doing so remains limited.
Balmoral State High School – a large school in central Brisbane with around 1,000 students – recognised a growing need to explore the impact of student use of AI and began to question the impact on student success and academic integrity. This year, a group of cross-curricular Balmoral teachers are collaborating with the University of Queensland (UQ) Learning Lab by joining the Partner Schools Program to investigate why students use AI tools and how these insights can inform future teaching strategies and assessment approaches.
With support from the UQ Learning Lab, the team recently completed a literature review and student survey to lay the foundation for the study and better understand students’ current relationship with AI.
What motivates students to plagiarize?
Before conducting a series of focus groups with our students, we first wanted to find out what motivates them to plagiarize.
When reviewing the research on motivation, we found that there is an obvious gap in the existing literature regarding students’ motivation to use LLMs. For this reason, we focused primarily on students’ motivation to cheat or plagiarize.
Research has found that factors such as easy access to information on the Internet, pressure to perform, and misinformation have changed students’ attitudes toward plagiarism (Ma et al., 2008; Evering & Moorman, 2012). A lack of integrity, maturity, confidence, or experience with certain writing genres and a lack of understanding of how to complete certain tasks are considered other motivating factors. It has been found that students who do not commit plagiarism do so out of fear of being caught rather than for ethical reasons (Evering & Moorman, 2012).
These research findings provided a background for understanding how students might approach the use of LLMs in their studies and therefore serve to guide our practice as educators in designing assessments and explicitly teaching the ethical use of AI to students in the future.
Considerations for teachers
Overall, a promising perspective for the integration of LLMs into education is emerging in the academic community, with potential benefits in content creation, personalized learning experiences, and accessibility improvements for students with disabilities (Kasneci et al., 2023). However, concerns remain regarding interpretability, ethical implications, and the potential for impairment of critical thinking skills when LLMs are overused in complex tasks (Kasneci et al., 2023; Grose, 2023).
Currently, students use AI in education to complete time-consuming and repetitive tasks such as completing exam papers (Yan et al., 2023). The majority of research suggests that AI and LLMs are promising tools that can enhance the educational experience by automating tasks and supporting learning. However, their integration must be carefully managed to achieve the greatest benefit and avoid the obvious pitfalls.
We believe that AI in secondary education will improve engagement and personalization of learning experiences. In particular, it will greatly benefit students who need intensive language learning support and promote the development of problem-solving skills in all students. However, researchers warn against over-reliance on LLMs as it could inhibit critical thinking and independent learning skills. Most studies argue for a redesign of traditional assessment methods to encourage creativity and reduce the risk of LLMs undermining academic integrity (Crawford et al., 2023).
How are Balmoral students currently using AI?
In the second trimester of this year, our students were asked to complete an anonymous survey about their use of AI platforms for assessments and general schoolwork. A total of 399 students (163 girls, 203 boys, 15 nonbinary, and 18 who did not disclose their gender) from grades 7 to 12 participated. This survey gave us valuable insights into how and why students use AI in their classrooms. This will help us design our future focus groups to better understand student motivation.
The survey data shows that 53% of students said they had not used AI during their school years. Of this percentage, 41% felt that they did not need to use AI, while only 11% did not use it for fear of being caught and 13% admitted that they did not know how to use it.
The majority of students currently engaging with AI (80%) use AI for English assignments, followed by science (50%) and humanities (42%). Their main motivation was to obtain examples of “high-quality” work or to provide a starting point for writing, correct grammatical and spelling errors, and improve the overall quality of their own work before submission.
Student feedback on motivation and ethical behavior
When students were asked to comment on their motivations, some of their reflections were as follows:
When I need feedback on a specific part of my work, it helps in finding errors or expanding on a particular idea and serves as a guide.
I feel like I don’t understand how to do the task and I’m not sure I can do the task.
I use it to give myself a structure for English so that my stories are planned out and I can start writing faster with a more solid plan.
Nowadays, answering questions by Googling is too inefficient. AI allows you to ask a question and get a relevant answer immediately. It is a great tool for research.
It helps me a lot when I miss school material and when I need feedback because the teacher is not there or there is a substitute teacher.
The survey also examined the ethical aspects of using AI in teaching, asking students whether they considered it cheating. On average, only a third (34%) of students perceived the use of AI as cheating, and this percentage decreased as grade level increased: 42% in 7th grade versus 29% in 12th grade. Male students across all grade levels were slightly more likely to view the use of AI as cheating.
When students were asked to express their stance on whether the use of AI is considered cheating, their statements highlighted the complexity of this issue. Many respondents acknowledged that AI should be used as a co-author rather than a ghostwriter on their work:
…it’s easy and accessible to everyone, and while it’s not good at doing the work for you, it’s helpful if you need feedback on a particular aspect of what you’re writing, etc.
This doesn’t give anyone a special advantage or a better understanding of the subject, but it does require people to use some part of their brain to access and use the information that the AI provides them.
It depends on how AI is used in school. If it is used to do one’s work without the student understanding anything, I consider that cheating. If a student uses it to teach themselves something or explain something in an exam, that is not really cheating.
Writing an entire essay using AI would be cheating, but using it for small tasks and research is fine.
Although our survey results and responses still need much more analysis, our action research already paints a picture that our students view AI as a valuable tool to improve their work. However, it also underscores the importance of teachers guiding their students in the ethical use of such a valuable tool.
Where to next?
From here, we plan to conduct further analysis of our school-wide survey data to deepen our understanding of students’ use of AI in learning.
The aim of this initial survey was to provide insights into our students’ perceptions of AI and to guide the development of authentic focus groups as the next phase of our research project. We anticipate that these focus groups will enable us to gain a more comprehensive understanding of students’ perceptions of AI to ultimately answer our research questions: “What are students’ motivations for choosing to study an LLM in their studies?” and “What implications do these findings have for the design of learning and assessment?”.
References
Crawford, J., Cowling, M., & Allen, K.A. (2023). Leadership is needed for ethical ChatGPT: character, judgment, and learning using artificial intelligence (AI). Journal for university teaching and learning practice, 20(3). https://doi.org/10.53761/1.20.3.02
Evering, LC, & Moorman, G. (2012). New considerations on plagiarism in the digital age. Magazine for youth and adult literature, 56(1), 35-44. https://www.jstor.org/stable/23367758
Grose, TK (2023). Disruptive influence. ASEE Prism, 32(3), 14-17. https://www.jstor.org/stable/10.2307/48734149
Kasneci, E., Seßler, K., Küchenmann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Günnemann, S., Hülsermeier, E., Krusche, S., Kutyniok, G., Michaeli, T., Nerdel, C., Pfeffer, J., Poquet, O., Sailer, M., Schmidt, A., Seidel, T. & Kasneci, G. ( 2023). ChatGPT forever? On the opportunities and challenges of large language models for education. Learning and individual differences, 103. 102274. https://doi.org/10.1016/j.lindif.2023.102274
Ma, HJ, Wan, G. & Lu, EY (2008). Digital cheating and plagiarism in schools. Theory and practice, 47(3), 197-203. http://www.jstor.org/stable/40071543
Turnitin. (April 9, 2024). Turnitin celebrates the first anniversary of its AI writing detector with millions of essays reviewed worldwide. (Press release). https://www.turnitin.com/press/press-detail_17793
Yan, L., Sha, L., Zhao, L., Li, Y., Martinez‐Maldonado, R., Chen, G., Li, X., Jin, Y., & Gašević, D. (2023). Practical and ethical challenges of large language models in education: A systematic scoping review. British Journal of Educational Technology, 55(1), 90-112. https://doi.org/10.1111/bjet.13370
