M. Peters1, D. Angelow2, C. Déri1, A. King3
Since the arrival of generative artificial intelligence (GenAI) tools in education, university students have found creative ways of using these to produce their assignments, among other things, to quote and paraphrase authors. The use of GenAI tools can be useful, and efficient but can also result in less learning and academic integrity misconduct. This research paper was interested in examining why students choose to cognitively unload to GenAI central tasks such as idea reformulation and quoting. Is the choice motivated by a desire to write a better paper or is it to avoid dealing with complex tasks? This multifaceted issue cannot be reduced in terms of laziness or plagiarism but would benefit from investigating various factors such as the training students have in quoting and paraphrasing, the pressures they are under, and the lack of clear guidance on the accessibility of using GenAI tools and the possible dependence on them.
A qualitative research methodology was used to examine the discourse of 21 group interviews (between 1 and 6 students in each group) about their use of paraphrases and quotes in their assignments. These students (n= 93 students) were in four difference Canadian universities, at the undergraduate level, studying in various disciplines, in French or in English programs. Interviews were on average 42 minutes long and were done in the universities, in person. The interview guide contained questions on different aspects of writing an assignment in university. Specifically, 6 questions were asked regarding quotes and paraphrasing and the use of GenAI tools, for example: How do you choose between quoting and paraphrasing when writing your assignments?
All the interviews were recorded, transcribed and analysed using NVivo software. To check the analysis’ consistency, three out of the 21 interviews (15%) were double-coded, with a strong agreement yielding an average Cohen’s Kappa of 0,83.
Results show a wide diversity of reasons why students choose to quote or paraphrase, some using GenAI and others preferring to use their own skills to do it. These reasons were grouped into four types. First came the idea of being more efficient and gaining time. The other three reasons can be grouped under the heading: Fears! For some students, they were uncertain about their knowledge of the norms for quoting and paraphrasing. For others, they expressed their linguistic and writing insecurities and finally, the fear of not being sure of what was expected of them lead many students to choose to use GenAI tools for quoting and paraphrasing.
These results suggest that GenAI tools are not used solely out of laziness or lack of time. These tools are viewed as a potential solution for what is often not understood or has not been learned properly. Students’ use of GenAI tools to quote or paraphrase seem like a reaction to unclear academic expectations and maybe a cry for more training for students to develop their skills in quoting and paraphrasing with and without GenAI. Conclusions will highlight the urgency of clarifying pedagogical and ethical guidelines on how and when GenAI can be used in academic writing.
Keywords: Artificial intelligence, quoting, paraphrasing, university.