4. Discussion

A recent publication (Author, 2024) presented an analysis of the pre-survey responses. Results show inequality in the level of familiarity with and prior use of GenAI by students. Furthermore, students expressed a desire for autonomy in being able to choose whether to use GenAI or not, with the vast majority agreeing that students should retain an option to opt out of its use. Finally, most were uncertain whether they were going to use GenAI or not in the course, perhaps due to a lack of certainty over what types of uses were permitted. In the end, 58.8% of the groups declared using GenAI in the group written article. A recent study of 6,300 university students in Germany found that 63.4% of students reported using AI-based tools for their studies (von Garrel & Mayer, 2023) which is consistent with our findings when considering that the rate in our study was not for individual use but for use within a specific group assignment. While two-fifths of the groups declared not using GenAI, overall, students reported a significant increase in their familiarity with GenAI tools, their use of GenAI tools to assist with their schoolwork, and in their plans to use GenAI to support their future learning at the university. This aligns with findings from previous work, where students acknowledge ChatGPT, as a valuable complementary learning resource and describe it as helpful for learning (Sánchez-Reina et al., 2024; Shoufan, 2023). Also, by the end of the course, no significant changes were found in student beliefs related to students having the option to opt out of using GenAI in class, nor to beliefs that grading criteria should differ for students using AI tools versus those who do not.
In relation to the first research question, R1. Would first-year students be accepting of a course learning agreement on GenAI use?,pre-survey results show the vast majority of students supported the use of a course learning agreement for governing GenAI in higher education (M = 4.45, SD = .77). Most utterances (86.1%) related to either preventing students from being unfairly disadvantaged or helping students learn to use AI properly as reported by Beardsley et al. (2024) which provide insights into student expectations of the learning agreement. At the end of the course, student support had decreased significantly but still averaged 4 out of 5 on a Likert scale (M = 4.00, SD = .85), suggesting continued support for the learning agreement approach but, perhaps, a failure to meet student expectations. In relation to the second research question, R2. Would students adhere to the terms of the learning agreement in their group assignment? , all student groups did make a declaration involving technologies used, yet none felt the need to cite their use of GenAI tools, and only 1 of the 10 groups that did declare use of GenAI acknowledged its limitations as agreed upon in the learning agreement. Hence, it appears students did not adhere to all items in the learning agreement – especially items 3 and 4. Findings related to the third research question, R3. After experiencing a course learning agreement, how would students improve on its use?, provide insight into possible reasons for low adherence for these items. The lack of specific examples of what is not permitted, class time for further discussion of issues related to the learning agreement, and opportunities to re-engage with the learning agreement may have contributed to the lower post rating – as these were suggestions for improvement. Additionally, several utterances related to student motivation to adhere to the learning agreement with some (8.33%) suggesting a stricter enforcement of the learning agreement terms being needed while others (11.67%) suggesting the approach should be more participative and supportive of student autonomy. The latter aligning more closely with self-determination theory and nicely encapsulated by one student’s response,
The focus of the learning agreement should be on the student’s goals and needs, rather than on the instructor’s expectations. Also, making it more accountable, where the student should be responsible for meeting the goals that they have set for themselves, and the instructor should be responsible for providing the support that the student needs to achieve those goals.