The Use of Artificial Intelligence (AI) to Generate Case Studies for the Classroom – Faculty Focus

Artificial Intelligence (AI) has become a tool that is used in the classroom. The integration of technology in education has historically been gradual (Holmes, Bialik, and Fadel 2019). Some educators lack the training to effectively use AI in the classroom, which may limit the ability to design AI-based coursework (Amado-Salvatierra et al. 2024). In addition, the lack of understanding of how AI usage can be applied pedagogically can potentially deter AI integration (Afzaal et al. 2024). The focus of this article is to share examples of AI input prompts to generate case studies as a learning tool to help students learn course topics and learning outcomes. 

Case studies provide opportunities for students to learn and/or reinforce what they have already learned. While using AI to create case studies, it is important to use correct prompts to obtain appropriate output. For example, a single prompt asking AI to give five cases on five different topics may not provide sufficient detail. However, prompting AI to generate a case study on one topic will potentially generate a more appropriate case. Using as much detail in the initial prompt helps provide better output. An AI prompt example might be: ‘Assume the role of a professor teaching an introductory accounting course. Generate a case study for students to use to learn the very basic format of a balance sheet. Make sure the case study is real-world relevant.’ The stated role, course, and topic can be tailored to fit an appropriate class. The case study generated should be reviewed to verify alignment with the specific topic(s) and learning outcome(s). 

Cases should be generated in a format that allows for measurable assessment of the students’ learning. Specifying this detail in an AI prompt while generating a case study will help ensure the AI output is measurable. An additional AI prompt example to include with the previous example might be: ‘The case study is to include an assignment portion that allows faculty to measure the student’s performance.’ Careful evaluation of the output from this prompt needs to be performed to ensure topic alignment. 

Once an appropriate case study has been developed, AI can provide a grading rubric for the AI-generated case study by prompting AI to generate a rubric. An AI prompt example for generating a grading rubric is: ‘Provide a grading rubric that aligns with this case study.’ A review of the grading rubric’s use to measure appropriate topics and learning is recommended. 

At any point in this process, AI can be used to change the output. For example, an appropriate AI prompt to modify something in a case study might be: ‘Update the above-generated case study to include 5 assets, 3 liabilities, and 2 owners’ equity accounts.’ At any point of revision, assessment of the previously generated output (for instance, the rubric) may need to be regenerated. It is recommended to note the AI prompts that generate acceptable output so that those prompts may be reused for future AI-generated case studies. 

Another output for the case study could be an answer sheet for faculty to use and to share with students afterward to enable self-evaluation of performance. An AI prompt example might be: ‘Provide the answer sheet for this case study. Make sure to include details of any calculations and definitions of key terms.’ 

To add additional depth to using AI in the classroom, faculty may want to create two case studies on the same topic: one to be performed by the student without the use of AI, and one with the use of AI. This two-case-study method could allow students to learn how to use AI appropriately. A list of appropriate AI input prompts for the students to use would assist in the students’ learning how to engineer appropriate AI prompts. This effort would help students because research indicates they have a diminished sense of preparedness when they have insufficient exposure to AI application (Hsiao and Han 2023). 

An example of an AI prompt to produce a two-case study with and without the use of AI is: ‘Assume the role of an Accounting Professor teaching an introductory accounting course. Generate one case study for students to use to learn the very basic format of a balance sheet without the use of AI. In addition, generate a second case study with the same format as the first case study for students to use to learn the very basic format of a balance sheet with the use of AI. Make sure the case study is real-world relevant. The case study is to include an assignment portion that allows faculty to measure the student’s performance.’  

By using this two-case-study method, faculty can measure changes in student performance, allowing students and faculty to see how AI use can assist with case study topic comprehension. Providing the answer sheets to students will enable them to compare their performance and critically analyze the AI output. 

A valuable measure to assess would be the amount of time the student spent on each case. A line item on the case study can be added by AI by including a prompt such as: ‘Provide an item at the end of the case to enable students to report the amount of time that they spent on each of the case studies.’ 

In addition to the non-AI case and the AI-usage case, students can also perform conceptual evaluations. Such evaluations can be qualitative, allowing the student to critically evaluate how AI assists in efficiency and accuracy on the topic. Another focus could be qualitative evaluations on how AI can be used in their future careers based on the specified topic in the case study. For example, a conceptual question might address how someone in their profession would benefit from AI to help them perform better in their future careers. If these cases are created to be used throughout a course term, students can gain a clearer picture of how AI might be applied to their classroom experience. 

In conclusion, faculty can use AI to create tools, such as case studies which are focused on specific topics, to expose students to real-world scenarios and reinforce student learning. It is very important, as faculty and students use AI, to acknowledge that current AI outputs may not always be accurate. Faculty and students should evaluate the accuracy of AI-generated output and adjust as necessary. Experimenting with various AI inputs will allow faculty to become more comfortable with using AI. Recognizing AI’s role in the classroom does not replace faculty, but it can help provide excellent learning opportunities for students and demonstrate how AI can be used effectively. 

Rhonda Gilreath is an associate professor of accounting at Tiffin University in Northwest Ohio. She enjoys exploring new opportunities to implement in the classroom to improve pedagogical approaches to prepare her students for career readiness.

References  

Afzaal, M., Shanshan, X., Yan, D., and Younas, M. 2024. “Mapping Artificial Intelligence Integration in Education: A Decade of Innovation and Impact (2013–2023)—A Bibliometric Analysis.” IEEE Access 12: 113275–113299.

Amado-Salvatierra, H. R., Morales-Chan, M., Hernandez-Rizzardini, R., and Rosales, M. 2024. “Exploring Educators’ Perceptions: Artificial Intelligence Integration in Higher Education.” In 2024 IEEE World Engineering Education Conference (EDUNINE), 1–5.

Holmes, W., Bialik, M., and Fadel, C. 2019. Artificial Intelligence in Education: Promises and Implications for Teaching and Learning. Boston: Center for Curriculum Redesign. 

Hsiao, D., and Han, L. 2023. “The Impact of Data Analytics and Artificial Intelligence on the Future Accounting Profession: Perspectives from Accounting Students.” Journal of Theoretical Accounting Research 19 (1): 70–100.