By Dr. Vlad Krotov
Generative AI has hit business educators like a freight train. In just a few months after its launch in 2022, ChatGPT had acquired 100 million users; 200 million users are predicted by 2023. Following the suit, Google has released its own conversational chatbot, Google Bard this year. Google Bard is powered by the same technology as Google’s search engine, so, unlike ChatGPT, it seems to be more aware of recent news and developments.
Shortly after the release of ChaGPT by OpenAI, several professors from top business schools announced that ChatGPT was able to pass their exams. While some business school professors still act as if ChatGPT doesn’t exist, a growing number of educators believe that Generative AI is a disruptive technology that will quickly and permanently alter the century-old rules and pedagogical approaches in business education. If this is true, then educational institutions must introduce changes and create policies to ensure that students use this new, disruptive technology in a way that does not impede their learning.
Nowadays, most students know what ChatGPT is and how to use it for completing homework assignments. How to mitigate the academic integrity issues associated with the use of GAI by students seems to be of the utmost importance to business schools, since academic integrity is important for quality of business education and is a formal requirement of all major international accreditation bodies, such as AACSB. In this article, I outline three simple strategies that business educators can use to mitigate academic integrity issues caused by GAI use. I also discuss each of these strategies’ pros and cons.
Punitive Strategy
Despite the rapid advancements in GAI, Many educators choose to teach their courses “as is”. Some specify in their syllabi that the use of Generative AI for completing assignments is prohibited and punishable under the school’s Academic Integrity Policy. AI-detection tools, such as ZeroGPT, are used to monitor student submissions for AI-generated content.
Pros
-
- Trusting students to make ethical choices and punishing those who do not is an old, simple, and, perhaps, wise approach for ensuring academic integrity. When students want to cheat, they will find a way to do so – by using GAI, hiring someone to do their projects, or in some other way. It is important for educational institutions to have an admissions process that screens out students who are likely to cheat in the first place. Instructors should be able to trust most students and not act as investigators and prosecutors at all times. If a cheating student is caught, the punishment should be severe enough to deter others from even considering unethical behavior.
- A minimal amount of effort on the part of faculty and the business school is required with this approach
- A business school may want to take this approach in the short term if they want to “wait and see” what happens with Generative AI in business education before making any important decisions or investments.
Cons
-
- AI detection tools are often ineffective at detecting AI-generated text, even though this issue is not as serious as some educators believe. False-positives are also quite common. It is possible for students to revise AI-generated text to make it unlikely that it will be flagged by such software as ZeroGPT. Furthermore, proofreading tools such as Grammarly and WordTune can produce false positives as well.
- GAI is here to stay, most likely. It may not be wise to prohibit students from using GAI tools, since these tools may soon become essential in the business world.
- In the near future, business schools will probably discover new ways to improve student learning by implementing GAI. Those business schools that do not adopt GAI for teaching and learning may soon lag behind those that do.
Flipped Classroom Strategy
It is possible to “flip” a class so that most learning and important assessments take place in a physical classroom, in front of the instructor, and with very little use of computers. For example, all exams can be administered face-to-face. Important learning exercises can be done in class as well. Major projects, while carried out outside of the class, should be presented and defended in class as well.
With this approach, the instructor can offer students help and make sure they are the ones doing the assignments. An instructor can ask individual students or student groups to demonstrate and explain progress on their project work in class every week. The grading weight devoted to attendance and participation can be increased as well, encouraging students to attend face-to-face classes. Students are free to use GAI tools outside of the classroom in any way they see fit (e.g. to prepare for a particular class session), but they should be able to demonstrate their competence face-to-face.
Pros
-
- Performing teaching and assessment face-to-face can be quite effective for attaining course learning objectives.
- It’s much easier for an instructor to detect cheating when most of the work is performed in front of him or her.
Cons
-
- While it’s possible to ask online students to take major exams or defend major term projects on-campus, this strategy is obviously not well-suited for asynchronous online courses.
- May not be appropriate for large classes, since this approach requires individual attention to every team and, sometimes, every student.
Integration Strategy
Instead of excluding GAI from the classroom, an instructor may choose to embrace the technology in a way that actually assists teaching and learning. This is probably one of the most effective yet difficult approaches. GAI is still a new technology. Many educators lack a solid understanding of how to use this technology ethically and productively.
What’s clear though, is that this approach may require a radical redesign of each course’s pedagogy. GAI can be used by students to answer basic questions at the “understanding” level of Bloom’s Taxonomy, for example. In fact, students may be instructed to ask ChatGPT or Google Bard questions in relation to the subject matter of the course and then to read and evaluate the responses as a part of an assignment. Thus, the point of the assignment is to interact with GAI and not to provide “correct answers”.
When it comes to higher-order cognitive skills, an instructor may ask very complex and context-specific questions in a format not supported by major GAI tools in order to decrease the likelihood that GAI will be used to complete these assignments.
Ideally, these assignments should be in the form where GAI still lacks capability. For example, instead of just asking to analyze a case study, the instructor can ask students to create flowcharts or UML Activity Diagrams based on the case. Alternatively, students can be asked to record videos with their analyses and post them to YouTube. Thus, even if students ask GAI for assistance, the final product will largely be their own work.
Also, while GAI tools such as ChatGPT and Google Bard are becoming increasingly knowledgeable in many topics and increasingly capable of performing very complex cognitive tasks, they often lack knowledge and understanding of very narrow and specific contexts. For example, an instructor can ask complex questions in relation to specific local individuals or organizations that may be known only to students and the instructor. For example, let’s say there’s a small, local software company where the university is located. The instructor can talk about this company in class and then asks students to come up with strategies that are suitable for this company given its unique local context
Regardless of which approach is chosen for assessing higher-order cognitive skills, the instructor can quickly run his or her questions and assignments through ChatGPT or Google Bard to see what kind of responses students are likely to get when they ask GAI for help. If GAI can answer these questions with ease, then additional context or complexity needs to be added to the question. ALternatively, the format of the assignment can be changed (e.g. from an open-ended response to a visual diagram).
Students can rephrase these specific questions as generic ones and submit them to ChatGPT. But generating a quality response will still require analyzing and accommodating the local context (and this is where most of the learning will occur) or putting the responses in a format that still entails some learning. The instructor can deduct points if the local context is not properly accounted for or the format instructions are not followed. A grading rubric can be created that evaluates students’ work based on the extent to which a solution is contextualized for a specific company or individual and the extent to which all the directions were followed. If necessary, an oral defense can be scheduled so that students can demonstrate their mastery of the material.
Pros
-
- Since GAI is likely to become a permanent fixture in business, this approach will make students more prepared for the era of GAI
- With this approach, the instructor can automate some of the basic tutoring tasks and focus on developing higher order cognitive skills among students via complex, contextualized, and innovative assessments.
Cons
-
- This approach can be time consuming, since it requires quite a bit of thinking and curriculum revision
- Many educators may lack time, expertise, or motivation for integrating GIA into their curriculum in a way that is ethical and conducive to student learning.
In conclusion, it can be said that Generative AI is very likely to become a permanent fixture in business education. Individual educators and business schools will have no choice but to adapt to this new, disruptive technology and find ways to accommodate in an ethical and productive fashion. The list of strategies for accommodating GAI in the classroom provided here is not perfect or exhaustive. What’s important though is that every educator and business school should have a strategy in relation to GAI, or they will quickly find themselves in a disadvantaged situation. Having a strategy in relation to GAI is better than having no strategy at all.