Students have wasted no time in figuring out how to use generative artificial intelligence tools to help them in their studies. Need an essay outline? Done. Stuck on a tricky concept? Just ask a chatbot. Run out of ideas? Ask a GenAI tool for help. These tools are changing the way students learn, think and, more importantly, complete their assignments.
This raises some big questions. Are students using it wisely? Where do we draw the line between “helpful assistant” and “academic shortcut”?
Last year, we talked to students across a range of disciplines, on what they thought of using GenAI tools in their studies.
- Spotlight guide: Bringing GenAI into the university classroom
- Keep calm and carry on: ChatGPT doesn’t change a thing for academic integrity
- So, you want to use ChatGPT in the classroom this semester?
Some of the ways they were using the tools, we discovered, were pretty useful. But some students felt that they were getting too reliant on technology. We asked them about how they crafted prompts, navigated assignments, tackled research and, yes, occasionally misused these tools. Here, we’ll outline our findings – and identify how educators can make the most of GenAI.
Use of GenAI in learning
Students have been using prompts and GenAI tools in various ways to enhance their learning experience. They use it to formulate prompt questions with contextual knowledge, iterating multiple times to refine their queries. For many, GenAI tools are invaluable for summarising long studies or reports and saving time. Yet we heard from a few students who felt their learning was impeded if they used GenAI tools, and consciously tried to steer clear of using them.
For many students, GenAI is deployed as a personal tutor on a perpetually opened browser to help answer the various specific questions that come up. It is especially so when they need an answer and cannot find time with the professor. GenAI tools have been instrumental in helping them understand complex topics, removing barriers to their learning. A favourite prompt for students is to use these few words at the start of the prompt: “Explain it to me like I am a 12-year-old…” It helps to simplify and rephrase unclear instructions or tutorial questions.
To provide more intentional guidance to students’ learning, institutions can consider developing their own chatbot platforms, linked to reliable large language model providers via application programming interfaces. Faculty can then customise the chatbots with intentional guidance, and input documents to the chatbot’s knowledge base for greater relevance and specificity. Learner inputs and chatbot responses can be logged, then analysed to find out how our students are learning through the questions they ask. It is also an opportunity for faculty to review the learning materials and questions used in their lessons.
Another activity we heard about from students is the use of GenAI tools in class as part of in-class discussions. Students can be asked to generate solutions to a specific problem or discover what the concept is about using the GenAI tools. Through the discussions in class and with critiquing the outputs provided by GenAI tools, students are able to learn, as well as be more discerning of the contents generated.
Guidance in assignments
GenAI tools have been invaluable in helping students to write reports, fix programming problems, correct grammar and generate an organised structure in assignments. This is how our students use GenAI tools to enhance their work. Students have mentioned how they use GenAI in generating the flow, outline and ideas for assignments, especially for those where the topic is difficult to grasp. In statistics or IT-related assignments, GenAI tools can help them to generate multiple functions or programming code, which will lead them to finding a working solution.
Our students do also acknowledge that AI-generated ideas and content are often generic, that using its content can lead to similar answers among their peers. They find that they need to have even more original ideas to get an A grade, which highlights their depth of knowledge and creativity. They said that if they simply “cut and paste” with some tweaking, they get a minimal passing grade or even a failure. They also realised that they must find, credit and cite original sources.
The first step to review our assessment is to imagine answering the question like students. Students will use various GenAI tools and iterate the process until they get a reasonable answer. Only then will we be able to find out the vulnerability of our assessment and how best to amend it.
Next, we will need to reduce or completely remove the amount of marks allocated to grammar, structure and organisation. The assignment itself must require higher-order thinking or be context specific, and students are assessed on their original and critical thought. Projects are based on real-life situations, centred around industry partners and the problems they’re facing today. These partners can also provide feedback on the students’ proposed solutions later on.
It is also advisable to remind students of repercussions for misuse of these GenAI tools. Most students would like to do well and would not want to jeopardise their degree with academic dishonesty or low-quality work.
While most of our students find GenAI tools invaluable as a useful learning assistant, they are also keenly aware of its limitations. They seem to use it wisely enough but we need to discern and refine responses obtained from GenAI ourselves. The best way forward is to acknowledge their usage and provide guidance for strategically integrating GenAI into teaching and learning. With that, we can build on our students’ intrinsic motivation towards mastery and maximise their potential capabilities for leveraging GenAI.
Miriam Wun is a senior educational developer and Nah Yong En is an educational technologist, both at Singapore Institute of Technology.
If you would like advice and insight from academics and university staff delivered direct to your inbox each week, sign up for the Campus newsletter.
comment