Artificial intelligence (AI), and specifically “Generative AI” tools such as ChatGPT, have quickly become everyday words.
If you’re teaching a Nursing course with any significant writing component, whether it be a writing assignment, or even a Canvas discussion board, you should be prepared for the possibility that some number of students might be using these tools in inappropriate ways. And, unfortunately, we still do not have any meaningful way to stop it from happening.
However, that doesn’t mean we’re helpless.
Here are some proactive steps you can take to make sure students know exactly what is and is not allowed and what you can do if you suspect some form of academic dishonesty including the use of generative AI.
First, make sure your syllabus is clear and explicit regarding what you do and do not allow:
There are two resources available to help you with this.
- The Provost’s Office has created this syllabus template which includes text specific to the use of AI which you can modify to suit your particular needs.
- The Center for Teaching and Learning has created this page which includes 3 different tiers of language you can include depending on if, or to what extent, you wish to allow the use of AI tools.
Second, familiarize yourself with ChatGPT so you know what its results look like in advance:
ChatGPT is free to use, and you can (and probably should) set up your own account: https://chat.openai.com/
And one way you can use it to help you spot possible plagiarism is by entering your writing prompt(s) or exam question(s) to see what ChatGPT says.
For example, if you have an assignment that requires students to respond to a 3-part question, enter each part individually into ChatGPT to see how it responds. Then, enter each of the 3 parts sequentially to see if the response is different.
Familiarizing yourself with ChatGPT and other tools is one of the best ways to be able to easily spot red flags. And, if I can help with anything, don’t hesitate to contact me.
Third, make sure you are familiar with common signs that writing content has been generated using AI:
Common signs of AI-generated content include:
- Incorrect and outdated information.
- Generative AI tools *appear* very smart, but frankly they’re kinda dumb. And often, this will show up in AI-generated responses that contain incorrect or erroneous information. Many AI tools have even been shown to “hallucinate”.
- Lack of depth and personality.
- Unlike machines, humans are unique individuals, with specific ways for thinking and of expressing our thoughts. So look for the opposite of that: look for clues that seem shallow or generic in the writing that may indicate it’s not genuine.
- Repetitive language and overuse of jargon.
- There’s a great deal of terminology and jargon specific to Nursing. And AI tools may tend to overuse terms or even misuse them in ways that will help you identify potential plagiarism.
- Significant differences from previous written work.
- Many of our writing courses tend to include a LOT of writing, so you may have access to previous written work you can compare with that will help you judge whether or not you think AI may be involved.
- An obvious example would be a sophomore who suddenly writes better than most graduate students.
Finally, for higher stakes exams and assignments, consider adding additional layers of security.
- If you want to completely prevent students from using specific websites, one way to accomplish this is to use Honorlock. If you’d like to discuss this option for your course, please contact me.
- There are additional options available, including human proctors, among other things. And, if you’d like to explore these options, please contact me.