Some interesting facts and helpful approaches from a courseware expert.
GUEST COLUMN | by Deborah Rayow
Academic integrity is not a new conversation to schools and educators, but the explosion of ChatGPT has complicated the already fraught discussion. According to McCabe’s Academic Integrity Survey in 2010, a high percentage of students admitted they have cheated or plagiarized at least once or more than once. Factor in the rise of generative AI, and educators have yet another variable to consider for their classrooms.
‘Factor in the rise of generative AI, and educators have yet another variable to consider for their classrooms.’
ChatGPT and other tools like it represent a significant leap forward in the capabilities of generative AI. For obvious reasons, this has caused quite a stir of concern for educators. However, there are also benefits of this technology for both students and teachers. The question then becomes how to identify when students are using ChatGPT inappropriately and how to teach students how to use the technology appropriately in their schoolwork.
Identifying ChatGPT in Student Writing
At first glance, generative AI seems like a cheater’s paradise and every educator’s nightmare. Students can input their homework or test question and receive a well-formed answer in minutes – hence the valid concern that students are not learning or conducting honest work. However, because ChatGPT pulls information from the internet, the answers students receive might be factually incorrect or lack the nuance that original student writing could include. ChatGPT can also include personal details that clearly have nothing to do with the student submitting the work (e.g., full paragraphs about a first-year student’s experience in a large public high school from a student in a small, private middle school).
Additionally, along with the flood of new generative AI tools has come corresponding detection tools. These tools work in a variety of ways. Some look for signs of generative AI and report the probability that a text sample was created by AI. Others look for signs of human-written text and report the probability that a text sample was written by a real person. Still others rely on digital watermarks from specific tools (for example, it looks for signs a text sample was written by ChatGPT 4). What these tools have in common, though, is that they alert teachers that there is a high probability—but not a certainty—that a student used ChatGPT or a similar tool to craft a response to a prompt. This level of uncertainty can be uncomfortable for educators determining whether and how to confront a student they suspect of academic integrity.
Teachers can combine alerts from detection tools with their familiarity with their students’ work as they review answers or text that seem superficial or might not make sense. Knowing a student’s writing style becomes important to identifying when their writing is straying from their usual voice, so that teachers are well-equipped to catch these inconsistencies.
Conversations About ChatGPT
In addition to regular discussions with students about plagiarism, teachers should include conversations about when and how to use ChatGPT as relevant to class assignments. Some educators feel that avoiding any mention of generative AI in classroom conversations around plagiarism reduces the likelihood that students will use these tools. However, ChatGPT is on everyone’s mind, and students are certainly going to learn about these tools outside of the classroom if not within – so educators should look to get ahead of it.
Some students may not understand that using a tool like ChatGPT could be considered plagiarism at all. It is important for teachers to make this connection clear for students. The discussion around generative AI and plagiarism should also be aligned to and appropriate for the student’s age and grade level as a student in elementary school does not need the same explanation as a student in high school.
Constructive Use of ChatGPT in the Classroom
A constructive approach to embracing ChatGPT is discussing ethical ways that educators and students can incorporate generative AI in the classroom.
For example, teachers can use generative AI to their advantage to create different examples of text for student assignments, such as teaching grammar, research skills or providing subject-specific reading samples. Using the technology in this way helps save the teacher time to work one-on-one with students, prepare other assignments or spend time on creative ideation or activities that support emotional and mental wellbeing. ChatGPT can also help with content ideation for students who might need some inspiration to decide on an essay topic. Students could even be asked to write a more substantial, research-based writing sample based on a surface-level ChatGPT-produced work.
The hard reality educators, parents and students must grapple with is this: ChatGPT and future generative AI technology are here to stay, but that doesn’t mean we have to fear its capabilities. The best approach for educators is to learn about the technology and find ways to embrace evolving tools at the world’s disposal. As educators, it can be tricky to show students the value of advanced technology. But at the end of the day, it is up to teachers to inform their students about academic integrity standards, why it is important, and what the consequences can be if they use technology for dishonest purposes.
Deborah Rayow is Vice President of Product Management, Courseware at Imagine Learning. Connect with Deborah on LinkedIn.