“I’d rather see grammatical errors than read another paper written by AI.” Sincerely, Kristen Linton, MSW, Ph.D., a CSU Channel Islands, Health Science Department Associate Professor

I am reflecting with raw emotion as I am halfway through grading a midterm for an undergraduate health science research methods course. Just now, as I read my 21st paper of the day, I was glad to see an incomplete or run-on sentence while grading, because it was a sign the student didn’t use AI. This is a juxtaposition to me, even one year ago. Potentially, I would have marked this student’s grade down for improper grammar, but now I see grammatical errors positively, because it is a sign of his/her authenticity.

In the previous 20 papers I have read, at least 10 had significant signs that AI was the true author. Even though I provide a template for the midterm assignment, the AI papers often use different formatting, including single-spacing, no indent for the first word of a new paragraph, and bullet points. The most obvious is that students didn’t learn the terminology used in those papers in my course. Some terminology in the “suspected AI” papers would be consistent with a graduate-level education.

As an AI Digital Learning Mentor at CSU Channel Islands, I hold this role because I’m a huge fan of AI. You probably didn’t think that after reading the first two paragraphs of this blog. I have intentionally shared with my students’ appropriate uses of AI. As a health science Associate Professor, I know that the healthcare industry has welcomed AI, thus students need to know how to prompt and use it. However, before my midterm assignment, I specified that they should not use it for writing this assignment. I wanted to see their knowledge, not AI’s, on the research methods they have learned throughout the course.

Next semester, I will be teaching a Graduation Writing Assessment Requirement (GWAR) course. I write this blog as a reminder for myself as I prepare for that course, but also to provide writing assignment tips for you as we teach in an AI world.

Tips

1. Include an AI policy in your syllabus, such as “AI tools can support your writing by checking grammar and identifying typos. However, using AI to write or generate any part of your paper is not permitted.”
   • Tip 1 Justification: Justification for AI should be directly correlated with the purpose of the assignment. If the writing assignment intends to understand the student’s knowledge, then you want the student to write it themselves.
2. Have students write 10-minute “thinking drafts,” outlines, or reflection summaries in class or on Google Docs (for asynchronous courses), giving you access to editing history before submitting their full paper. Dr. Sonsoles de Lacalle (Chair of Health Sciences) shared this overall idea with me last week. Students are prompted in the assignment instructions to use their draft to write their final. You can compare topics, tone, syntax, and reasoning in the final submission.
   • Tip 2 Justification: This places students’ thoughts at the center of the assignment. Students need to be reminded that their ideas are highly valued. Students’ idea generation builds self-efficacy and confidence. Using AI to generate ideas does not.
3. Assignment prompts specifically asking students to apply course concepts and their real-life experience examples only.
   • Tip 3 Justification: I can easily detect an AI paper when the terminology used is highly advanced or slightly different terms compared to the content taught in the class.
4. Rubrics or grading should reflect priorities. For example, weigh the quality of analysis and integration of course concepts (used in video or in-person lecture- AI cannot fake that) higher. Weigh the surface-level polish or grammar (which AI can fake) lower.
   • Tip 4 Justification: AI can fake course concepts if available in PowerPoints or readings provided to students. They can upload them into paid AI software, such as GPT-5.

I truly hope the tips above help prepare for teaching in future semesters and design writing assignments in a way that prioritizes the voice and knowledge of our students. As AI becomes more popular, I see how we might be comforted by simple things, a typo or small grammatical error, as a sign of human creation. Imperfection is actually what makes life fun. I left a typo there just to show you I’m human.