This includes using AI to write discussion posts and essays in which the professor wants to hear the student’s opinions and experiences. The results are often sterile, bland, and airless. But they are spelled and punctuated perfectly, free of grammatical mistakes.
Kaye Adkins, a professor of technical communication at Missouri Western State University, wrote about this dilemma in a Facebook group dedicated to discussing AI use in higher education. I asked Adkins if I could share her post because I think it captures the dilemma well: Should some traditional methods of grading be rethought in the age of AI?
“Does anyone else find themselves tempted to give a higher grade to a paper than they might have in the past because, although the paper has issues like fragments, run-ons, or poor organization—thus signaling it was not written with AI, the paper shows deep engagement in the topic?” she asked. “In first-year writing courses, I have given such papers Cs or Ds (depending on the point in the semester), but these days, I’m so relieved to see a student actually engage with a topic that it’s tempting to give them a higher grade than I would have pre-AI.
“I’ve been part of the content vs. correctness wars of the past,” she continued, “and I’ve always come down on the side of ‘the two can’t really be separated.’ But with AI, I’m wondering if that’s true anymore. Should we be thinking about going back to the bad old days of giving two grades on papers — one for form/ conventions and one for writing?”
In the comments on Adkins’ post, some respondents expressed concern that not teaching proper grammar, punctuation, and spelling can harm students. Others wrote that they are not grading as heavily on those components to encourage authentic writing.
I reached out to Laura Dumin, a professor of English and technical writing at the University of Central Oklahoma and founder of the Facebook group where Adkins posted, to get her take. She noted that a version of this grading debate has been going on for decades. Some academics have long argued that students should be allowed to write in their own voice, even if it’s not technically correct.
Dumin puts herself in that camp for first-year composition courses. She figures that if she can get students to first focus on what they want to say, they can work on grammar, spelling, and sentence structure later. One way to encourage that is by grading grammar as less important. “I think sometimes when students get so freaked out about having perfect grammar,” Dumin said, “they forget that you can have grammatically correct sentences that don’t say anything.”
Of course, context matters, she notes. A freshman composition course, for example, is one in which the instructor likely wants to encourage critical thinking and original ideas, while a technical-writing course emphasizes accuracy and clarity. Developmental writing might be another place, she said, where an emphasis on grammar and sentence structure actually builds confidence in the writer.
Many students enter college believing they are bad writers, Dumin noted, because their high-school teachers gave them low marks for failing to meet the more technical aspects of writing. That’s something college instructors should keep in mind.
“If you’ve got students who are coming out of that, who have often been told that their writing wasn’t good enough or wasn’t what the teacher was looking for,” she said, “they may be more tempted to use AI to write things because they feel like that’s what the instructor wants. And I’m very clear with my students that I want to hear their voices, that I privilege their voices.”
Dumin says she builds time into class to talk about how AI writing tools can needlessly alter perfectly good writing. She and her students also play around with AI tools in class so they understand how they work. She tells students it is OK to use such tools as long as they are transparent about how they did so.
I checked in with Adkins, too, to see whether she has made any changes to her grading criteria after posting her question last month. She has since added a category called minimum expectations that outlines the least a student can do to earn a D. “I’m still working,” she wrote, “on how to account for bland, AI writing.”
Have you changed your approach to writing assignments or your grading criteria to encourage students to write authentically? If so, let me know what you did and whether it worked. Write to me at beth.mcmurtrie@chronicle.com and your story may appear in a future newsletter.
ICYMI
- In this Chronicle Advice piece, Marc Watkins describes the rise of AI reading assistants and the challenges they present for learning.
- Meanwhile, in this Academic Platypus post on Substack, two professors describe how they asked students to use one such assistant, Google NotebookLM, to improve their comprehension of scholarly texts.
- More than a third of faculty members said they feel more constrained in their ability to speak freely while teaching, participating in faculty governance, or living as a citizen, according to a new survey by AAC&U and the AAUP. You can read The Chronicle’s coverage here.
Thanks for reading Teaching. If you have suggestions or ideas, please feel free to email us at beth.mcmurtrie@chronicle.com or beckie.supiano@chronicle.com.
— Beth
Learn more at our Teaching newsletter archive page.