Regina Martin, Chair of Denison’s Writing Committee, shares resources and ideas on teaching writing in the age of AI.
 |
In my W101, in most class periods, students do a graded in-class writing (ICW) assignment. A few weeks ago, I did an experiment: I required students to use a large language model (LLM) like ChatGPT or Gemini to do the ICW.
I gave them the following instructions: “Use an LLM of your choice to write a response to the ICW prompt.
Students frequently ask AI for feedback on their writing, so we need to teach them how to interpret that feedback. This week I am sharing some materials I developed to do that. In my W101, when students peer review rough drafts, I am now integrating lessons on AI feedback. I teach an 80-minute class, and during the first half, students work in pairs on a traditional peer review exercise.
 |
| I spent my summer developing a suite of AI literacy assignments that are appropriate for writing courses. Over the next several CfLT newsletters, I will share those. I developed them for my ENGL 342: Utopian and Dystopian Fiction class, which I am currently teaching. The first assignment consists of two parts. I’ll elaborate on the first part this week. |
 |
| This fall I have redesigned my W101 to experiment with new methods of assessment given that many students will use AI to write their essays. Because the quality of a written artifact is no longer necessarily a sign of student learning or a reflection of their ideas, I have shifted some of my assessment criteria to focus on evidence of student learning. |
 |
| An important aspect of addressing AI in writing instruction is helping students make better decisions about when and how they use AI to offload cognitive labor. Lori Kumler in ETS created a slide deck with a classroom activity for getting students to think critically about how AI use affects their learning and cognitive abilities. This activity would be appropriate for a class period early in the semester. |
 |
| As we enter year three of generative AI, where do you fall on the AI spectrum? Do you want to refuse AI? Adopt it minimally? Or Embrace it? This article, “How Can Your Courses be AI Aware,” links out to resources for making writing classes “AI aware” along each category on the spectrum. For example, if you want to resist AI in your writing classes, do so more intentionally. |
 |
| In my March 25 TTT contribution, I promised some suggestions for rethinking assessment to ensure students are meeting learning goals even if they (mis)use gen-AI to create their high-stakes writing assignments. And in my April 22 TTT contribution, I discussed an example of how I am re-weighting low-risk assessments. As promised, in this edition, I share a new type of assessment I am adding to my writing classes: paper conferences. |
 |
| In my March 25 TTT contribution, I promised some suggestions for how to rethink assessment. If a high-stakes writing assignment can be adequately created with gen-AI, assessing that piece of writing does not provide an accurate measure of how well a student has met a set of learning goals. Instead, we need to seek additional evidence. |
In a CfLT-sponsored webinar on assessment and gen-AI on March 3, Leon Furze (longtime educator, education administrator, and now Ph.D. candidate working on a dissertation on AI in writing instruction) shared an academic article, “Validity Matters More than Cheating,” which argues that academic integrity needs to focus on “assessment validity” rather than “cheating.” Cheating is a question of moral integrity, whereas validity is a process by which educators ensure that students have met learning outcomes.
My approach to AI in my writing courses is primarily to teach students how to make informed decisions about when it is appropriate and inappropriate to use AI. Part of this involves helping them understand what they do and don’t learn when they use AI. To that end, I include an AI policy statement on every assignment as well as a table that identifies what students learn if they use AI for some tasks vs.
